IPP4 - Pre-press conference - 1997

iPP3 Proceedings of the third annual pre-press technology conference 1

Held at Bangalore on November 28, 29, and 30th 1996

Welcome to the delegates by Naresh Khanna

WELCOME TO INDIAN PRINTER AND Publishers third Prepress Technology Conference which we have been calling IPP3. The first conference was held in Delhi in 1994, the second one was held in Bombay, also in November, in 1995. This is the third one which we had promised to do in the South. And it is part of our effort to establish a regular technical plat- form in South Asia and although we hope to take the fourth pre-press conference to Delhi next year, we hope to take the fifth, to Kathmandu and eventually to Colombo and Dhaka. So far the conferences have been growing, the first one was two days and it was dedicated to newspapers and mag- azines only, in Delhi, and the second one was also two days but included commercial pre-press and a bit of packaging prepress and it had a small ex- position. The third conference has grown into three days and a little bit more on packaging and commercial pre-press, on non-Roman languages and Internet publishing and there is also an expo and so that means it grows to some extent every year. We are quite serious about a technical plat- form for the subcontinent and to make it happen it is essential to have an annual event because both the international calendar and even the national calendars have so much static. A good source of information is one aspect of the conferences’ role and the second aspect is that this is a techni- cal platform and that means a forum for techni- cal articulation for engineers in the industry, for the technologists in the industry to themselves become presenters, and for the users to become the presenters.

One way of learning about technology is from trade fairs, from salesmen, from brochures, from even being trained in various hands-on technical situations but the actual absorption of technolo- gy will require a community, a texture for us to share and that means we must go away from the

 “Over time, and we think it will take a long time to build these conferences, we must ask the senior organizations to understand that it is part of their duty and responsibility and in fact in their interest to allow their technical people to become articulate sharers and to interact with the community of technologists and engineers.”

situation that when an individual or a company broke its head and solved a problem the solution immediately become a trade secret. The individ- ual must share with the team and the team must share and interact with the industry at large. We want the industry in India to be mature and also in the subcontinent, to share, that means we must share with Dhaka, we must share with Nepal, we must share with organizations and individuals that want to improve themselves and show how we have absorbed new technology, how technology works for us, so that we develop this culture of sharing and absorption. Otherwise I don’t think that this subcontinent will really be able to mediate technology, it will not attract the best people and it will not be able to filter or deal with technology to its best advantage or cost effectively use or share

Introduction of Burjor Poonawala

I would like to introduce to you our keynote speaker, Mr. Burjor Poonawala and as you see his presentation is in the documents given to you, there is quite a bit of the information about him. Addi- tionally, Burjor is part of the important tradition of bringing colour technology to India and he brought this technology from Europe, from the West. Burjor, was among the people who brought film to India. Earlier glass sheets were used with various photo sensitive coatings. In his working life he had to push for the adoption of film, some- thing we all take for granted and now still in his own working lifetime he is seeing in the era in

“Burjor is one of those people who understood and implemented colour correction in terms of these mathematical relationships and also the first in our country to make the transition to the automation of scanning and colour correction. He is the person who most appreciates what the new technology is bringing and he is always eager about it and open to it.” our resources. We will keep buying what is obsolete and we will keep buying the last machine out of the door of the factory. We will keep running to Europe or America when the answer may be next door.

Our conferences have to be cost effective, we Europe don’t think the organizations in this part of the world can easily afford to send people to 1 or America, for six months or two years at a time to deal with problems of technology. It is not easy to even build up the conference and do all this, because the same people who are the technical hot-shots in the industry, because they are hot- shots, are very important in keeping their plants running or their projects moving forward. That means it is very difficult for a large commercial prepress company or a newspaper to spare two or three key people to come away for two days or three days to share with us. Over time, and we think it will take a long time to build these conferences, we must ask the senior organizations to under- stand that it is part of their duty and responsibility and in fact in their interest to allow their technical people to become articulate sharers and to interact with the community of technologists and engineers. Which people are talking about computer to plate and not having any film at all, going right from computer to press. Burjor is as you know, Director of Comart Litho, he is a hands-on technologist, more hands-on than one wishes for any one

When colour printing became very scientific people in Rochester at Kodak and various oth- er institutions working in colour technology developed a whole way of colour correction and masking which was mathematical such as the Nege hauer equations which were third and fourth order equations. The earliest use of computers for colour separations was basically to come up with the right filter combinations, exposure times and masking for colour correction. Burjor is one of those people who understood and implemented colour correction in terms of these mathematical relationships and also the first in our country to make the transition to the automation of scanning and colour correction. He is the person who most appreciates what the new technology brings and he is always eager about it and open to it.

Keynote address by Burjor Poonawalla

IT IS ALWAYS NICE TO HEAR NICE THINGS said about you and I have been asked to speak on insight to colour, insight to pre-press or more or less amalgamate the two and more or less it is the same subject because colour is pre-press and pre-press colour. In every walk of life today, we think, we see, we observe in terms of colour, on a television screen with cable movies and in- formation and if we sort of happen to see a pic- ture, say of a moving film in black and white, we just tend to switch it off, as oh, it’s in black and white. So we are very oriented to colour now. One doesn’t think in terms of black and white although black and white has its own sort of view point and it used to be interesting earlier and now today it is all colour and that is why perhaps we are going to give more emphasis on colour.

It is a privilege for me to be addressing this gathering and I thank my friend, printer, publish- er, editor, print consultant and organiser of this conference, Naresh Khanna for giving me the opportunity and the privilege and the honor. I shall not delve into the details of various process developments in depth, because presentations on the various subjects in detail will be made by the experts during the respective sessions. I suspect that I was chosen to address this gathering on this subject because I believe that like a vintage wine matured over a period of four decades in which time I have been in the business of graphic arts, and graphic reproduction, an industry in which revolutionary changes took place every 15 years, or some times every 15 months, or now-a-days even in 15 days. The number of years I have spent in this business makes me feel ancient when I see most of you young people in your 20s and 30s and perhaps early 40s. You wouldn’t believe me if I were to tell you that I joined my father in busi- ness when I was only 5 years old. That would make me 45 years old. But no. That’s too difficult to believe so I will let it pass, but I mention that I have witnessed the transition of a skilled craft pre- vailing in the first half of the century to a technology oriented process in the second half, which no longer requires craft and experience except technical skill. Thanks mainly to the age of digi- tal computers. We deal with digital computers, we deal with colour using digital computers and so we talk about digital colour.

The demand of those employed in the graph- ic industry has changed at all stages of the pre-press field. The professions of typesetting, reproduction, retouching, lithography, assembly and platemak- ing amalgamate into one profession these days. The highly experienced craftsmen of the past are now, like myself, relics of the 20th century. Young people entering into the graphic arts industry, once spent years of learning the trade from 

❝It is not enough anymore to know about ink-water balance in printing or creating a page with QuarkXpress or to know how to use a computer. A graphic arts operator will have to be completely at ease with both the computer and the learning of new skills throughout his career, as an ongoing process. So as technology changes if you don’t keep up with the technologies you are going to get left behind and perhaps, out of a job, I think.”

experienced hands. Today they operate computers which are rendering traditional craftsmanship obsolete. Today a pre-press entrepreneur and a technician must have a profound knowledge of soft- ware and hardware capabilities. He must also un- derstand the limitations of such technologies. Technology is evolving so fast, both in terms of hardware and software that by the time new equipment is purchased and installed it is already out- dated, not obsolete, but outdated. By the time you install it you are always already thinking of one of these trade journals where you read about some new developments and you feel you may have pur- chased too soon and that is if you waited three months you would get the latest but it doesn’t work in that way. You’d keep on waiting all your life, because changes keep taking place, technologies do improve. They get more compact and with more capacity and more user friendly. Cautious optimism about the market and a bewildering range of new technologies for communicating in colour make purchasing decisions very critical and they are critical indeed. You can make a mistake and find yourself in trouble and these technologies don’t come cheap. As this century comes to a close, the Graphic Arts industry must learn to absorb the new technologies profitably. The end must justify the means.

The industry must participate in the on-go ing transformation of traditional methods and tools into new digital ones. The advance of technology affects the graphic arts labour force as well. New tools mean new skills. It is not enough any more to know about ink-water balance in print- ing or creating a page with QuarkXpress or to know how to use a computer. A graphic arts operator will have to be completely at ease with both the computer and the learning of new skills throughout his career, as an ongoing process. So as technology changes if you don’t keep up with the technologies you are going to get left behind and perhaps, out of a job, I think. I believe in the West, it is an process, technologists and technicians continually attend evening classes to keep up with the technologies.

New technologies have already entered the markets abroad and in this country digital cameras, short run colour printers, personalised computer to print, computer to plate technologies and more are coming in. I shall make brief mention of these technologies as we go along. I will not talk of these in depth, not that I am competent to, but we have the experts here, and you will have an opportunity to ask questions, One can begin to appreciate the marvels of modern digital technology in any field, provided a person, like myself, has lived through the various phases of development over the years. One may relegate the past to history, but that would be folly. Because it is the awareness of the t past which helps us in understanding the intricacies of modern day technology. Further it is the comprehension which synthesizes the past, the present and the future. The graphic industry is no exception to this phenomenon. The progress and modernization of every industry has been accelerated by electronics, that is analogue earlier and digital as at present.

In 1950 the first huge IBM mainframe com- puters were developed, a piece of equipment worth millions of dollars, and useful in data pro- cessing. The next big upheaval came in the 1960’s when the mini computer arrived to open the computing market. This might interest some of you because it was also in the period when programmers and computer professionals became two distinct professions. 1981 saw the introduction of the personal computer popularly known as the PC causing a market revolution which among other things, produced many millionaires in the 200 billion dollar industry, with a growth rate of 13% or more per annum. Believe me, there must have been some who fell by the wayside during those days. In the old days, one imagined a computer to be worth millions, occupying a room full of electronic equipment, which was capable of relatively simple calculations. Today such computers into a shirt pocket and are capable far more complex calculations instantaneously at the press of a button.

Such innovations don’t mean much to a generation which has grown up with pocket calculators, VCRs, colour television, computer games and digital quartz watches, but for those of us who have existed and experienced life without these modern gadgets, we cannot but stop and marvel at these so called objects of desire. I need not dwell too much on these modern marvels since all of us are aware of their presence in the market, thanks to the media, in print and television, through which the consumer is exposed to the various advertisements and knowledge and in- formation, to bring about awareness of the avail- ability of the various products on the market so that the consumer can exercise his right to choose. That is the expression we often see and hear on television, advertising, the right to choose. You can exercise the right if there is communication and you have the knowledge. In the end it is always the consumer who reigns supreme. His likes and dislikes determine the success of many a project.

With the opening up and liberalization of the economy, advertising and marketing which was hitherto considered a futile expense in this country became an absolute necessity for the success of any enterprise or venture. Advertising on television and print media has now become the order of the day and inevitably so is colour which today plays a very important role. I am one of those who believes that no matter what strides and progress are made in the audio-visual medium of television, the demand for the printed word will never di- minish, but these days I often wonder whether it. will be true in the next century. In my lifetime- long having crossed the halfway point of no re- turn, I have witnessed the ushering in of the graph- ic films and sensitized glass dry plates. We are now about to witness their exit, in a limited way for the moment, considering that computer to plate and computer to print technologies already exist and are gaining worldwide acceptance. So, to our friends in the film manufacturing industry that is not a very good prospect but they have to learn to live with it and perhaps diversify into other technologies. One has to wait and see.

In this country, it is quite popular when you are in dilemma, in a problem you go and consult an astrologer. So I did it. The best of astrologers have failed to predict which direction technology will take in the future and although I am not a politician I have consulted the best. Colour is very alluring. Much progress has been made since the days of Mr. Alois Senefelder the father of li- thography who invented it, more by accident than by design, in 1798 or there about, that is, almost 200 years ago. The principle on which lithography is based is the mutual repulsion of grease and water as we all know. In spite of the technology advances over the past 200 years, the basic principle of lithography remains unchanged to this day.

It must be understood I was going to say few more words about Mr. Senefelder and how lithography was developed but I’ll be brief and men- tion that in those days there was no such thing as a reproduction. A lithograph produced from stone was an original itself. You needed an artist, a crafts- man to produce colour pictures. In the earlier days, one had to paint. If you made 50 pictures, each picture was painted by an artist in several colours. They all resembled each other but they were never the same. Then later on with the development of stone lithography artists started drawing on stone and they used to produce im- ages in colour using maybe 15 to 20 colours to produce a single image and the process took any- thing between 3 to 6 months. Such prints today are collectors items and there are not many that exist and they could produce maybe 40 or 50 in six months. So you can imagine that we have come a long way since those days and that whatever took between 3 to 6 months can be done in a few min- utes today. It must be understood that progress in lithography depended on the progress of several component elements, during the past 50 years, mainly.

The continued growth of lithography is not the result of one single advance but due to many improvements in every element and phase of the process. Direct lithography from metal plates nev- er too successful as an all round method in co- lour printing was more and more replaced by the offset press. Now to just make a brief mention of the offset press. Why did we change over from basic lithography, that means direct printing in lithography to offset lithography. The reason was colour.

“Today a student fresh from a printing institute needs not more than three weeks training to produce colour separations on the electronic scanner within minutes. The quality of scanner separations surpasses by far the prize winning reproduction of yesteryears. “

It was very difficult to produce colour prints by lithography. Main reason being dimensional stability of the paper used and the ink water balance. You have to use water in lithography as we all know and you have to dampen the printing plate and when a piece of paper came in direct contact with printing plate, not only did it lift off the image it also lifted off a bit of moisture from the plate. which means that the paper was not stable any more and the subsequent colours printed were not in register. Therefore, lithography was never popular in colour printing in the earlier days. In the earlier days it was always letterpress or gravure and there was another little known process called collotype. Collotype was very popular for print- ing colour. It almost gave us a continuous tone result with no halftone screen used. Letterpress was used for the four colour process using the conventional screens as we know them today. So what happened was that as printing developed I don’t know whether it was by accident or by intention that the offset press was developed. The impression from the printing plate was lifted off onto a rubber blanket. The rubber blanket had the affinity of attracting the image and the ink but it did not absorb a lot of water and when it transferred the print onto paper, the paper hardly got any moisture from the plate. Therefore, we succeeded in improving the dimensional stability of paper and so we got better registration and from those days onwards colour printing became more and more popular with offset lithography and less and less with letterpress. Till, in the 1970s, I think letterpress was almost completely overtaken by offset lithography.

Constant improvements in plates and presses, and also in photographic materials and colour correction masking techniques, paper, inks, roll- ers and blankets have made lithography what it is today. The offset press and its share of black and white and colour printing market grew and developed at an increasing rate. Press sizes became larger, and running speeds faster. Two colour, four colour and multi colour presses as perfectors became available. Today’s presses are available in many shapes and sizes but the most important innovation is the computerisation of various strin gent controls required on a high speed multiple colour printing machine to minimize wastage and improve consistency. Earlier there used to be a lot of inconsistency. The first print did not even match the 5000th print or the 10,000th print. It all depended on the machine minder to keep on adjusting the ink water balance and the flow of ink and everything to make sure that there was certain degree of consistency maintained. Well, it was a well known fact in those days that it was not possible to run off all the 10,000 prints which looked exactly the same. Not that we can do it now, but the degree of inconsistency is reduced con- siderably.

The demand for colour printing grew dramatically as a result of these advances in the seven- ties and eighties when the national newspapers started to switch over to printing by offset lithography. With the coming of the age of colour scanners, colour printing took a front seat in the industry. Suddenly there was a deluge of colour newspaper supplements, periodicals, magazines flooding the consumer market with news and advertisements setting off the age of information technology. They used to say that there was a certain sort of image, a look of the 80s and the 90s. It was totally different to the look of printing of the 50s and 60.

Insight to colour or prepress in the print medium needs to be dealt with on two fronts. Colour scanning or colour separation and print- ing. In order to fully appreciate colour scanning one needs to understand colour. It is a particularly complex phenomenon, made up of three basic components, a light source, an object and a human observer. The colour that we see is based on the amount and kind of light reflected from coloured objects, which means that if you see a coloured object in the tungsten light, it will give you a different impression to when you see it in tube light environment or a totally different im- pression when you see it in daylight. It also makes a specially in photography when you take pictures in the morning, afternoon and evening. The change of light has an effect on your final images. So light is very important to colour. Without light there is no colour.

It is important to remember that there are three ways to view colour. One is in nature around us, on the computer monitor or printed on paper. The widest range of colours is the visible spectrum. This spectrum consists of colours that most people are capable of seeing. Some of these co- lours can be encountered in nature, others visi- ble on paint that is man-made and like I said ear- lier, colour you see all around you. However, qual- ity of colour printing has always been dependent on colour corrected separation halftone film pos- itives which in the earlier days involved a rather complex and tedious process involving lengthy and time consuming photographic tone and co- lour correction masking and retouching tech- niques. Back in the 50s and 60s even an urgent job required at least two to three days of process- ing time, compared to what it is today just a few hours or sometimes even in minutes. Much de- pended on the experience, expertise and crafts- manship of personnel involved to produce quality work. Comparatively, today a student fresh from a printing institute needs not more than three weeks training to produce colour separations on the electronic scanner within minutes. The quality of scanner separations surpasses by far the prize winning reproduction of yesteryears. Electronic colour separation, otherwise known as colour scan- ning was used by us in India for the first time, I think in 1969 when our company installed the electronic engraving machine popularly known as Vario Klischograph manufactured by Dr. Hell, the same company as we know it today and was installed in our company and we successfully produced colour separation positives or letterpress plates in four colour, maybe two or three sets in a day. That was some achievements for those days because it took perhaps a week to produce a good set of colour separation positives or letterpress plates.

On this machine scanning of the original and the engraving took place simultaneously, produc- ing a marked improvement in quality of images and time taken to output a set of four colour plates within hours. The modern electronic digital scan- ner can do better and faster within minutes. All present day scanners separate coloured originals using three additive primary colours of light in the form of Blue, Green and Red to produce in a single pass yellow, magenta and cyan separations, black being generated with a combination of all three respectively.

In spite of the sophisticated and digital computerisation with complex built-in software, the principle and theory of colours in a scanner, as propounded by Clark Maxwell way back in 1890.

continues to hold good. That means the principle of colour separation as laid down theoretically by Clark Maxwell almost 100 years ago still holds good. We still need blue, green and red filters to separate out the colours. Those were the filters we used for conventional colour separation in the camera as well.

To say a few words about the development of scanners, the earliest scanners were what we called same size scanners, they could not enlarge or re- duce. That means a 35 mm transparency when scanned, produced a set of 4 colour 35 mm size separations and if you enlarged them they were not very good because you lost the sharpness and so another branch developed simultaneously and that was colour studios, Colour studios sprung up all over, wherever prepress was. They specialized in blowing up colour images from colour nega- tive colour transparencies to the required size for printing. So that if I got or received a 35 mm co- lour transparency from a customer, I’d it send it to a photographic studio and they would blow it up and they produced a fairly good match of the enlarged size in 10 by 8 inch or whatever size was required. You slapped that on to a same size co- lour scanner. You scan the image. You can get an excellent set of four colour separation positives or continuous tone negatives. You put them in contact with a contact screen, expose and produce a set of 4 colour halftone positives. Now that tech- nology was prevalent, I think in the seventies. Then they started making scanners which could enlarge but not reduce.

Those were mechanically sort of designed scanners so that enlargement took place mechanically. They had registration problems and so they did not last very long and then came in digital enlargement. We had an analogue scanner with dig- ital technology for enlarging so that we could enlarge images in perfect registration and it was still continuous tone. We could not use the screens.

“In spite of the sophisticated and digital computerisation with complex built-in software, the principle and theory of colours in a scanner, as propounded by Clark Maxwell way back in 1890 continues to hold good.”

We needed far more powerful light sources to expose the film. Then came the scanners with contact screen. They were known as contact screen scanners. So that we did not have to produce continuous tone images any more. You could enlarge and produce direct screen positives.

With the contact screen scanners you had the cumbersome process of introducing a contact screen in contact with the sensitized film and then expose that film through the contact film. It, of course, needed a powerful light sources to expose. Then the first digital scanner, to the best of my knowledge was put on the market by Crosfield and we were one of the first to go in for the digital colour scanner. One of the main reasons or the main causes for us to go in for a totally digital scan- ner was that you could store job values digitally, on analogue scanners that was not possible. With the result that eventually scanners as we know them today became totally digital in producing images as well as in enlarging and reducing.

One big advantage of electronic scanning is the consistency of optical density, range, sharp- ness of image, which means that quality separations can be produced on the scanner uniformly from day to day. Unlike the earlier analogue scanners, all manufacturers now offer laser dot generating scanners with plotters and with image-set- ters, which have many advantages in terms of high speed, economy and sharper details. During the past thirty odd years of development, scanners have come a long way. All goals have been achieved, full colour and final correction, complete control of range compression, under colour removal, gray component replacement, almost unlimited en- largement and reduction, direct electronic screening, independent input and output through systems and image setters and much more.

High end colour scanners, invariably drum based are now getting flat. Drum scanners have dominated the high-end quality colour separation market for the last two decades. The introduction of the flatbed scanner with CCD technology in the late 1980s has now begun to challenge the drum based scanners in terms of most applications, quality, speed, flexibility, economy, user friendli- ness and productivity. The comparison between, or the advantages and limitations between the drum based scanner and flatbed scanners is de- batable. There is no clear cut solution that is, ah yes, this is where I use a drum based scanner and this is where I use a flatbed scanner. Both produce good quality, but there are advantages and disadvantages. I think one of the panelists will be discussing that a little later.

Digital scanners gave rise to page makeup and retouching work stations. It became possible to record and store digital data on magnetic tape or on the disc for later reuse, to transmit the data information from one device in or on to anoth- er, location-wise. Also, to control graphic place- ment of data, to merge with other digitally stored data and visually manipulate with special effects, and to retouch images by altering or modifying data. All this became possible, thanks to digital in- formation.

The latest innovation is the development of the frequency modulated, or the random dot or the stochastic screening method for printing with- out any fear of moiré screen interference patterns. The conventional halftone screen angles which have been in use for over a hundred years and still continue to remain the popular screening sys- tem has its limitations. They are most inherently susceptible to interference with finer details of the image and limit the number of colours that can be printed to the conventional four colours. It is a well known and established fact that the pigments, for colour printing inks for cyan, magenta and yellow are far from perfect and hence can- not reproduce the brighter and cleaner colours. You may have observed that clean oranges, blues, greens, emeralds, pinks etc. are lost in the con- ventional four colour process. However, a wider spectrum of clean colours can be obtained if it is possible to print with more than four colours de- pending on the nature of colours required to be reproduced from the original. Usually an addition of blue, green and red should suffice. With ran- dom dot screening it is possible to print more than four colours without fear of moiré patterns. The printed result with this new technology produc- es a result closer to continuous tone to resemble collotype prints from the thirties and forties. This process of reproducing colour images covering a wider spectrum of colours is popularly known in the trade as “Hi-fi colour”. Soon this technique will gain momentum and perhaps by the begin- ning of the 21st century it will gain precedence over the 4 colour process. Now this is quite a possibility, a lot of scanner manufacturers are looking at this very seriously and soon perhaps press manufacturers might go in for 6 or 7 colour or 8 colour presses which exist today but perhaps for different purposes. But these press might be required for printing more than four colours.

Minimizing downtime, reducing fixed costs and increasing throughput are the key competitive factors to any prepress printing operation. In order to increase throughput and speedup productivity imagesetters can now expose a line and halftone image directly onto the printing plate, thus, wholly eliminating the intermediate stage of producing film positives. There is minimum dot loss on the plate which means sharper and more details in the highlight. Although plate-setters as they are known, are very expensive for the mo- ment, the prices are likely to come down as the demand and their popularity increases. Direct to plate saves labour and film costs and has become popular with newspaper printers especially as well as book publishers. Colour and black and white files can be prepared, imported or received digi- tally on disc or through ISDN, that is international services digital network, telephone lines as well as by satellite, especially when operations of editorial and press are located in places distantly apart. Time of India is a case in point. I shall elaborate a little further on ISDN a little later in my talk. The focus of distribution of various international publications such as Time and Newsweek who used to print in one centre and then distribute are now shifting to different printing locations and centres all over the world and confining distribution to local agencies. So that the same magazine is print- ed in various parts of the world producing the same image consistently and with the greatest of speed, because all are published at the same time on the same day.

Of course, it should become more economical as well, depending on how popular the process becomes. Increasingly, print has to compete with other media like television, computer based interactive presentations etc. Further, the Inter- net and world wide web which are virtually break- ing the sound barrier in the communication and information media, of which the printed word must continue to remain a part of and coexist. In many markets abroad, this is creating a new de- mand for short-run colour printing personalisa- tion, just in time and on demand, more targeted runs and faster turn-arounds.

Such a market is being created thanks to the digital computer to press developed by several manufacturers since 1992. Until a few years ago, there were almost no options for getting good quality colour in small quantities. Then in 1992 the well known company of Heidelberg essentially revolutionised colour printing with the GTO DI digital colour press. This press which still uses printing plates but images them on press directly allows digital data to be sent directly to the print- ing machine. A year later a company called Indi- go and Agfa of Germany, after years of research and development introduced digital presses that totally eliminated printing plates by transferring images directly onto a reusable drum from a com- puter and further advanced the concept of short- run colour printing by making runs of even just one piece possible by making runs of even just one piece. The end does not justify the means but to- day, it is possible to produce just one print with- out having to waste any sheets of paper for regis- tration or any thing of the sort or adjusting or con- trolling the machine.

Since then there have been a number of de- vices from various vendors that address the short run colour market from an expanding selection of colour copiers, printers to the new smaller format digital presses from Scitex, Xerox, Fuji and others. A new trend has developed of late with wide format colour inkjet printing technology carving out a niche in the market. Images produced and stored digitally on a disc can be downloaded onto the PC Server which drives an ink jet printer or electrostatic printers. It is usually wide format from 1 metre to 5 metres wide, primarily used for printing in colour limited runs from 12 to 300 DPI for indoor posters to single piece billboards in high quality and performance. The 36 inch printer can also produce large tiled images for hoardings, posters, banners and can also perform step and repeat functions. It can print on various substrates including glossy paper, vinyl back lit film, transparent film as well as canvas. This is what they call non- impact printing because it’s just an ink jet, there is no printing plate, there is no contact with pa- per, just an ink-jet producing the image.

“All goals have been achieved, full colour and final correction, complete control of range compression, under colour removal, gray component replacement, almost unlimited enlargement and reduction, direct electronic screening, independent input and output through systems and image setters and much more.”

The demand for colour is growing exponentially. In fact, with the latest innovation in photography, it is now possible to capture an image in colour directly onto a host computer screen through a digital camera. While digital raphy makes image capture faster and cheaper, the existing graphic arts business model does not always enable the exploration of these advantag- es. There are several manufacturers world wide that allow you to shoot 3-D objects much like a regular camera except that you don’t have to wait for film developing and processing. Portable units are presently limited in storage and image size and hence suitable for images up to post card size only. The studio models are more bulky and offer larger image formats and dynamic range with fairly long exposures but require attachment to a host com- puter. This technology is not yet a substitute for conventional reversal photography although it may reduce downtime. Further developments and research are in the pipeline. I don’t believe it will be long before digital photography replaces the conventional one as we know it today. Photographic colour images are already available in digitized format on CD-ROMs, that is read only memory. It is now quite possible and practical to shoot objects with a digital camera and mix the image into a page layout with the text keyed-in in colour, and output a wide format colour poster, on demand, all in a day’s work. It would be normal, I mean it wouldn’t be a rush job as such. The equipment is designed to produce such jobs and you can justify producing just one poster in four colours. That was hitherto not possible or practical.

All colour images can be digitally stored on magneto optical discs or Syquest cartridge, or Zip drive or several such devices available on the mar- ket. DLT or the Digital Linear Tape is the newest media on the market. The image information is stored in bits or pixels and an A-4 size image in 300 DPI can occupy approximately 25 Mb of space. In a regular, fairly large size system environment with multiple workstations there can be a fair amount of picture information being archived and retrieved from scanning stations to the outputting through image setters, plotters and other devic- es not necessarily from the same manufacturers. It becomes a fairly complex operation and if not carefully monitored can cause confusion.

OPI that is open prepress interface and data management are important issues which need to be addressed on even small configurations. With customers becoming more demanding for 

“think the frequency of such conferences has to be maintained to at least once a year, because that’s when we can exchange thoughts, ideas, share our knowledge and be better off to that extent”

throughput, quality and price on the one hand, and the increasing number of services on the oth- er, offering a greater variety of graphic arts ser- vices requires a very efficient operation. Magazines, newspapers, catalogues, packaging and advertising have their own sets of production requirements and their own unique workflows. Data man- agement and archiving are further complicated by multiple hardware platforms, multiple file types and multiple storage media such as on-line hard discs, off-line magnetic tapes, optical discs, Pho- to CDs, Exabyte, Syquest etc. In such digital environments, precious production time is wasted just finding and moving data from one location to an- other. Data management and archiving solutions are now made available to ensure smooth and un- interrupted workflow. Such devices as CD-ROMs and MODs, ink box, DLT are large storage and archiving solutions.

This means additional financial investment for strategic digital data management which involves networking with servers, database search software, storage media and ongoing customer training, which is always an integral part of any solution to- day. Apart from data management, colour management has also become a prime necessity. For a long complexity of exchanging colour data across platforms within workflows and across media prevented colour becoming a commodi- ty. The question very often asked is, will I see on the screen what I will get on paper? No! That in my experience has never happened.

You will see image on a colour screen that resembles what you get on paper but it will never be exact. It cannot be exact unless you are view- ing an image through the same medium. Here you are viewing an image with an additive process of colour lights. You are viewing the printed im- age with subtractive process of colour lights and they will although they might resemble they will never be the same. So what you see is what you get is really a tall claim.

Colour management software is a means of controlling and adjusting colour information throughout a multi workstation system. One of the problems facing today’s colour imaging industry is that there are many different kinds of systems

“While digital photography makes image capture faster and cheaper, the existing graphic arts business model does not always enable the exploration of these advantages.”

that work in different ways and so a unified, industry-wide, colour managed environment needs to be created, in other words we are looking for some type of standardization. Eventually the specific aim of colour management is to bring about a uniformity of colour images and information across different media, scanners, monitors, digi- tal proofing devices, plotters and image setters. Colour management systems have been developed and solutions being offered by various companies like Color-blind, ColorSync, Kodak and a few oth- ers. Thus software and hardware manufacturers are working towards a standardized solution.

The day is soon approaching when it will be common practice to exchange colour files through modem between clients, prepress studio and print- er or with a remote unit by wire, optical cable or satellite. Such a facility is available in a limited way through MTNL, that is our highly respected tele- phone system in this country through ISDN lines, that is Integrated Service Digital Network lines, which will transmit data at 64 Kb per second and more. For example an A4 size image of 24 MB can take 6 to 7 minutes to transmit from one lo- cation to another, this is what I presume, I have not tried it. However, the speed of transmission can be improved if the image data is compressed 5 to 6 times without visual loss of information or detail. There are several data compression soft- ware available on the market. Data can also be compressed and decompressed for final use.

ISDN is becoming a world wide accepted stan- dard. Information can be exchanged through ISDN for quick error free, cheap communication. It takes one tenth of the time to transfer a file on ISDN as it would with a modem over ordinary lines. In the fast expanding international and In- dian markets, more publishers start off by print- ing their publications in various locations, rath- er than shipping their printed material daily by road and air.

The ability to perform remote proofing also speeds up production cycles tremendously. Dur- ing initial stages of production, files can be transferred to remote sites such as an advertising agency for soft proofing on screen. In reality, for final proofing prepress shops and publishers still may have to rely on the actual printed proof or a Cromalin or hard copy digital proof. The other day I read a news items very aptly captioned and I quote With progress in technology, distance will become a thing of the past. The article goes on to state that with advances in information and communi- cation technology, distance is increasingly becoming a thing of the past in medicine. Just like it has already happened in prepress, computers are find- ing their way into several areas of the medical pro- fession, including diagnostics and patient monitoring. It is felt that the case of electronic transferability of data and images will change the face of the medical profession. Now this is what the medical profession claims and it’s already starting to happen in this country. Depending on the size of the files it can take minutes or even hours for transferring images, and therefore dedicated lines become a must for regular users, mainly large newspapers and publishers. The digital information so received can be worked upon, modified, corrected and sent back to the station of origin or can be output on a digital proofer or imagesetter or platesetter. This means that it will no longer be necessary to transport colour separation film positives to the printer and instead colour image in- formation can be downloaded onto an archiving device and the cassette or disc can be forwarded to the printer who can then read the files to pro- duce his own film positives.

Time is running out, one can go on and on and yet not cover all the aspects of this vast subject. I have tried to put forward a general picture of the digital technology involved in colour pre- press. However, if there are any specific questions and if there is time and it is within my competence, I will try and answer them but I think we can leave them for the later sessions.

You will agree that we have come a long way from the days when prepress was a product of skill, craft and experience upto the present age of digital colour prepress. I trust I have given an ‘Insight to Digital Colour’ in prepress as we know it today: I have made submissions and I am open to questions later on.

I extend a hearty welcome to all the delegates here and I thank my friend Naresh once again for having organized such a conference and since it is on-going process of learning, I think the frequen- cy of such conferences has to be maintained to at least once a year, because that’s when we can exchange thoughts, ideas, share our knowledge and be better off to that extent.

Proceedings of the third annual pre-press technology conference Part 2

Held at Bangalore on November 28, 29, and 30th 1996

Introduction of Peter Janusik by Burjor Poonawala

I have indeed pleasure to introduce Mr. Peter Janusik. His bio-data has been given to you in the conference kit and as I have been talking to him I feel like I have known him a longer time than those few minutes talking to him. The reason be- ing that he has started his career at the grassroots level. I was surprised to learn from him that he started his career with hand composition. Now that is way back to the sixteenth century. When I also started my career a little earlier, using the wet collodion process which nobody has heard of today, it was a process of coating a plate with sensitized liquid called ‘collodian’ and then exposing it in a camera in the wet state, developing and process- ing the image while it is still wet. The solutional coating that coated the glass is still wet and then the image was fixed. It took about 15 to 20 min- utes to produce the image on a glass plate. Now this technology of producing negatives was prevalent during the time when hand composition was quite in vogue and when I dealt with representatives of advertising agencies a few years ago, it was hard for me to convince them to switch over from hand composition to phototypesetting which was a new technology which was available. I told them it will give you faster turn-around time, better quality, and it will be a cleaner process. Now, Pe- ter was born in 1964, completing schooling in Mel- bourne, Australia,  and served apprenticeship as hand and machine compositor at the Victorian Government Printing Office in Melbourne. That is what I call starting from grassroot level. He worked in all aspects of composition, including Linotype, Monotype, hand-setting, digital type- setting, such as Penta Systems and Macintosh PostScript systems. After the Government print- ing office, he moved to Ogilvy and Mather Advertising, where he worked on high quality press advertising on Berthold equipment. Following Ogilvy and Mather, he spent time at Mattingly Advertising, where he was involved in the com- missioning of a large Berthold installation there. During this time he got more involved in Systems Management. Now this is the progress of his career, he started as a compositor and now he is moving into prepress, electronics, comput- er and digital technology. In 1987, he moved to Sydney and was employed by a start-up company to commission and manage the typesetting area using Penta Systems and Linotype image- setters. As I said earlier, typesetting is no longer an individual profession. It was in the eight- ies. Now it is merged with prepress technology which starts, I think, from advertising agencies onwards. At the end of this project Peter moved out of the industry and became involved in com- puter sales, and finally ended up at Delairco, who were the Australian distributors for Lino- type equipment. Now he is going to be in In- dia with EAC Graphics. We will welcome him with his knowledge and we look forward to shar- ing his experience with us.

Peter Janusik on Considerations in choosing a high end scanner

Thank you for your introduction. EAC Graphics has decided in India to consolidate their prepress business into a single business unit with respon- sibility being held by one person, for sales, service, spare parts, application support, etc. The idea behind this is to merge it into one unit merged in one year, where we can keep full control, and we can start to develop the business into a system integration business. Customers felt that it was a lot easier to deal with one vendor for their hub, for their scanner, for their imagesetter, etc. So, I’ll be looking after the Linotype Hell part of this business in India for EAC Graphics and also for the Heidelberg DI press part as well. So in the future please don’t hesitate to give me a call with anything at all I can help you with in prepress

The topic that I was kindly asked to talk on is “Considerations in choosing a high end scanner for high end colour”. I had little bit of a think about this because it’s all very easy to start talk- ing technical in conferences without really knowing at what level the expectations of the audience are. One person out there may already know everything that I am going to talk about whereas some others will not.

We’ll look at what is high end colour and I think we can maybe look at trying to define it but perhaps not in the way you would think. We can look into it in two ways. We can say high end co- lour is first class reproduction that is technically excellent. And if we look at it that way we can make a definitive judgment. But we can look at it an- other way. We can say it’s a result that exceeds or achieves your, and your customers’ expectations. Because that is what you can sell and make money from.

The major part of our presentation today is that you don’t always have to look for the most expensive tool or the ones with all the bells and all the whistles because that’s not what’s always going to give you your best value for money when you’re out there producing and trying to make some money from your business. 

What we want to look towards more is find- ing something that fulfills all of your requirements that you need and your customers’ expectations which you should carefully outline before you start looking for a scanner. And by high end scanner I suppose it’s little bit hard to put into a category but we are talking about the top end of the flat- bed market and any of the drum scanners. We are going to try and look at that a little bit more to- day and try and get you thinking about if you’re looking for scanner, to put some things in writ- ing as to what you want to achieve and start focusing on the things that are going to make a difference to your business in the long term, not on short term things such as price. 

We all know how confusing it can be to go out and look at equipment; it doesn’t matter what it is, for example, looking for a new PC. I was in the office the other day and the IT man came in. He brought a lot of brochures and I started go- ing through them and all the specifications look the same and I wound up going for a model that I knew back from Australia because if you just look at the specifications it can be very very difficult. There are many vendors out there in the market today looking for your business as far as scanners and it can start to get very confusing very quickly, believe me, it is no different here than in Austra- lia I’m sure. There’s a lot of people out there look- ing for your business, they’re all wanting to make some money as well and that’s including ourselves. 

It’s most important when you are out there looking for a vendor that you start a partnership with someone and you start a dialogue with them, so that they start to understand what you want to  “We can say high end colour is first class reproduction that is technically excellent. And if we look at it that way we can make a definitive judgment. But we can look at it another way. We can say it’s a result that exceeds or achieves your, and your customers’ expectations. Because that is what you can sell and make money from 35 achieve because if you do not give them the in- formation they need they would just put scanner- X or scanner-Y in front of you and your decision may not be the one that is the right one. But if you start to form a partnership about not only the scanner, but about how it fits in with your busi- ness plan and your ongoing business, then per- haps you’ll end up with the right machine, and you may be thinking yeah, but there’s not that many scanners on the market. Yes, but there are. There’s, I don’t know, 20, 30, 40 once you really start looking around. They’re all a little bit different. You must be careful but it is not that hard if you look at it in a very clear and methodical manner.

So today I think we can start off with some discussion points and this list by no means is the be all and end all. These are some things that I put on when I tried to put together a presenta- tion that wasn’t going to be too long and put everyone to sleep but that wasn’t going to be too short either. What I’d like for those of you who are out in the marketplace in the near future, you may want to start thinking about some of these points but you also may want to add to them and start compiling a checklist, which you can use that is suited towards your business that you can use as a starting point in your discussion with your busi- ness partner, the vendor of the scanner. And that way you will start off in a very nice even footing and you’ll be able to make some very informed judgments as time goes on. So some discussion points here, beware of specifications. Don’t always, well never, ever, make a decision on those alone. Because you’ll find that quite a lot of the specifi- cations are exactly the same. You’ll have to delve quite a bit deeper into the issues in hand.

Another very interesting discussion point. Flatbed or drum-I have couple of slides on this but I do not intend to answer the question. I sup- pose I can add a little bit to it. But again I think, it is looking at what your business needs and what your requirements are and at the end of the day

I think that they both have their place. 

I think another thing which we’ll touch upon is the ability of the scanner to save in various for- mats. And maybe the printers are out there saying: so what, all we want is CMYK. We’ll touch upon why CMYK is not good enough these days. So that’s another little point we’ll have a chat about. 

Productivity, is a very very big issue, and very hot because in the clamour for scanner vendors to build new and sexy user interfaces, that makes everyone very excited when they use them on their Mac, you have to look and make sure that these things can produce the amount of scans per hour that you require to make some money. So that’s another issue we will have a little bit of a talk about.

Colour management is a topic that’s already been mentioned once this morning and I am sure it will be mentioned almost every hour on the hour for the next three days. It is important and it is still very confusing and I think the best way to look at it is to start going down the road. It is not an expensive thing to get into because it is mainly small modules of software. So it’s easy to experi- ment and to learn from it. 

And I think the next important thing is the level of support from the vendor and that has to be driven by the customer base. I think a lot of the vendors out there now are looking much much more seriously at the support they’re going to have to give to their customers and that doesn’t mean just in spare parts-it’s the training that is given to the applications staff that trains you the way the business is handled, the coherence, networking, servers all that sort of thing is back- up and support to enable you to put in a system which is going to work 100% for you. So there are my discussion points. I am sure there’s a lot more as well, but I picked enough to fit onto one slide.


So, back to the specifications debate. I remember going back a few years when desktop flatbed scanners were released, and I’m sure a lot of you people do as well. A lot of the people manufacturing these said, the drum scanner is dead. And these little scanners that are out there for four or five thousand US dollars, these are going to save the world from these horrible big drum scanners that were operated by these operators who knew this black art. But it is not like that anymore because these all had nice things and it’s all easy to get consumed by the specifications. DPI in a scanner is not so important unless you are looking for a high enlargement factor. What you want to know when you are looking at a scanner is, can it capture the information you are trying to give it? And the term you really want to look at is the Dmax of the scan- ner and what sort of range it can capture? Because if you can get the information in and process it, then you can almost certainly be able to do some- thing with it. So it’s very very important to look at that one particular item especially in a flatbed scan- ner. On a drum scanner it really should not be too much of an issue. Most drum scanners these days have that range and that’s why they’re still around because they have a very large range. They can capture just about all the information that’s presented to them. So for a flatbed, be very care- ful that you’ve got a high Dmax so that it can capture the information that’s presented to it.

The user interface on the Macintosh is very very important. I mentioned this a little bit earli- er and I am not going too much into the productivity issue here and of all the tools that may come with it but just the simple user interface to do one simple scan. And the question is going to arise be- cause most people are migrating from a scanner that may be a drum scanner. It may be a Crosfield or with a bit of luck it may be from a Crosfield to a Hell or something like that. But can the opera- tor that you are going to have work this scanner, can he sit down and make a quick scan? Can he do it the same way he could do it on a drum scan- ner? I have heard quite often before from many customers in Australia after they get their scanner that is controlled by a Macintosh, “I can’t do a scan as quickly any more.” There are two things that can be relevant here. If the operator has not taken a big step into the Mac envithat a issue you look at very very carefully. Can scanner operator be trained on the Mac? Does he want to be trained onto the Macintosh and secondly, what’s being presented to him, does it really al- low him to do the job you are asking? So it’s very very important that we look at that carefully.

A couple of points. I mentioned Dmax before. You should expect to be at a minimum of four when you are looking at specifications on scanners. That should be around about where you are looking at for some quality. Anything less than that and you’re probably not going to capture the in- formation that you’re going to need and your results may be a little bit disappointing. So the points on this one. Specifications. Be careful what you look at, with DPI, and all that sort of thing, be care- ful of the user interface, make sure that your operator can be at home with it very quickly, because if he can’t it’s going to cost you a lot of time. And the third point here, that I believe is very important is that before you go head fast into talking with the vendors, about “I need a scanner, send me a proposal”, why don’t you send them your proposal first, and say what your requirements are and send them a page. They are going to be so happy to see that. It makes the job a little easier but it also means that they get a feeling for what you are trying to achieve and it is so important. In fact I got a letter the other day, my second day in the office, my friend Jacob who I’ll be work- ing with, handed me a letter from a customer which said: “please send me a proposal for pre- press equipment. And this outlines what not to do. We don’t know anything about this guy, I could give him anything from a system worth, I don’t know, 150,000 US$ to a million US$. So where do I start?” I’m throwing it back at you as the customer, to help us as the vendor to help us give you the right solution so that we can be seen to be professional as well.

Flatbeds and drums

So we get to the interesting one about flatbeds and drums and all that sort of thing. Almost all the scanners that you see in the market can beeated by a Macintosh either directly or oper- indirectly. In fact, I have to say I made a bit of a mistake in this slide, for any of you PC fanatics out there I should have mentioned PCs as well. Because it is important, I mean, the Windows 95 interface or the Macintosh interface, we have scanners that operate by both, but if they are operating by what we call an open system, that’s fine. Windows or Mac interface. If it’s not something that is com- mon like that, you really would have to look at that and say why isn’t it migrated to one of these great open platforms that we can use, and if it hasn’t, give it some second thoughts, I would say. Because there is no need anymore for proprietary systems controlling a box as simple as a scanner. At Linotype-Hell all our scanners are either con- trolled by Macintosh or Windows PCs. And that is from the top of the range 3900 which is PC driv- en down to the Topaz and even Saphirs, the small flatbed which is driven by the Macintosh OS sys- tem. So bear that in mind, you really don’t want to reinvent the wheel and learn some weird op- erating system just to operate a scanner. Your scan- ner should come with a software that operates with one of those systems.

So we’ll keep going on this flatbed versus drum debate. It is a very difficult topic to talk on because we are not a vendor that has our foot in one camp or the other. We have our foot quite firmly in both. And I think for good reasons. But let’s just talk a little about flatbeds at the moment because drum scanners have been around for a while. And I think flatbeds are still a good topic of conversation as to where they’re going and how they’re doing. At the top end of the range the flat- bed scanners are not toys, they do come with some great technology. First of all I mentioned that Dmax and user interface is very important. Keep again very aware of that even when you’re buying a flatbed scanner.

The CCD technology that drives these flatbed scanners is technology in which a lot of R&D is still going in, whereas photomultiplier technolo- gy is not really R&D driven so much any more. The work’s been done, it’s happened. And sure you’ll probably see new drum scanners come out but the actual technology that goes into capturing the data probably hasn’t changed a lot. 

You’ll see things like faster scanners, you’ll see new elec- tronics and all that sort of thing, but inherently the technology doesn’t change, whereas when a new flatbed scanner comes out, the CCD array generally will be a new one, it will be upgraded and it will have different and new specifications. I personally think that the flatbed versus drum argument may die off in the next few years and I don’t think either one is going to disappear by the way. Both have their places. And with the advances in CCD technology and that will mean that some flatbed scanners, not all, some of the high end flatbed scanners will start producing quality rivaling some drum scanners, and again, some drum scanners only.

So if you like there may be a crossover matrix here. But again the point I want to make is that it really doesn’t matter whether flatbed is better or drum is better, that is not the point of my discussion today. The point I want to make is, which fits your requirements better? Is the flatbed and the samples you get off that, good enough for what you want to do? It may very well be. It may be you have a space limitation and the flat- bed fits into that and the quality is fine as well. That’s great. But it may be you need the very high 

“The CCD technology that drives these flatbed scanners is technology in which a lot of R&D is still going in, whereas photomultiplier technology is not really R&D driven so much any more. The work’s been done, it’s happened. And sure you’ll probably see new drum scanners come out but the actual technology that goes into capturing the data probably hasn’t changed a lot.”

productivity and very high quality that only a big drum scanner can give you. That’s okay as well. But I think it’s dangerous to say, “Hey flatbeds are no good” and try and shove them aside. And it’s time for repro people to perhaps get off their high horse a little bit and accept that flatbeds are out there and they have that spot in the market and they do an excellent job especially for the lower dollar value that’s being asked for them. So that’s pretty important.

Productivity tools should also be examined when you’re looking at flatbeds versus drums. And I say this that because a lot of the flatbeds come with some new and sexy software whereas some of the drums tend to retain the old user interfac- es as well. I think also productivity tools should be looked at-batch scanning is an absolute necessity. If your scanner doesn’t have that, your pro- posed scanner, don’t even think about it because you’ll find it will be slower than the operator working your traditional drum scanner. Some of the new flatbed scanners have some pretty interesting tools as well. Some of the new ones have the ability to take a film supplied from your imagesetter or a set of films, scan them and make them into a TIFF DCS image and then incorporate that in imposition. So then, all of a sudden you can stop patching up so many films on the bench. This is brand new. Some releases are happening this year. One of ours, one our competitors released a while ago. It is pretty exciting technology, very simple to use as well. Our product is called the Topaz Copix and it’s selling very well. We’ve sold a few in Australia. And we hope to introduce it to In- dia fairly soon as well. But that sort of technolo- gy allows your flatbed to become more than just a scanner for your colour transparencies or your black and whites. All of a sudden that advertise- ment that is the only missing part in that page imposition can be scanned, incorporated, and you can output one whole film complete. It’s fairly attractive and it will be interesting to see how the marketplace in India accepts it.

Why scan in CMYK?

So this is an interesting one as well. CMYK. Those of you in the audience who are printers may think this guy’s a bit silly putting up a slide that says, “Why scan in CMYK” But the world is changing dra- matically, and the reason I’ve said this is based upon some experiences I had back in Australia. with some of our very large customers who decided that they need databases. They went through the process of looking at how do we store our images, we’re used to CMYK. It’s no longer the case that CMYK is the way that your file may be used best. I’ve noticed since arriving in India that the Internet is a hot topic here. Just give me an idea how many have Internet access or are looking at in the near future? Not many. Oh, slowly the hands come up! OK. But I mean I think it’s indicative. It’s happening and it’s happening quickly, and a lot of printers back in Australia have supplemented their sales having a Web page that’ll allow peo- ple to do an estimate themselves, a quote online. So that they’d enter a few of their requirements and they’d get a rough estimate of what the cost of the job was going to be. That’s pretty interest- ing and that really came from the States where a lot of the big printers and repro houses offer almost their full services over the Internet as well. 

But I’ve diverged a bit from the point of why scan in GYMK? I’ve got something here that says “but what about multimedia and the Internet?” And I think that it’s really important that we touch upon that. Because when you do a scan into CMYK you throw away a lot of data and a lot of colour information and it makes it fairly flat to reproduce a CMYK back into RGB to put it into amultime dia presentation or your web page or any one of the many other ways that you can use that scan these days. So the better way may be that you would have your scanner scanning into RGB or Ciclab and store that in your database using a colour man- agement system and then only when the job is output at the RIP, be it a film recorder, be it an imagesetter, or even be back to a monitor that the colour management takes care of looking after what happens to that particular image to that particular output. And we’ll talk a bit more about colour management. The key is just be careful, if you want to set up a database, make sure the scan- ner purchase you’re looking at can scan and save into things like RGB or Cielab in particular, is a very good colour space to use if you’re looking at an image database. Don’t start a database in CMYK, you’ll end up being quite sorry about it.

Image databases are best stored when you have the maximum amount of information before you start throwing things away. When you’re looking at a scanner and when you’re looking at, “well, how can I save my file when I’ve done it”, don’t think CMYK alone, think about RGB, Cielab, Sci- tex CT, think about Kodak YCC format. I mean, all of these things, you don’t know when you’re going to use it, but believe me, you’ll regret it the day you don’t have it. So that’s another good point to look at when you’re looking at a scanner.


Often forgotten and always missed! It’s something I thought about on the plane coming over. And it really is a bit like that. When you talk to custom- ers, and I think our customers are the same in Australia, they’re the same here. Sometimes it’s very easy to focus on things like price to be hon- est, but make sure that a new purchase can fit into what you want to do, that your workflow is right and that the scanner suits the workflow that you want to use with it.

Background scanning. I mentioned it once before. It’s absolutely essential that the scanner can be scanning to a hard disk or a file server or wherever while you’re checking up on a prescan. It’s absolutely essential, and if it doesn’t have it, don’t think about buying it. This is fairly new. Twelve months ago a lot of the scanners on the market didn’t have this feature, but most of them now do and those that don’t will either have the software fairly soon and if they don’t intend having it, don’t look at buying it, again, it’s just too important.

If you don’t have some form of background scanning you’ll find your scanning workstation will become the major bottleneck on your network. Another question you may want to ask is can you take the low resolution prescans that are done off your scanner and do colour corrections on anoth- er workstation apart from the one that’s doing the scanning? I think also that would be something nice to have, it’s not essential but it would make your life quite easy. If you think about doing ten prescans on your scanner very quickly and then having three operators working on the colour corrections rather than one and then you are send- ing them back into a master queue and having fine scans done. That’s fairly attractive and it means that you can get your scans out to where they should be, that’s into your pages, in a very short amount of time. Another important point, on productivity is that there a need for an off-line set- up process. And by that I mean something as simple as can you get additional trays and drums and drum mounters and all that sort of thing you need so that when you finish scanning one round of transparencies or reflectives, you can pull that off the bed or off the machine and get another one on there, set up ready to go straight away. And there are lots of different ways that all the vendors handle this. Some are good and some are not so good, but investigate this thoroughly, it’s well worth your time because again, what do you want to do? The main aim of having a scanner is to have the drum spinning and delivering scans on to your desktop now. It’s not the same as it was a few years ago, where the main point would be to have that drum spinning and have films coming off the end. What you want to do is to have your scans delivered on to your file server or onto your Macintosh with the colour corrections complete. You cannot go back into Photoshop when you’re working with these scanners and start fiddling with colour correction. Your productivity will go out the door. So look very carefully at the colour correction capabilities of the software that you’re looking at It’s paramount, and I’ve seen customers who have almost gone out of business because they didn’t look closely enough at the colour correction capabilities even though they had them and they had skilled operators on expensive workstations fool- ing around with Photoshop getting colour right. And that colour should have been correct off the scanner the same way it used to be when you used to get colour correct film off the back end of the scanner. There is no difference here, you’ve got to get it right the first time.

Colour management

I’m not going to try and give a lecture on colour management because it is a little bit of a complex topic as to how it exactly works. I’ve tried to keep it very simplistic and not be too technical about it because it can really eat you up a little bit. The first thing I want to ask is, can the people in the audience raise their hand if they know what co- lour management is, or are using it? All right, so there are not all that many. And that’s okay and not surprising, because I would only get a small percentage I would say, in any audience I’ve spoken to, because it is a new topic and one still treated with a great deal of skepticism I must say. Because not many people really believe that what one can see is really what you’re going to get. I tend to 

“… batch scanning is an absolute necessity. If your scanner doesn’t have that, your proposed scanner, don’t even think about it because you’ll find it will be slower than the operator working your traditional drum scanner.”

agree that that is the case, but we have some tools. to get you a lot closer than what you used to. So what is colour management? Colour management is a simple concept. In fact, it’s been done for many many years already. The scanner operator who does the scans that you ask him to do, has been doing it. Operators make a judgment, they look at the tyranny and they are told that it is going to be printed on stock x or y, on machine x or y and now make some decisions on how to treat that. And that’s colour management. So most people have it up here already. What we need to do now is to introduce colour management to desktop devices because there’s many more outputs now. and the outputs can be anything from a dye sub- limation printer, to a bubble jet printer, to a plot- ter, to a monitor, when you’re looking at a CD and Internet, to printed matter in many forms, offset, flexo, etc. You need to move to device independence so that when that scan is done and placed in the database, you don’t have to worry about where it’s going to be printed. In fact, you don’t even have to know that. All you have to worry about is doing a scan in Cielab with some very basic requirements. So colour management takes care of that. Basically it looks at the output and creates a fingerprint of that and does some changes to that file as it sends it to the output device. So it basically maps your input to your output, so you are going to get a result that is consistent. So that if you’ve fingerprinted your monitor, your colour printer, your imagesetter and your printing press, and you have done a good job of it, the differences that you would have had four or five years ago or you would have had an operator who had to make a lot of of how is the best way to deal with this, don’t have to be worry about because it is a simple task of making a decision on which colour management system do I use and there are a few around. Linotoype Hell works with the Ap- ple ColourSync system. But there are quite a few others, some of them are very good and basically what you do to have a colour management system work is, on the input side you scan a target that is then compared internally to a target supplied with the colour management software and it then cre- ates what is called a profile of the scanner. For out put you would print a job on your printing press after making films and plates using a standard method that doesn’t vary, that can be controlled, and you compare and measure that printed out- put using a spectrophotometer to a reference model that’s included in the software and you can actually create a profile for a printing press and all of a sudden you start to get a loop of calibra- tion.

So I’m talking about colour management because if your scanner doesn’t support it or can be used with it, beware. Because it something that is a great tool to use. It’s not going to send you broke not having it but it’ll make your life so much easier to be able to use it as time goes on. And it’s something that I would urge you, those of you who already have prepress, experiment a bit. Have a bit of a play with it. See where it takes you. It is a low-cost way of getting into high quality and making sure your quality control really does move up the scale a bit.

Colour management, what devices can it work on? Most new colour scanners, printers, monitors can actually be supplied with a profile which works with Apple’s Colour Sync program. That’s the one I generically talked about. Then again, colour management systems can work on just about most devices these days. And if you had old hardware, that shouldn’t be a problem because you can create profiles using some simple software from com- panies like Linotype-Hell for instance that allows you to measure the output of that equipment. Colour management is something I’ll be happy to talk to people about on an individual basis. Once you start talking about it at length it can tend to get a bit confusing for everyone, including me.

One of the things we did back in Australia. After we set up EAC Graphics and decided we weren’t doing it the way we should have been, was to set up project management of installations and I want to look at that while I’m over here with EAC Graphics. And that leads me to backup and support and will your chosen vendor support you with managed installation. Are they going to let you know what’s happening. Are they going to look after you and ensure that the equipment is com- missioned correctly. Talk to them about that. It’s important that you have an understanding of where you are and for you to let them know what your expectations are as well. Prepress is a tough game and you need to get the gear in quickly and get it operating very quickly. And to make sure that the staff that are training your staff are qualified to do so. Don’t be afraid to ask the questions. Don’t be afraid to have a talk with these guys. It really shouldn’t be a problem to do that.

This is a very simple slide but I think one that I’ll just highlight. Make sure that your choice can integrate to your existing equipment. Make sure that when you get this thing delivered to your place, it will plug in and it will work straight-away without, “Oh we have to buy an interface card”, or, “Oh we have to do a network bridge”, or “Oh we have to do that”. It has to be, has to be put to bed early, because again you want it in quickly, you want it in without any drama. So again, this is all part of the checklist you may want to look at Network connectivity. How does it connect to what I want to do. Does the scanner work direct- ly or does it go via the Mac. Ask the questions.

The last slide I’ve got is on backup and support. I won’t take too long with them but I think they’re important because it talks a little bit about the direction that I want to take EAC Graphics in in our prepress digital environment as well. When you purchase a scanner you have to remember you are purchasing a part of a whole system that’s got to be productive for you and when you start transferring scans in PostScript format over a complex system, that your network has to be up to scratch. That you start looking at things like “do I need a hub for my network, do I need fast Ethernet, which has been a good lead for us”. We did that back in Australia in the last six or seven months with all our servers and it was a very worthwhile exercise. Do you need to recable to make sure your cabling can support fast Ethernet, thin coax can’t do that. You’ve got to make sure that all those little things can be taken care of. You can have a trouble-free installation, you preplan it. Ask the vendor for a configuration diagram. It’s not going to take him very long to do it.

We found it was very helpful when we did that, because quite often we found errors before it even got to signing contract. By just doing a diagram and going over it with our potential customer. To make sure that we’re all happy with how things are connected together. That the network load would be okay, that we don’t have to put something else in to help cope with the additional traf- fic. It’s important because scanning is no longer a stand alone thing, it is something that feeds into your network. And all of a sudden you’ll have to be looking at things like file servers, perhaps im- age databases, where you’ll store a thumbnail image of all your scans so that they can be called up quickly. All these are things you may want to consider. You may not need half of them but think about what are your requirements and don’t think about now, think about six months ahead, because I’m sure many of you out there, have, three or four months after an installation, said “T’d really like to do this now” and you’ve found out the prices are quite expensive. You could have afforded it at the time but all of a sudden you’ve got to get it out of your cash flow and it’s not so easy. So you’ve got to try and get all these little items in the main body of your purchase

So you’ve got a nice modular system ready to run than have to purchase them three or four months down the line. At the same time, if you buy a modular system that allows you to add at minimal cost as time goes on, that’s also okay. But again, look into all the bits and pieces and find out what they’re worth, some of them will be ex- pensive, as long as you know. I think the worst thing is a surprise when you come to this sort of equipment, when you think that something is going to cost you less and it costs you a lot. So it’s about doing your homework as well. It’s I suppose all very easy for a vendor not to want to tell you about the item that costs a lot if you don’t ask. So ask him and maybe he’ll tell you.

I think that the other thing you’ve got to look at-the final thing-is quality as well. As long as you can produce the quality you want in the time frame that you allow for it, you’ll be fine. Things like servers and all that, they’re part of the whole equation. So the point is that you’re not just look ing at a scanner these days, you’re looking at a scanning solution. And the solution may involve a fast server. If it does, don’t halk at the fact that it will, because there will be nothing worse a few months down the line than for you to stop scan- ning every couple of hours to copy all your jobs to a SyQuest and put them somewhere because you’ve got no room left. It doesn’t work well that way. So there’s a lot of things to think about.

I can’t go on for much longer because I’ll be here all day and it’s all too easy to get caught up in this whole exciting thing of prepress and to talk about well what if we did this and how would we do that” but for now that’s my little contribution. I hope that I can meet with a lot of you during the course of the next few days and have a coffee or a tea with you or maybe even a beer afterwards and get to know a bit more about my new coun- try of residence and get to enjoy it a lot. Thank you very much.

Niels Ohrgaard on High end scanners

Thank you very much. My name is Nielsen Ohrgaard. I am from the Danish company called Scan- view. I have been asked to talk a little bit about different scanning technologies, that is one of the parts of the presentation, we can also talk a little bit about the difference between drum scanners and flatbed scanners. We are going to cover a lot of ground in terms of where we are in the world with regard to appreciation of desktop scanning, etc.

What I would like to say first and foremost is that this is by no means a lecture. I regard it as an opening statement for a discussion regarding the issues that I’m going to raise here. I think you should use all the experience gathered in this room today to have frank and open discussions about what equipment is right for you, viz. where are we headed in the prepress industry, etc. I must say that I enjoyed Peter Janusik’s presentation very much because there are a lot of things I agreed with even though we are coming from two differ- ent companies, both trying to do more or less the same in some areas at least. But we agreed to a lot of the things that he has raised here today.

Unfortunately, Mr. Janusik has two advantages over me, one is that he speaks English, I don’t, I’m from Denmark, I am a Danish guy, and the second is that he is representing a company which is very well known. Therefore, I will start my presentation by using a few minutes just to introduce my company, so we all know why I am actually here, what am I doing here? And if you thought I was going to give some answers to you during the presentation, what scanner’s the best? Drum scanner? Flatbed scanner? I can tell you, definitely not. I am going to raise a lot of issues. And actu- ally it’s for you to consider these issues. This is a basis for you to may be come up with your own answer as to what technology suits you the best.

Innovation in prepress

Scanview is a young Danish company. It was only founded six years ago in 1990. We heard before, the mention of a large mainframe computer, costing several million dollars, which is quite true. The vision of our company, the way we were found- ed was actually that the founder of Scanview had a vision like the mainframe computer which was downsized over the years from the main frame computer to a PC, that it should be possible to do the same with a large hi-end drum scanner and down size it into a desktop environment. So that is the basis of Scanview. Our mission statement is to become a trendsetter in the industry, we want to be in the leading edge of what is going on, whatever direction desktop publishing is going to. 

And we also strongly believe in the flexibility and freedom of choice to use standard platforms and interfaces for the equipment that you are using like we also heard before. To give you just an impression, Scanview has shipped more than 4000 drum scanners into the market, and we have a subsidiary in the United States, and we are a company in Denmark of approximately 140 people.

And what are we doing? We are covering the full range of prepress processes. What we believe we are able to do today and I will finish off by showing, is that we can provide a production solution from input to output. Scanview is today having a range of input devices and when I say input devices, image capturing, it’s because we are handling desktop drum scanners, flatbed scanners, and we are also manufacturing a digital camera. Now on the output side we have a range, a family of imagesetters and I am sure most of you are familiar with that term but where we have actually focused a lot of our attention and where our future development is, is in this groove in the middle which is in software. We believe a lot of enhancement, a lot of additional value to already existing products on the market can actually be done in software. So that is where we are focusing our attention.

Let’s try to take a look at the scanner revolution because if you look at it from outside it may be seen that a revolution has been going on, within scanning or prepress over the last few years. In fact, for us who are working within the trade, I would regard it more as an evolution. It is inevitable that the changes will happen and there have been lots of changes and more changes are yet to come. If you look in the 1980s, colour was the exception, print runs were few and large, you had very very large print runs and quality was very consistent and relatively high due to the high numbers of high end scanners on the market. In the early 1990s or actually in the late 1980s where we saw a lot of flatbed scanners coming on to the market, every- body wanted to do colour, everybody wanted to do it themselves. What happened was due to the poor quality of the first initial flathed scanners coming on to the market, due to the non-profession- als trying to do colour, the overall quality of co- lour on the market for publications, for in-house publications etc., for a period of time was actual- ly very much degraded. It was only in the 1990s that colour became a de facto standard. We want colour, everybody wants colour, you can’t sell al publication unless you have colour and the way 


❝If you look at it from outside it may be seen that a revolution has been going on, within scanning or prepress over the last few years. In fact, for us who are working within the trade, would regard it more as an evolution. It is inevitable that the changes will happen and there have been lots of changes and more changes are yet to come 35


to differentiate one company from another, one publication from another today is better and better and more and more colour. So what we have seen is that desktop colour, which initially started on a trial basis (flatbed scanners of course are most- ly used for placements), has now become a very known and accepted quality level on the market.


Scanners on the market today


I think, I am going to introduce today a term called heritage, because I think most of us in this room today have some sort of experience with high end drum scanners and there is a heritage about these scanners that I’ll come back to. There is a very solid perception of what a high end drum scanner does for us today. However, there are also other scan- ners on the market today that we should discuss and those are actually what we call the professional flatbeds. You might know some of these scanners already just from the presentation of them. We also talking about the professional desktop drum and we are going to talk about where they are in comparison to each other. We have some slide co- lour scanners and we have a colour scanner as well. What they do is that almost all of them represent the same sort of specifications. There are differences of course, some of these differences are the reason why you have to make a choice. You have to find out exactly what requirement you have in this big scope of different scanners here. That is where, as you heard before, DMax comes in and you have to look into resolution.


Of course, this is a consideration that is not just for today. I think it’s very important that an actual purchase of a scanner today relies heavily on what is actually achievable with that same product tomorrow.


Let’s have a bit of a technical approach to the two different scanning technologies. We’ll try to do it not too in-depth and if you have any questions later on we’ll just take them in the panel or over a cup of whatever. But in principle there are only two scanning technologies available on the market, one is PMT and the other is the CCD 


“That is where, as you heard before, DMax comes in and you have to look into resolution. Of course, this is a consideration that is not just for today. I think it’s very important that an actual purchase of a scanner today relies heavily on what is actually achievable with that same product tomorrow.”


technology. There is no other. Now if you’re looking at a high end drum scanner, the heritage is that the scanner is incredibly sharp which gives a very good result, and another heritage is that the traditional high end scanners have very high resolution, good DMax etc. So let’s see why a professional desktop scanner has the same heritage, and has inherited the same qualities as that of a high end drum scanner.


One of the benefits of a drum scanner is of course, that they have single point optics, it means that they look at all the pixels with exactly the same optics here in the same angle, always at a 90 degree angle on the transparency that is scanned. A photomultiplier is a device that takes light and transforms it or a photon and transforms it into an electron and every time the electron hits one of the plates inside the photomultiplier, it will actually triple, let’s say one electron here becomes 3 here, becomes 9 here, becomes 27 here. That amount of data is taken into analog to digital (A to D) converters and into the scanner and han- dled. In principle, here we are talking about pix- els always treated exactly the same way. Another heritage is the density. We know that a high end scanner gives good density, good optical density, but I think there’s more to it than that. We can also look at the bits, we know that most scanners on the market today when they are connected to a Macintosh, they do scan in 12 bits. And if we make a simple calculation, if we have a 100% light, we have a piece of film, we have some die grains of certain thickness, and let’s say that this certain thickness here will allow 10% of the original light to come through. This is a level of gray we are talking about here. In principle, if we think that the die grains will double in thickness, we would believe only 5% would be allowed to come through, but that is not the case. When you double the thickness you will have only 1% left to come through the film. That is why density is represented by a member and the formula for density is logarithm to 1 divided by the transmitted percentage of light. 


Now that’s interesting, because in theory and this is a theory, if you say black is zero and that  white is 4095, it means that in just one step from the darkest black to the next level of gray only 0.024% of light is transmitted giving a maximum possible density of 3.6. This is in theory. What 1 am trying to say with this is that 0.025% of the transmitted light is very very little and it will very quickly be eaten up by noise distortion from the electron- ics introduced in the scanners So we can’t just say that the density or the capability the scanner of a density of 3.6 in principle is the only measure of how good you can actually reproduce a scan. We need to take the oth- er factors into consideration as well. Again, junto dwells a little bit on the high end scanners, they are always also perceived a as giving very high resolution. Why is resolution important? Peter mentioned that the resolution is not that important except when we are talking about high enlargement which is very true. But when you are talking about high enlargement, the way that an imagesetter or rather a RIP works, to come with the output from the scan input sample is that we have a recommendation of two scanner samples per screen dot, for the transformation in the RIP. Now whenever we are talking about interpolation which is a thing you have heard very much about because many of the low end flatbed scanners they do not have very high true input resolution, so, you are taking a 750 DPI true optical resolution and interpolating it up to 4000 DPI or whatever, and what is it that interpolation actually does? Does it gen- crate more data or information? Yes and no. In- terpolation generates a lot more data but it can never generate more information. So by nature it would blur the contours like you can see on the sample here.


What is meant by this slide, and I’m not sure I fully understand, but what I would like to say with this is that when you are talking about a high end scanner, you have a colour computer on board. The colour computer is in the scanner itself. When you are talking about the newest developments of high end scanners or high end desktop drum scanner or flatbed scanners, these controls are all found as a colour computer based on a non-proprietary platform, a Mac or PC. The important thing is that the scanning, again if we are talking about 12 bits, the human eye can only perceive two-thirds of the information that the scanner ac tually obtains. Therefore, it’s important that the information that we work with from the scanner itself is done with as many or as much informa tion as at all possible. We cannot work only with the 8 bits in the computer. Already that will de grade the final results. On all high end systems you will work with the 12 bits that you will have in most scanners, 12, 14 or 16 bits in the scanner itself and only when all the corrections have been made you will then take it into 8 bits for the com puter.


And if we compare this to the other segment of the market, which is the flatbed, we’ll just fin ish off by having a look at what drum scanners are giving us. The professional drum scanner or the high end desktop drum scanner, the heritage is that it is an accepted and proven quality. It gives a high dynamic range, good shadow detail, and a high level of colour definition-very consistent, smooth and sharp from edge to edge because of what we discussed before, the PMT technology. The high end of the market with the high end scanners, they give you very high true resolution and on the other hand the heritage is that they are complex devices and they are expensive. We are of course aware that the high end scanners are coming down in price. But we should all be a aware that so is the quality.


The high end scanners of some years ago at the highest prices are not the same as the much- cheaper scanners of today, but those of today have so many other advantages over previous models. Again for the high end desktop drum scanners, complexities, size, and price, have come down and on a drum scanner it is still possible to mount in oil. And I think it would be wrong to say that you would never need to mount in oil-that would be the same as saying that you would only have brand new originals without any scratches. There will always be a need to do high enlargement in oil mounting. Of course, one of the drawbacks is that cardboard mounting is not at all possible on a drum scanner.


Therefore, there might be some room for very high end desktop flatbed scanners. In fact, a lot more technology development has been going on in the CCD area than in the PMT area as we have also heard before, Most of the high end desktop scanners today are using exactly the same tech- nology as implemented in the high end systems. Three PMTs, three power supplies for the PMTs, etc. So the innovation in the desktop drum scan- ners has only been downsizing. The technology as such is the same as has been used for several years. On the other hand, in the CCD market, there has been a lot of development going on. The reason is that a CCD element is actually used in


 “One of the benefits of a drum scanner is of course, that they have single point optics, it means that they look at all the pools with exactly the same optics here in the same angle, always at a 90 degree angle on the transparency that is scanned.”


the space industry where there is a lot of enhancement and there is a lot of requirement for better CCD elements.


In the early stages or even today, you could look at several different markets. You can question why is high end flatbed scanner so much more expensive than low end flatbed scanner? What is the difference? There is a lot of difference. Actually we always say that you get what you pay for. There is lot of difference in the way that the CCD scanner is working and what electronics and optics have to be built into these scanners to achieve a high end result. If you have a 6000 pixel element, if I have my element here, the flatbed here, I have a light source here, if I have 6000 pixels elements and I have an A4 scanner, the maximum optical resolution you can get on a Al size flatbed scan- ner is 750 DPI through optical resolution when we are talking about a fixed lens scanner. A fixed lens scanner is cheaper than some of the other models we are going to talk about today.


Other parameters are important. If you want to have higher optical resolution on your flatbed scanner, not over the entire bed, but on parts of the beds, then you have to use a multiple lens sys- tem. It means that you would have a lens placed in various places in the CCD scanner, allowing you to scan a smaller area at the full maximum reso lution of the CCD element itself. That is why on some of the very high end CCD scanners today you can actually achieve 4000 or 8000 true opti- cal resolution down the middle band of the scan- ner and of course as you move the lenses further out it will degrade and go down to 1800, 1200, or 1000 DPI depending on the scanner design. That is a multiple optic scanner, there is one more scanner type on the market which is quite advantageous in many respects. The zoom lens flatbed scanner would give you a very very high degree of flexibility and high degree of levels of chosen resolution for various areas on the plate. Further- more, it gives you some benefits in terms of scan- ning 3D objects.


More than the CCD element itself, as we will discuss little bit later, why CCD elements are so important, is the fact that the development for A to D (analog to digital) converters has also been  picking up very sharply lately. Again from the space industry but also for the automobile industry, 14 bit A to D converters have been brought down in price, so it is now possible to implement them in CCD scanners. It was not possible before because the price of the scanner would simply be enor mous. And finally, also for a CCD scanner what is very important is the light source. Some CCD scanners are actually using a light bulb. They are rotating RGB filters in front of it. Others are us ing a standard light tube and some others like Scanview have invented a very continuous spec- tral light tube that we are using in our CCD scan- ner


We are looking at some of the problems that we have with flatbed scanners. We are looking at some of the reasons why until now, at least till a year ago, there were problems in manufacturing a CCD scanner that can actually compete fully in quality with the high end of the drum scanners. One of the biggest disadvantages of a CCD has been that a CCD element actually is a piece of material from which you could manufacture a lot of different electronic components. It’s a piece of silicon that normally has been very very expen- sive because it could be utilized for other prod ucts like transistors, and chips, etc. But now a way has been found to increase the reliability of the CCD elements and actually also the performance of the CCD elements. We are talking about some thing called pixel lag and line lag in CCD scan- ners and I think to put it very simply, a CCD ele ment is in its nature a charge coupled device. It measures light and it transforms it into electrons. If you visualize that you have a CCD element with 8000 pixels may be you can visualize 8000 firemen in a line, each with their buckets. They have to put out a fire so the guy who stands all the way to the left, has to empty his bucket onto the fire be fore the next one gives him the next bucket, that is how a CCD element is working


You cannot empty the information in one cell, into a bucket until the cell next to you has actual ly been emptied. So this is a process that has to be done and until a few years back you were only able to get a CCD element that would be able to avoid pixel lag (able to empty the bucket), let’s say up to 85% or the like. Today we are talking about 92 to 95% ability to empty the bucket of water or empty the information. The practical meaning of this is that if you looking at the far- thest right hand corner, this information, let’s say it’s a pure white, when it actually comes all the 


“More than the CCD element itself, as we will discuss little bit later, why CCD elements are so important, is the fact that the development for A to D (analog to digital) converters has also been picking up very sharply lately. Again from the space industry but also for the automobile industry, 14 bit A to D converters have been brought down in price, so it is now possible to implement them in CCD scanners.”


way down to the end, and empties the last buck- et, its down to 95% white. So therefore, we are talking about the fact that CCDs some years ago were only able to giving you a plus or minus 15% accuracy. All these problems mentioned here, are problems that have been overcome to a great ex- tent. They are not fully there yet, but the quality. of CCD elements has improved so much we can give a very high quality on high end CCD scan- ners.


Spatial resolution


We spoke about enlargement, interpolation, tonal resolution and noise. How do we reduce the noise and keep a good optical density and sharpness, we have format issues, speed and performance is sues, we have so many issues that I have not at all touched upon yet and there is the case of use which also has to be a subject that is a part of the con- sideration of what type of scanner should be cho- sen. We will actually come back to some of these issues a little later today.


What we are talking about in terms of world- wide presence is that the first step for desktop publishing was actually to get acceptance from the market, that the quality you could get from desk- top publishing was that of the high end scanning. or the quality of high end scanning. This is not necessarily true though. I think that the impor- tant issue to discuss is actually what quality is re- quired and what is necessary for the type of jobs we are doing and what is the profitability because it all comes down to the last point. It has to do with the initial investment you have to make in a equipment for the future.


What is the price you can charge for the job you are doing? And is your customer happy with that quality you can provide for the money. What has been going on is that the colour flatbed over the last few years has gone from the segment of, for positioning only or home/office scanning, and they have started to move into the magazine, bro- chure and catalogue work. We are talking here 


“You cannot empty the information in one cell, into a bucket until the cell next to you has actually been emptied. So this is a process that has to be done and until a few years back you were only able to get a CCD element that would be able to avoid pixel lag (able to empty the bucket), let’s say up to 85% or the like. Today we are talking about 92 to 95% ability to empty the bucket of water or empty the information.”


about the scanners I have mentioned before with the fixed or multiple lens system. Professional flat- beds are scanners with zoom lens technology. These scanners give very high end results. We have seen the growing, very high-up in the market, fine art, advertising, company profiles, etc. We have also seen over the years, all over the world, in Eu- rope, United States, Asia, that the quality achieved is very very high and the price point has come down.


So, how can you achieve the best of both worlds? I think there are many ways to go about it. Again the important thing is that nobody can help you with this. You have to make the decision, because you have to base your choice or further requirements you have for your jobs. We have still the-low-end-of-market, slide scanners. We have colour flatbed scanners, we have professional drums, professional desktop drums, and of course the professional flatbed scanners as also shown here.


Now there is one other solution as well and that is the combination of the two where you ac- tually, on one workstation, combine the high end desktop scanners with the flatbed scanners. This allows for those jobs that you cannot do in drum scanners, to be done on the flatbed scanner with- out the need to invest in extra computer power. Besides, there are a couple of devices in the market with the zoom lens technology in flatbed scan- ning and of course they are positioned in this range of scanners very near to that of the high end desktop drum scanner or to the traditional high end scanner. The reason why I have chosen to show a different design than the other flatbed scan- ner you saw before in the circle, is that besides the point of drum versus flatbed, there are a lot of other considerations to take into account work- flow and like we have also heard before, that of batch scanning and productivity. I think I will bring you something to consider. Wouldn’t it be nice, if it was possible to do something, like I am try- ing to describe now, six steps to performance, now  what is it that we are all looking for? We are look ing for a scanner that is fast and we are looking for scanner that will give you very high produc tivity and thirdly we are looking for a scanner you almost don’t have to touch. The scanner should be able to do the job itself. Just to give you an ex- ample, if you have a scan, which takes 2 minutes and 10 seconds, if you can do 27 typical cover pages scans per hour. If that is good enough speed for you, then the first requirement is met. Secondly, batch scanning is defined by 20 scans and in prin ciple that would give you a 44 minutes coffee break.


However, there is a removable plate systems on the high end scanners. It means that the cof fee break is over and you have to do a mounting of the other plate while the scanner is scanning. Now we have some thing little bit different, be cause, when you are actually in the middle of pre- paring the second plate, that preparation might be finished long before the batch scanning of 44 minutes is ready and the operator is now ready for his coffee break. But it is today possible in sys- tems to go in and pause a job, meaning that when we are doing batch scanning of 44 minutes, if you can pause the scanner, take off the plate that is in there being scanned, put in another plate, do a two-minute preview and then after that contin- ue your batch scanning. So it means that in prin- ciple you can set up and crop everything for the third plate while the scanner is just finishing the first batch scan.


An additional feature would also be to have a scanner so that whenever you lift the job, it would also automatically pause, meaning that you could replace the plate in there and preview another plate [system clearing in there with another plate and have a preview done for you.] Then you close the cover. There is a bar code on your plate and it is scanned and if there is nothing to scan it would do the preview. Another additional feature is some- thing we call mouse-free scanning and in princi- ple what it does is that you don’t even have to be in, you don’t have to touch the Macintosh at all. You are controlling everything from the cover to closing the cover and doing the scanning com- pletely mouse-free. Lastly, I think we talked a lit- tle bit about colour concepts in a previous pre- sentation, and it is going to be more of the same, but from our perspective we are very much in agreement with what has been said today about the workflow. The process of where you work has to be very and in principle you should be

“We are looking for a scanner that is fast and we are looking for scanner that will give you very high productivity and thirdly we are looking for a scanner you almost don’t have to touch. The scanner should be able to do 

the job itself.”


able to do the following: cross platform clients where your scanner is connected to a Mac and PC, in a network should not matter. It should not matter whether it’s a Mac or PC your software is running on. It should be possible to connect your scanner to a Windows NT cluster server. It should be possible from any of the work stations in the network to access and drive the scanner across the network. It should be possible to distribute all the previews made from the scanner to all work sta- tions in the network. It should be possible for the scanner to recognize any stored data with a pre- view and get it from the NT cluster server here. Furthermore, what is also possible today is that in one box we can have file server, scanner server, OPI server, RIP server in one and the same box driving the imagesetter, driving the scanner and of course, the real benefit you get from this is much less load on the network, and much higher speeds because you are not transferring the high mega- byte files over the network. OK. What is the con- clusion? As I said before, this is a topic for discus- sion.


The conclusion is you should get what is right for you. You have to come up with the require- ments. You have to set the specifications for what you require for your business to be profitable. Colour scanning and separation are easier than ever. You can go from colour flatbeds to any size drum quality today. You have to decide whether a flatbed scanner today can deliver the kind of quality that you are doing in your shop, or you have to decide that you cannot use a drum scan- ner because you are using large originals and you have to do lot of line work, a lot of reflection art. These are the para-meters you have to consider. Anyway, open scalable solutions that enable you to grow, I think are the recommendations that I would come up with here today. I am looking for- ward very much to have a discussion with any of you who might be interested in hearing a little more about the experience we have acquired from our customers around the world. Thank you very much.




BURJOR POONAWALA I am sure the audience will have lot of questions to ask and I might start off with one question and that is, we haven’t talked much about line scanning. Because to the best of my knowledge we had excellent drum scanners, and we could produce excellent colour separations but when it come to line scanning, we had a prob- lem. So can you enlighten us on the line scanning as far as flatbed and drum scanners are concerned?


PETER JANUSIK I have to express a little bit of a lack of knowledge on this, mainly because of my background and what we were doing back in Aus- tralia, it is not a big issue to have to do scanning of type and that sort of thing, but yes, it can be done and the results are better than what they used to be. Especially from the Linotype-Hell perspec- tive they have put a lot more effort in the last cou- ple of versions of software to increase the speed of line scanning and getting better quality. Also, with the new Copix option for the Topaz II scan- ner it is pretty much designed to do that and it produces very high quality type as well. But I wouldn’t say that the vendors are putting in a huge amount of R&D into this because we would cer- tainly consider this to be a market that is decreas- ing rather than increasing.


NIELS OHRGAARD From our point of view, since we welcome both the high end drum scanner and the high end flatbed scanners, we definitely see an advantage of doing it on the flatbed scanner rather than drums, the main reason being speed. I can also say that today the advantages of a flat- bed scanner over the drum is that the drum has a tendency to be very focused, be very sharp and giving some jagged edges that are easier to com- pensate for, when we are talking about the flat- bed scanner. I think that what we are doing to- day is trying to focus more on that application, not because we are in disagreement with what Peter says but merely because we at Scanview, are lagging behind some of the other vendors in terms of offering a very high speed solution for line art scanning and we see the demand being there for the next two years to come as well.


BURJOR POONAWALA With line art scanning, I mean this is little more peculiar to our country because here we still haven’t come to the stage when we accept the majority of our artworks to come in the digital form. They usually come in hard copy form and they come in different sort of qualities. In fact, I would say that most of our typesetting and line artwork is done on computer


“…with the new Copix option for the Topaz II scanner it is pretty much designed to do that and it produces very high quality type as well. But I wouldn’t say that the vendors are putting in a huge amount of R&D into this because we would certainly consider this to be a market that is decreasing rather than increasing “


 but unfortunately instead of passing on this information to a prepress house digitally, it is usually sort of translated through an imagesetter on to bromide paper and then passed on to us, with the result that it usually comes with corrections which have small bits of type pasted on and sometimes they are not very clean or sharp, the papers have a slight tinge of yellow, with the result that we in India have to contend with a lot of these problems, and so, line work scan- ning is a little more important to us, perhaps than in your countries. So the point is that once again we have been a little perplexed because there are certain file formats which give you a lot of information as far as line scanning is concerned whereas certain file formats compress this information into less data and therefore, we have to use more disk space. Can you enlight en us on this factor? Because bitmaps I believe give you a small line file and the other formats give you a lot of data with the result that you run short of disk space.


PETER JANUSIK I would tend to think that the problem lies more with the ability of a lot of the desktop programs to handle a type image as such. They are not really designed to do that. You have to treat it as a picture then you colourise it and all that sort of thing but it starts to get pretty cumbersome. When you look at the programs in the market such as QuarkXpress, they are really designed to create type and to type things in not as a page layout workstation and that sort of work really tends to be best handled by the traditional page make up systems available I believe, and it’s a little bit tough, I think PostScript doesn’t han- dle it all that well. We went through a very simi- lar thing in which some customers we were deal- ing with wanted to do exactly that, they wanted to scan line-art and wanted to colourise scans on a Mac and make the Mac work like a traditional system. It doesn’t do it. So it’s a little bit of a tough question, I think it’s a matter of looking at indi- vidual requirements as they come up. If you want  to be very innovative with what you do with that type you want to scan in, you want to be very careful because I don’t think that the desktop does that real well. I think the desktop is good for taking what’s been created electronically and goes through the whole system that way.


RAJESH GULSHAN My question is to Peter. I am drawing your attention to the separated film to the TIFF image, where you mentioned that Magazines can take up advertisements and impose them on the system. Now what is the quality, which basically will come out using the software which you mentioned?


PETER JANUSIK With regard to quality there are two options you can look at using the system I mentioned which is the Topaz Copix. You can scan the originals in two forms, one as what is called a Copydot and that basically takes all information that’s on those films and reproduces them as they are and the quality of that is very high but it’s not right for all instances where you may have colour type and that sort of thing and in which case you may have to go to a descreen with type quality that is good but for images it’s certainly excellent. So you have to be a little bit careful with what appli- cation you want to use for this and do some test- ing. We found back at home that just saying, “Hey we could take this separated film and do all these wonderful things with it was a very broad state- ment and you can’t do that all the time. So it is a matter of looking at requirements and saying well if you’ve got x amount of film and all you really want to do is just create a copy of it to merge to file and nothing with.


RAJESH GULSHAN So basically the descreening is built into this software?


PETER JANUSIK The descreen gives you a good result for your images.


RAJESH GULSHAN I am asking you this because, in our country and may be all over the world the advertisements require very high quality. A 2% loss of quality is not accepted by the advertising agencies, so if the separated films have come to us and we have to use them as part of the magazine and we do it with 2% loss on the quality, the people will not accept it at the advertising agency. Say we have to have 100% quality back. Will this software give it back to us that way?


PETER JANUSIK Some of the results I’ve seen have been very much a facsimile of the original. Abo it depends on the subject matter, there will be some that wouldn’t be good enough to do. But there will be lot of work which is fine, I think it will depend upon what the content is. Colour type can be a problem as I mentioned but on black type images, Copylot will give you very faithful facsimile of the original. I think it is quite useful.


PARTICIPANT This question is addressed to Niels Now you are talking about CCDs. You have said that the CCDs, just collect the information and pass it on from one person to another person. things like that, because they are passive CCDs. That is, after collecting the information, that is the signal they collect from the scanning, they will pass on to a processing system. They do not process themselves, isn’t it? They are called passive CCDs, and they are able to just collect the information. The further processing is done by a separate electronic signal processing system. So during the passing on of the information from one CCD to the next CCD there is a loss of information. Now I have read an article wherein the Jet Propulsion Laboratories, who are the number one facility in manufacturing CCDs-in fact all CCDs we are using come from JPL-have come out with a solution that is called an active CCD, that is to say, a after collecting the information, the information will not be passed on to next system. They will pro- cess the information themselves. That is, they have got processing signals like for example the active CCDs is based on CMOS technology (Complementary Metal Oxide Semiconductor) that is the same technology used by your CPU, your memory, which are all active elements. They can process the signal after receiving it. So if active CCDs come according to the JPL a scanner which costs say US$2500 is going to cost $250 in a span of three or four years. That is what JPL. says that they re- lease the active CCDs, to manufacture digital cam- eras and flatbed scanners. As far as I can see there are only two devices in prepress which use CCDs. One is flatbed scanners, which use a linear array, the other is a digital camera, which uses a matrix CCD in the form of array. If they release the ac- tive CCDs in the market, it seems that the cost will come down from 100% to 10%, that is, a scanner which cost about $2500 will come down to $250 in a span of three years. Is this true, and if it is true, what will be fate of people who buy a scanner to- day, after three years?


❝ advertising agencies, so if the separated films have come to us and we have to use them as part of the magazine and we do it with 2% loss on the quality, the people will not accept it at the advertising agency. Say we have to have 100% quality back. Will this software give it back to us that way?”


NIELS OHRGAARD I think this is about technolo- gy and innovation that is going on. I think it is a very valid question. But I also think that it is very often used as an excuse for not buying today, and delaying your purchase till tomorrow. And the thing is that no manner how many days you wait, there is no change in the rapidity of development going on. Just look at the computer, look at the PC that you bought a year ago. What is it worth today? Where are we now? The one thing that is being focused more and more is upgradability of various products. That means if you buy a PC, you can upgrade to higher speed etc. But I think with the technology, I do not agree with CCD technology and with the active chips, because what we have seen so far, they are very far from giving us some products we can actually use at all. So it will still take some years before a new technology like this can be implemented. But it is one of the areas where people are doing most of the developmen- tal work, I believe, in scanner technology.


PARTICIPANT AGAIN It is not a question of releasing. It has already been released, used in space crafts for capturing the images using the CMOS- CCDs. So any research when it comes out with new technology, it will be incorporated in space technology especially as far imaging is concerned. Now they are saying that we will be shortly releasing these chips to industrial companies.


NIELS OHRGAARD A few years ago you could buy a CCD chip of around 50,000 Danish Kroner, that is $5000 a piece. Now it has come down in price so that one can actually use it in your scan- ner. But it has taken a long time, a lot of development from the initial phase where people believed they could get the quality of CCD elements we are using. And I think that Peter may have experienced something similar.


PETER JANUSIK Yes, I think an important thing to remember when you are looking at especially high end flatbed scanners is that the CCD element is a very small part of it, really, you’ve got a very  accurate transport system, where you’ve got your original, you’ve got electronics that have to have very little noise, you’ve got a lot of R&d) that drives it, and so the cost of the CCD in relation to the whole thing may not mean very much at the end of the day. If you want to call a little desktop a flat- bed scanner where the CCD is probably the big- gest cost, it may impact it more, but there’s a lot more R&D that goes into one of our scanners than just the CCD array. And I think that would be true of all the manufacturers. I mean we buy ours from Kodak, so they’ve already done that and we’ve got a lot of R&D to do even after we get the CCD ar- ray.


FRED POONAWALA Should there be any differ- ence when you see a result done on a high end desktop drum scanner as compared to a high end traditional drum scanner. I mean in terms of, I believe, that the desktop scanners can only accept eight bits and therefore you could expect a little less quality out of them?


NIELS OHRGAARD Maybe I could answer first because I think this might be an issue where do not have exactly the same opinion necessarily. For the desktop drum scanners, as I said before, especially the high end drum scanners, some of them are using the same technology and even further, some of the same parts as in traditional high end drum scanners. I have known high end drum scanners for many many years, and of course there is al- ways a preference for something that you know and have been working with for a long time. How- ever, what we are seeing with the quality from desktop drum scanners today, I think is very very good and can be compared. The thing I started to mention regarding 8 bits versus 12 bits or 14 bits is that I think you can achieve results as good as the traditional high end scanners. When you are scanning in the scanning device itself wheth er it is coupled to a Macintosh or PC you are ac- tually scanning in the data, in one instance, in 14 bit. So there’s 14 bits of information per colour in the scanner. When you are working with something called an interactive feedback window which I believe we both have, you will be able to manipulate the data before your final scan in the 14 bit mode. So when you are talking about what the human eye can actually see, the two-thirds of the information from 12 bits, if you can get the most important 8 bits, what is actually left in the com- puter, you have thrown away the bits that are not 


“What do we mean when we say quality? Quality is such a relative term. When we speak of quality does it mean sharpness, does it mean likeness of the reproduction to the original? What is it when we ask for quality? Because in my experience a customer no longer looks at the original “


necessary and the remaining eight bits is a very important part of the image that you’re scanning in. Therefore you can get a very high quality


PETER JANUSIK I suspect Niels is from an engi neering background, I’m not. I can only tell a story of some customers who went from our very tradi- tional high end scanner $3900, and then started looking at a our desktop controlled scanner of the same heritage. We don’t have desktop drum scan- ners, but I have to say that once they did all their homework, this is one of our very big customers called ShowAds who at one stage probably were one of the top five Hell customers in the world. they are very large, and they had bought some $3400 scanners for their Thailand operation af ter about six weeks of tests and they considered the results that they could get off a scanner con- trolled through a Mac to be as good as they were getting out of their other machines. From my point of view I would say you shouldn’t really see any deterioration in quality. But a lot of it has really got to do with how your traditional operators make the transition to a Mac-based interface as well. That’s important.


PUNDERICK The Copydot scan will have restric- tion in dot shape, size of reproduction in differ- ent sizes, and for stochastic or FM screening, will it make any difference in Copydot scanning while outputting?


PETER JANUSIK First of all, you cannot use Copy dot with stochastic screening. Secondly, the big- gest issue that comes into doing a Copydot type of technology that I have seen (It’s fairly new and we were installing only our second machine in Australia when I left) is that you have to be care- ful in that you have some form of consistency in the sort of film that you’re try to process and you’re not trying to scan these weird things where you have four different screen rulings in a sheet be- cause it doesn’t print very well. So you have to look for some level of consistency, as I said with the Copydot you’ve got to be a little bit careful so that you automatically don’t just get any bit of film that’s on the market and try to combine it onto your page You have to look for something that’s fair- ly close. If you’re trying to do a 4 up 150 line screen and you get something you try and Copydot from a 65 line screen, believe me it doesn’t work and the software cannot compensate for that. And you’d have to look at descreening it and then re- screening it again on output. Okay, so there are some issues there that you have to look at care- fully before you go down that path. But FM screen- ing cannot be Copydotted, so to speak.


SUNIL KHULLAR I will take Freddy’s question slightly further and wider. And my question is not just to the panel but to the audience as well and I am trying to build a consensus on the quality is sue. As far as typography is concerned, there is a general consensus that over the last 15 years, that is, since new open systems have come in, there has been a general deterioration of quality in typog raphy. There is a general consensus. Is there a general consensus, similarly, for colour? We are not talking about 8 bit or 12 bit. Are we today, not just in scanners but the entire workflow, bet- ter off then say seven or eight years ago? Are we still the same or have we improved? What are your opinion as vendors and what are the users’ opin- ions about their experience. Someone mentioned that in between quality did deteriorate. But where do we stand at high end scanner on colour quali ty level?


NIELS OHRGAARD If you’re looking at the two markets today, where the perception is that the quality of work or colour is the highest, it’s Ger- many and Japan. There you will see that there is a great mix and a great acceptance of desktop pub lishing and they are manufacturing high end mag azines and publications on desktop systems. The high end Lino of course on the Macintosh etc. I think in general, there are two aspects to it because seven, eight years ago I would have imagined, that you would have a higher level of quality because of the scanners or the recorders available. But if you look at the same time at the printing technol ogy or the paper used at the time, I don’t think that in general he quality was higher. I also think that traditional repro houses were under lot more strain in those days to output a lot of work, not only in colour but in black and white as well. So the experience from the operators of the high end system may not have been over all as great as for some specific trade shops. I believe that on a high end system you can do the best quality in the world. in terms of colour, if you have the right operator. It’s easier today on a desktop system with a un- skilled operator to achieve as high a level in terms of quality. I don’t think that there has been a de- grading in terms of overall quality.


SUNIL KHULLAR Has there been an improvement do you think? Apart from the ease of use and the ease of work flow. Do you think that we are able today with the new open systems to produce bet- ter quality than in the traditional scanners, I mean not the analog scanner but the traditional digital scanners of the last five or ten years.


NIELS OHRGAARD I don’t think it’s possible to make necessarily better colour quality. I think that what has happened is that there’s a broader hand- width of products of a higher colour quality. If you’re talking about a specific, whether a desktop drum scanner or whatever can produce the same quality in terms of colour, as a big high end with a professional operator, no, I don’t think so. But in broader terms where you see many more di- versified publications today have a lot higher qual ity standard than they used to have just a few years ago because of the widespread use of desktop pub lishing.


SUNIL KHULLAR Is that the general consensus of everyone present?


BURJOR POONAWALA I might wind up by putting another question to the audience here. What do we mean when we say quality? Quality is such a relative term. When we speak of quality does it mean sharpness, does it mean likeness of the re- production to the original? What is it when we ask for quality? Because in my experience a custom- er no longer looks at the original. Ten years ago we were asked to see and observe the original, see we’ve got such a beautiful sky and it’s not beauti- ful in this picture. Today they don’t look at it that way. They just say well I like this image or I don’t like this image. I don’t know why I don’t like this image. Please make it so that I like it. So these are the kind of dilemmas we are in and therefore when we talk about quality, according to me the image that I am presenting on paper might be excellent, but the customer doesn’t think so, he thinks it’s not good. So what quality are we talking about? Let us think about this.

Indian Printer & Publication                                         March  1997

Proceedings of the third annual pre-press technology conference Part 3


Held at Bangalore on November 28, 29, and 30th 1996

Introduction of Mr. M Sankaran by Naresh Khanna


Mr M Sankaran will make the next presentation for Paul Foster of HyWay Ferranti who couldn’t be here because of a family problem. Mr Sanka- ran is a friend of ours and like many of the pre- press professionals he is an electronic engineer. He worked for many years with one of the leading companies in the world, Monotype and Mono- type India Ltd. In the last two years he has branched off and started his own company called Graphsoft Integrators. Mr Sankaran will make Paul Foster’s presentation on imagesetter solutions.


M Shankaran and Paul Foster’s presentation


Good afternoon, gentlemen. I am being given this responsibility of presenting the paper of Mr. Paul Foster, on his behalf. I just want all of you to know that Paul could not come mainly because of a re- cent bereavement in his family and he was look- ing forward to meeting all his old friends, custom- ers both old and new that he has managed to con- vince the years to buy equipment, potential customers, and even some competitors that he has  been in contact with over many years. So he hon- estly feels sorry to be missing all of you and he has requested me to present this paper on his behalf.


This presentation is mostly not going to be a sales presentation about HyWay Ferranti or its products. We promise you that I will only be talk- ing in general overall terms and not promoting individual products and my talk will also make no attempt to give you an update on the latest hardware platforms available. I am sure this will come out in many ways through many of the other presentations that will be made by other manufacturers, suppliers and users. What I wish to achieve, is an understanding of the benefits that will become widely available to you should you choose to follow this new software technolo- gy route. Benefits in production speed as well as the one benefit that we are very much interested in looking at is cost-effectiveness. In short, what are the main areas that you should be looking at. Perhaps I should begin by giving my defini- tions of the terms contained in the core of this talk. The definition of system. System simply is an organised body of connected items, both hard- ware and software, and people working together to perform a particular task. The second is the term solution. Solution can be described as a method of solving the problem. Cost-effective means a reasonable cost for a system which will perform the task for which it was purchased. I am sure that we all recognise that computer tech- nology has progressed at a rapid pace, especially in applications within the prepress market over the last two years in particular. Memory has re- cently become cheaper, processor power and disk capacity have increased and therefore, you as customers are rightly expecting to pay less and less for your hardware products and honestly, no- body is willing to pay for the software. I would like to say that I do not believe that this is the end of the story. In my opinion, there has to be a limit to the price reductions. I believe that we are approaching that limit even if we have not already reached it. There have been many warnings over the last few weeks that RAM and memory component prices will rise over the next few quarters and there may be some price increases in the near future. Today, a typical PC configu ration for a RIP will look like this:


Pentium, 133 MHz processor




2×1 GB SCSI hard disks


1 free PCI slot


CD-ROM drive 


Windows NT 3.51 operating system


In the not too distant future and by future-I mean virtually now-I believe that we will be looking at a configuration that looks like the following:


Pentium pro 200 MHz processor 


192 MB RAM 


1 Free PCI slot 


CD-ROM drive


2×2 GB SCSI hard disk


Windows NT 4.0 operating system.


And all these for virtually the same price. I believe that all customers throughout the world should not just be looking at the cheapest solu tion, but the most cost-effective. This is not the same thing at all. One point that needs to be bome in mind about these words is that you purchase a system that will do the job for you. I can tell you now that India is one of the cheapest places in the world to purchase prepress solu tions. If you look at the total solution and not just the cost of the hardware, please ask yourself, are you ultimately doing the correct thing by driving the prices down to what, in my opinion, are far too low. Do not just consider the initial capital costs. You need to bear in mind that your costs do not stop on the initial capital outlay. Somehow, the cost of installation, the training, the warranty support and continued support has to be paid for. Eventually all companies have to make a profit so that it can be reinvested for further development. No reinvestment means no new development. No development means no new products. Is this what the prepress industry. really wants or indeed can handle? Even the largest prepress giants are struggling to make profits. None of them are making the profits that they were making only a decade ago. So let us go back to the definition of the term ‘cost-effective. The reasonable cost of a system which will perform the task for which it was purchased. This is a fairly generic term, so how can this be relat- ed to the imagesetter solutions available today? We are all familiar with the term, from some years back, Wysiwyg. We now have to start think- ing along the lines of ‘Wypiwyg. What is this? What you pay is what you get.


Let me stress one thing very clearly at this point. I firmly believe that in certain circumstances, the cheapest solution is also the best for that particular need. What you need to ensure, how- ever, is that you have defined your own needs and that your new system will match those needs. Analyse your needs first and then speak to the companies which can handle these problems with some sound professional advice and then help you implement the solutions. Towards the end 


“Axel Springer Verlag in Germany is the largest newspaper publisher in Europe. Their major production is the Bild newspaper, which has a daily circulation of 4.6 million copies. This is significant by anybody’s standards. Obviously time to issue was one of the major considerations in their buying decisions.”


of this presentation, I have made some attempts to broadly outline certain parameters that you have to consider while finalising your needs. Another question that we should be asking ourselves before any purchase decision is finalised is “Are the software solutions currently available taking full advantage of the latest hardware technology?”. The common belief is that the faster the hardware platforms perform, the faster the production. The reality is that this is true only to a certain extent. So it is the software that is of particular importance in this scenario. There is no point having a multiple processor, for exam- ple, if the application programme is only capable of handling one processor. The reverse is also true. So if you see a multiple processor and if the application is using a single processor it is not going to derive any benefit. At the same time a single processor where the application software is using multiple processors is also not going to give any benefit or advantage. So it is only the multiple processor with application software using multiple processors, that is going to give the real advantage. So this is actually the thing which you have to be looking at Linless the two are working together you will never get the best pro- duction throughput. I would like to be specific when we talk about the impact on productivity and therefore I would ask you to bear with me when I mention one of HyWay Ferranti’s customers. Axel Springer Verlag in Germany is the largest newspaper publisher in Europe. Their major production is the Bild newspaper, which has a daily circulation of 4.6 million copies. This is significant by anybody’s standards. Obviously time to issue was one of the major considerations in their buying decisions. Initially they required to use our RIP on a Spare 20 platform. When they realised how much the speed had increased on these platforms they decided to use Spare- server 1000’s. These are very powerful platforms, but they are also very expensive. But when they realised that after we had re-configured the RIP, their production was going through four times  faster than previously. So to them this was cost- effective. They now have 45 numbers of cost-ef- fective platforms, each costing 50,000 dollars. These platforms have got 4 CPUs, 7 rasterisers which means that they can output to an image- setter a 5 colour page fully separated in around 1 minute 40 seconds for all the five separations. There they are using Autologic / Triple I 3850 output recorders. HyWay Ferranti provided the software integration technology route in order for them to achieve this. We are very much aware that we need to keep pace with processor devel- opment. Every time a change is made to the pro- cessing speed and power, it involves us with some form of reworking of our software. It is not a task that we wish to do frankly we have liter- ally no choice.


So what should you be looking at is:


  • Application software and hardware plat- forms to match and complement each oth- er.


  •  Software that will enable you to grow as your needs change.


  •  A reliable local support company.

That is what is required, so first you must decide:


  • the selection of RIP that can process all of your jobs to meet the deadlines. 


  • the imagesetter that meets your needs in terms of speed, resolution and quality. 


  • the front-end system that will work with the imagesetter


  • the extent of training that you need 


  • the reliability that you require


  • the company or companies that you will pur- chase the system from


  • the applications packages and compatibili- ty of local language fonts to produce your type of work


  • your productivity requirements


  • and a full understanding of what is cost-effective to you. 


So I will briefly summarize what are the con- siderations to keep in mind while selecting the PostScript RIP. It could be the type whether it is hardware RIP or software RIP but nowadays the hardware RIP is getting obsolete and more developments are carried out on the software RIP and when you consider the software RIP, what is the platform, you would be looking at, whether it is Sun or PowerMac or IBM PC or Dec Alpha and the operating system related to  that. And the interface for the image recorder whether you are using a RIP and the recorder whether the interface is proprietary or the In dustry standard, PCI or SCSI, Sbus or Nu Bas and what is the data transfer rate between the RIP and image recorder. That again depends on the type of interface, and the number of output devices that a RIP can handle, the vari ety of imagesetters that are supported by the RIP and the Level-II implementation whether it supports all the Level-Il parameters. A user friendly operation and the speed of interpre tation and rasterization for colour pages, and facility for output control like film feed, me dia cut, film saving options, warning for end of media, cassette sensing, etc. and the calibra tion to balance the dot / exposure gain, re-


corder laser intensity and the test calibration files. Support for PS fonts like Type 1, Type 3 and TrueType and necessary tools for down- loading the printer fonts Screening mecha nism for monochrome and colour and support for AM and FM screening and the integrated spooling facility to handle multiple jobs in queues, and the connectivity of the RIP to front- end systems with the appropriate networks. Connectivity to RIP’s at remote sites for multi- ple edition publications and other characteris- tics like modular design, easy upgrade path, parallel processing to increase the speed, per- formance, and the upgradation in a cost-effec tive way.


  • Then the second point where we have to con sider various parameters for selection of the im age recorder. Basically what type of application? Which is the business that you are in actually? Whether you are in Newspaper/Magazine/Job Bureaus/Repro Houses? 


  • The image area, size.


  • The speed to match your productivity. 


  • The resolution required to meet your quality needs


  • The recording source (red/infra-red laser etc.), and the cost of media and its recurring expenditure. 


  • Quality: Whether your are looking at a capstan or drum. The screen ruling, accuracy and repeatability of the image recorder and spot size.


  • The design of the optics and the lens to give consistent quality


  • The material cassette Le how it is built in. 


  • The film punch registration facility and  


  • The cost of maintenance and spares.


And what are all the parameters to consider while selecting the front end system. What types of jobs are you handling?


  • Whether it is monochrome or colour. . And, the ratio of colour to monochrome.


  • What platform, whether it is Mac or PC The connectivity. The type of networking software, Novell/WFW 3.11/Windows NT/EtherTalk, etc, and the network cards you are using.


  • The traffic in the network.


  • And probably segmenting the network and the use of OPI to get smooth workflow.


  • And whether you need for high speed, NICS like 100 base T and switches, depending on network traffic based on the estimated vol- ume of data flow in the network.


  • The facility for image/document databases and archiving as is commercially required for newspapers and magazines.


  • Interface to news agencies like PTI / UNI. • Interfacing to the RIP hardware platforms with the appropriate communications soft- ware.


  • And finally the type of network cabling that you would like to use, i.e. UTP/ thin ethernet etc.


So, briefly we have summarised some of the check- lists for considering the selection of the imageset- ter. So if you follow all of this with professional local suppliers, you will finally achieve what you want: imagesetter solutions and productivity at a cost-effective price.

 Thank You!




Burjor Pola Mr Sankaran has presented this paper on behalf of Mr. Paul Foster, who unfortunately could not be with us this afternoon, I am sure you have questions with regard to selection of imagesetters, and the selection of the right equipment. The thing that puzzles me-and perhaps I am not an exception, is that we have being given during this discussion a lot of points to consider, how to come upon decisions, and what kind of equipment to select for needs. How does one go about collecting this information? Are there any consultancies available or how does one go about finding out? Even if we know our needs, how does one find out which is the right equipment, the right system and the right platform for our needs? How do we go about it?


Sankaran In fact, I was working in Monotype In- dia for almost ten years and we were doing couple of major system integration projects for news papers mainly and when I left Monotype India, I decided to continue on the same business and Grafsoft Systems Integrators, by the name of the company itself, we don’t ny to promote a particular imagesetter or anything, although we are tied up with HyWay Ferranti mainly for the RIP and the OP’I software. But in terms of the image setters, the type of scanners we provide a total system solution. First we discuss with the custom- er what is the quality level, what is the productiv ity he requires, and then we try to give a configured system based on his needs. So that is the way we are trying to give a system solution. We believe in giving a total system solution. It could 1 be that, they can source the imagesetter, they) can see the RIP, we demonstrate the RIP and take the benchmark test and then if he is satisfied, then we try to configure the system based on actual need. This is the way we are trying to operate and we have been fairly successful and a couple of customers are quite happy with our approach.


Burjor Poonawala Is this supply mainly to larger c business houses or middle business level entrepreneurs can also approach you for similar solutions?


“Even if we know our needs, how does one find out which is the right equipment, the right system and the right platform for our needs? How do we go about it?” 


Sankaran Yes, we have even configured a system for a customer in Madras, for a small company. using a DTP system. They wanted to move on to this. We have given solution even for the new en- trepreneurs as well as people who want to switch from DTP to imagesetter, large newspapers, etc. We are open for any type of work.


Burjor Poonawala In the prepress industry it is a well known fact that the equipment is fairly ex- pensive, at least in this country with customs duty at this rate. Would this consultancy add significant- ly to the cost of purchase or how is it?


Sankaran Not really. It is basically a relative term but it is not that expensive.


Burjar Poonmunia Exactly. So gentlemen you have a choice of taking advantage of Mr. Sankaran sit- ting here for some free advice. Why don’t we ask a few questions and get a few answers and it might help us.


Participant This is regarding the hardware RIP and software PostScript RIPs. Now suppose I want to drive multiple devices, suppose I want to drive a CLC, a film recorder, or a large format Inkjet printer, and one big electrostatic printer. Do you suggest individual RIPs or a single RIP which can do the job for all the machines?


Sankaran That depends on which RIP you are selecting, for example, if you look at the HyWay RIP, you can with a PCI type of interface support a cou ple of image recorders like the Agfa, the ECRM, the Linos. They are even prepared to develop an interface for a specified output device and of course, the development cost has to be borne by the customer if it is a very specialised image re corder. So with one two-PCI interface, one RIP can drive two image recorders and of course, Monotype has introduced the laser bus technology by which one RIP can theoretically drive 8 output devices. That laser bus is a proprietory in- terface so the laser bus concept is only for the Monotype supported image recorders. So if you are looking at the industry standard interfaces, HyWay Ferranti is coming with a SCSI interface. Theoretically, the SCSI can address seven periph erals. So you have that choice available.


Participant What you are telling is about the in- terface?


Sankaran Basically the type of interface between the RIP and the output device. You have to see whether you are going to use a proprietary one or one of the standard interfaces.


Participant That is what I want to know. Should we go for a proprietary system or should we have a open system?


Sankaran HyWay Ferranti always believes in open systems and they always develop and configure the RIPs as per the industry standard interfaces avail- able and they are coming with SCSI drives for all these recorders. But one thing you have to keep in mind. At a time it can address only one output device. So if your is going to be a key issue, you have to go with individual RIPS.


Participant I want to know the difference between the quality of capstan imagesetters and internal drum imagesetters for colour blends and trapping, because it seems that the internal drum setter gives 


“Now suppose I want to drive multiple devices, suppose want to drive a CLC, a film recorder, or a large format InkJet printer, and one big electrostatic printer. Do you suggest individual RIPS or a single RIP which can do the job for all the machines?”


good quality for colour print and trapping.


Sankaran Internal drum definitely has advantages in terms of repeatability, in terms of registra- tion quality, and it is far superior. The way the tech- nologies developed in the capstan and drum, they are entirely different. So obviously an internal will give higher quality and there is, in be- tween these two, the virtual drum which means the film itself forms the shape of a drum, which is called the virtual drum, introduced by Exxtra corporation. So that is a compromise between these two; for example, the repeatability available on a capstan typically is about 1.5 mil. and the repeatability which you get on a internal drum is less than 0.5 mil. and the virtual drum gives 1 mil. repeatability. So definitely you have, depending on your budget, the advantage of choosing the type of output device.


“Basically the RIP is the heart of the imagesetter. I would say. The imagesetter is only a slave, it is totally a dumb machine. What the RIP gives as the data, it just records. So the entire heart of the system is the RIP, the selection of the RIP has to be critical.”


Participant And do you have any punching system for capstan imagesetters, capstan type imagesetters?


Sankarun Yes, there are couple of them available.


Participant But for many four colour offset ma- chines, there are 5 or 6 methods for plate punch- ing.


Sanharun Yes, in fact, couple of imagesetter manufacturers, they do support the standard Bacher punch and other things and they also give you a custom punch for which they charge something extra, so you have the options available.


Participant Is it available for capstan also?


Sanharun Yes, it is.


Participant And I want to know whether any difference is there if you connect through a fiber optic for imagesetter output instead of Ethernet 100 Base T.


Sankaran Fiber optics as of today, is a very expen- sive proposition.


Participant Does anyone use fiber optics for image output connectivity?


Sankaran Between the RIP and imagesetter, to my knowledge so far, I have not come across any one.


Participant What are the advantages of fiber op- tics?


Sankaran You can get much higher speed. Data transfer is much faster.


Participant But if the RIP is faster to that extent, even Ethernet is fast enough for transfer speed?


Sankaran But basically when the RIP is fast and the imagesetter is also equally fast, then to balance this the transfer rate should also be proportionately increased.


Participant So the recording speed will be great- er in a virtual drum imagesetter or an internal drum setter?


Sankaran Virtual drum is still not that fast. Of course, the Exxtra Corporation is coming with a faster speed some time in January and they are also coming also with a wider format. (In response to a voice in the audience) Yes, in fact, it is already introduced.


Participant And I want to know whether the RIP is responsible for the type of output we get like flexo, or gravure or offset printing, or the image setter is responsible for the kind of film output?


Sankaras Basically the RIP is the heart of the im agesetter, I would say. The imagesetter is only a slave, it is totally a dumb machine. What the RIP gives as the data, it just records. So the entire heart of the system is the RIP, the selection of the RIP has to be critical. You have to be very careful in looking at some of the parameters, as I have out- lined.


Participant And do you have any advantage of hav ing an on-line processor with an imagesetter or does it matter if we take the output film separate ly and send it for processing?


Sankaras In my opinion, there is absolutely no difference whether you have an on-line processor. When your productivity is very important, people go in for an on-line processor so that you dump the job and it comes out processed. On the basis of whatever information I have, one disadvantage of the on-line processor is that there is some dis tance between output and the mouth of the pro cessor, so there is likely to be some amount of film wastage for each batch. So if you are sending a large number of pages in one-go then it may not seriously affect.


Rajesh Gulshan I will address the problem which actually Mr. Poonawala mentioned. Basically, I would like to draw the attention of the audience to what happened to Local Area Networks (LAN). operat LANS came because they were different ing systems and when they came a number of ver dors came and got their own LANs and they sold like hot cakes such as Vines and Novell. You have umpteen number of LANs which came but then along with that, came some kind of standardisa- tion. When they came with a 7 layer structure, they started coming with some kind of standardisation and that is where one was able to identify and prob ably do consultancy and express, how things will move as far as LANs are concerned. A similar thing is going to happen in the prepress industry also. Unfortunately here the equipment is large like scanners, work-stations, imagesetters, but there is no standardisation. There have been different companies following different formats. Every ven- dor follows his own format. So when a time will come where some kind of standardisation will come, then and only then you will be able to get some kind of feed-in as far as consultancies are concerned and how to get an equipment, what prices to pay for an equipment, and what Mr. Poonawala said is very right, the equipment is very expensive. There is no doubt about it. So it is only once the standardisation comes there’ll be more people to maybe inform bet- ter and you will be able to buy a system fully integrated at low prices.


“Front end standardisation is already there. In Mac/PC it is already there. I am talking about the back end where Crosfield has its own format, Scitex has its own, Linotype has its own; so it is only once there is some kind of a standard platform that they will come to, then only you will probably be able to address this kind of situation better.”


Sankaran I am not able to follow what you really mean by the standardisation of the front end sys- tem.


Rajesh Gulshan I am not talking about standardi- sation of front end. I am talking about the standardisation in the proprietary area or back end systems. Front end standardisation is already there. In Mac/PC it is already there. I am talking about the back end where Crosfield has its own format, Scitex has its own. Linotype has its own; so it is only once there is some kind of a standard plat- form that they will come to, then only you will probably be able to address this kind of situation better.


“The type of hardware and the software have to complement each other. There is no point in going in for fast hardware if the software cannot handle that kind of multiprocessing.”


Sankaran But slowly, you know, people are moving into standard interfaces.


Rajesh Gulshan That’s what I am saying. That’s why I gave the example of LAN. In LAN also, it is com- ing now, when they set a 7 layer structure, and how they set ISO standards for each and every layer. And as far as on-line processors are concerned, I think since we use it in our process house, it is of great help especially from the productivity angle.


Burjor Poonawala One thing is for sure with on- line processors, that you would not get finger marks anywhere!


Participant Regarding networks when Ethernet came, the speed was 10 MB/sec. i.e. that is the rate at which the data was going across the network. Then there was a fiber optic standard also called FDDI that offered about 100 MB/sec. speed and it was an expensive solution, when it came. Now today in the market you have got fast Ethernet, which also gives 100 MB/sec. So speed-wise it was comparable to FDDI, because speed is 100 MB/ sec. which is 10 times faster than the earlier Ethernet network. So you don’t get any specific ad- vantage by going for fiber optic cable that’s called FDDI standard, which is a very expensive standard.


Participant Mr Sankaran spoke earlier about the Spare 1000 being used by a German publishing group for upgrading its speed. Was it the sume RIP or was it a different RIP?

Sankaran A RIP has to be reconfigured. The type of hardware and the software have to complement each other. There is no point in going in for fast hardware if the software cannot handle that kind of multiprocessing.


Participant in the Sun Microsystem packages when they were introduced, most of the software packages started working straight away even without recompiling. 


Sankaran But do you get the advantage of the high speed hardware?


Participant You do get it. For example, when you have a 386 machine with some application run- ning and you buy a 586 machine the same application definitely runs faster because the proces Or runs Easter. But the whole program may not be optimised for that particular processor, but still it is faster.


Sehr Right. It is a relative term. Whether you are happy with that or not is a different matter.


Participant The Germany company was using the HyWay Ferranti RIP or a different RIP and was the old software straight away ported onto that machine and used or was it optimised and recom piled for the new machine?


Sakaru It was optimised for the Spare 1000.


Participant How does that compare with respect to the performance of the Spare 1000 RIPs with the Pentium 586 133 MHz and 200 MHz. How do you compare the same RIP working on Sun Microtems workstation platforms or on the PCs? How do we compare since some of us would like to select on the basis of performance since the Sun Sparc 1000 is an expensive piece of hardware and also the maintenance will be pretty difficult unlike the PCs which are antaibable commercially. Can you give us some approximate performance fig- ures? Even if it is like 5 times, 10 times, approximately?


Sankaran I can give you information on some recent tests I have done on a Pentium 133 MHz with a Windows NT RIP since I cannot run a Spare 1000 RIP. I have done some recent trials in a newspa per using a 50 megabyte colour page with four colour separations on our NT RIP driving an Ul tre machine and they had earlier been using a Hyphen DOS RIP. Right from the time you gave the print command it took 26.5 minutes. When we output the same job with the NT RIP from Hyway Ferranti it took about 10.5 minutes. So that. is the speed difference.


Participant You are talking about two different RIPs. I am talking about the difference in speed of the same RIP on two different platforms, for instance a 586 and a Spare workstation. And how do the software prices compare for the RIP? This would probably be interesting for many people here.


Sankaran Basically I don’t have any data immediately for the Sun Spare but I can probably send it across The RIP software prices for both the Sun Spare and the NT are the same.


Hemant Desai, Yaban The question about software RIPS We are also the vendor of Harlequin RIP’s We have been selling on a PC, Windows based, we have seen the performance on 386, 486 as well as Pentiums these days. Those who bought RIP’s for the 486, they moved onto Pentium. 


The op erating system still remained Windows 3.x. As long the operating system remained the same and the platform only changed, you could put the same 486 RIP on to the Pentium without paying any sin gle paisa to the vendor. But the moment you change it over from Windows to Window NT, or Windows 95, which are to some extent different operating systems, then you need to upgrade your RIP’s. Then you need what is optimised for Win- dow NT, or Windows 95. If you are already using a software RIP, and you change the platform, you don’t have to go back to the vendor as long as the c same Windows operating system remains.


Participant This is a very basic question. Suppose I have an image file, which is already a bitmap. Do I need to RIP it again? It is already in the form of pixels, say, for example, I have a file which is about 4092 by 2564 pixels, a 32 MB file. It’s a TIFF image file format. None if I want to make colour separations out of it, or I want to record onto film. should it be RIPped again ?


Sankaran You have to RIP it. If it is in TIFF for-


mat you have to RIP it again. Participant PS RIP should come in when the file is in the PostScript format. TIFF image format is not PostScript. The image is already in the form


of pixels, so what is the role of the RIP?..


Sankaran Basically, you want to drive the laser engine so you how to have a rasterisation process in between.


Participant What is rasterisation? It is very basic but I want to know.


Sanluran Rasterisation is basically, something like a TV raster. Your laser beam has to be turned on and off according to the data available from the front end system.


“The operating system still remained Windows 3.x. As long the operating system remained the same and the platform only changed, you could put the same 486 RIP on to the Pentium without paying any single paisa to the vendor. But the moment you change it over from Windows to Window NT, or Windows 95, which are to some extent different operating systems, then you need to upgrade your RIPs.”


Participant again So far I have a different view of raster. For example, if you take a PostScript file, it will be stored in the form of formula. It is not in the form of pixels, where each pixels has got a numerical value, that is actually the pixel image or the bit mapped image. A PostScript image is not in the form of pixels, it is in the form of formulas, for example a curve. I have to store the curve in the form of an equation for the curve. Now there is a circle in the image. It is stored in the form of equation for the circle. So it is not stored in the form of pixels grids. So the RIP is required to convert those mathematical formulae into the printed dots. Is that what is called a RIP? Can you also tell me what is a PostScript file?


Sankaran I am hearing this approach for the first time. I cannot comment on that. But basically our understanding is that the PostScript file basically defines the page coordinates. It is basically page description language


Participant My page has an image, nothing else. There is no text, no font, nothing else. It is just an image. It is stored in the form of numerical components. This is a very basic question. It is worrying me a lot, that’s why I have asked.


Participant The basic RGB format where you have 8 bits for Red, 8 bits for Green, 8 bits for Blue. You have a file say 1000 by 1000 and when you print it, the printing side you have colour separa- tions, yellow, cyan, magenta and black. Now each picture element is decomposed into different co- lours, yellow, cyan, magenta, which is one aspect of this. After you convert the image into the yel low, cyan, magenta, black, the picture elements are subdivided into dots which will either go to an inkjet printer or to film, later the film is taken for the printing press.


Participant I don’t think the picture is subdivided 


“TIFF is not just a raster, it is also a way of storing raster, whether it is black and white data, colour RGB data or CMYK data. It is in a way of very trivial kind of language.”


into dots, because you see, for each LPI you want two times the pixels, where is the question of subdivision? For example, I want to print at 150 LPI, you scan at 300 DPI, that is the factor of 2 For each LP’I, you got 2 DPI. For one print dot on the printer, there are two dots in the pixel file. Where is the question of subdivision?


Sunil Khullar Actually it is a very interesting question. What he is saying is, that his understanding of a RIP is when you have a higher level language like PostScript where everything is divided into formulae and finally your output device is nothing but raster and the RIP’ rasterizes. What he is asking is, suppose my input data is rastarized like a TIFF file which is already rasterised, why should I RIP it again to a raster? The answer to that is that TIFF is not just a raster it is also a way of storing raster, whether it is black and white data, colour RGB data or CMYK data. It is in a way a very trivial kind of language. Now, an out- put device, even if it is intelligent, though out- put devices are not intelligent unless they have a RIP, even if it is intelligent like a laser print– er with PCL, it understands a different language, so yes, you won’t require a RIP to con- vert a TIFF to output it, but you will require some program to read this TIFF file and read one raster and convert it to the raster which the printer understands. Now this program, if it is a high level program, you might call it a PostScript RIP or it could be a small program which laser printer manufacturers use, to test their laser printers.


Participant I mean pixels is not understood by the printer. So the pixels, image pixels, have to be converted into the printer dots. How do you get shades? In the paper you see the black and white


either as dots that are there or not there.


Lakshmivenkataraman I will read out one small sentence. “A PostScript RIP has two functions, one is to interpret the PostScript code and render that code into bitmap that the marking engine can image. During interpretation, the RIP processes the PostScript file to build dictionaries and paths. fetches fonts and other resources, and analyses the elements of the page. “In a high performance RIP such as used in imagesetters and colour record- ers the output file of the interpretation phase is something called the “display list.” Now your ques tion is if a bitmap file is there, it need not be RIPped, because it is already interpreted as a bit- map and that’s what is required for the engine to mark it.


The only thing, whether the bitmap is in the form off’s or TIFF’s is that the bitmap should be recognised by the output device, that there is an understanding, that the file conversion reading is there. But once it sees the bitmap, it need not RIP it again. However, if the resolution of your bitmap file and the resolution of the recorder is different or requires a change, the image has to be RIPped again. Otherwise a moire pattern will be produced.


Participant That is exactly what I want.


Venkataraman, Wipro Infotech In fact, I want to clarify what the other gentleman wanted to know as far as performance differences between Pen- tium and Sun systems. In Sun there are two pro- cessors which are available, one is called the Mi- croSparc, the other is called UltraSparc- now the typical Pentium system would cost around Rs. 15 lakhs, and an entry level Sun Sparc system would cost around Rs. 4 lakhs. The differ- ence in performance (these are ball park figures), would be 3 to 4 times more. Now an Ultra Spare system which is one of the latest systems available today would cost you about Rs. 6 lakhs, which is 4 times the Pentium system but you would get anywhere from 6 to 10 times the performance. This is only assuming that a RIP uses a lot of floating point functions where the Ultra Spare is far, far superior.


On the performance of the PowerMacs vis-a- vis Ultra Spare systems? In fact I would put it as the PowerMacs giving you the 60% of the per- formance of any of the Ultra Sparc System. The newer PowerMacs would be comparable to some of the Super Spares which is a slightly older gen- eration of Sun Sparc systems. The PowerMac 8500 I would put it as 60 to 75 % of Sun Spare systems (I am only trying to give you some ball park figures). If you put across configurations to me I would say it would be only 60 to 75%. Mac has got its own unique capabilities but as far as raw performance is concerned, I think the Sun is the ideal platform.


Introduction of R.P. Lakshmivenkataraman by Naresh Khanna


R.P. Lakshmivenkataraman is a young person whom we first met when he made a presenta- tion at the IFRA seminar in Bangalore a cou- ple of years ago. Actually he is an industry veteran who has been with The Hindu since 1982. Trained as an electronic engineer, currently he is a joint controller of production at The Hindu.

R.P. Lakshmivenkataraman An Update on Imagesetters


I’d like to share my technical views and I have pre- pared something about imagesetters. Imagesetters as you know, are standard output devices, normally they are PostScript, they need not be postscript all the time, but generally we talk about PostScript output devices, exposing on a photo-sensitive material. The differences in cost, quality, and speed originate mainly from the design of the ma- chine. The variations in the design are for exam- ple, in the film transport mechanism, or the light sources and the exposing technology. See like magneto optical rails, lead screw movement and the speed of the lead screw rotation, RPM, whether internal or external, how many beams it records simultaneously. Those are the things that decide


the cost, quality, speed. Roller mechanism feed normally we call it capstan type of imagesetter where the film is held between rollers, and the laser beam sweeps across it. And the other one is drum. There are two important classifications of imagesetters. Again in the drum type, there are internal and external type. Of course, lots of developments have taken place in the external drums which are based on the high end scanner recording section. Internal drum, people claim, was developed long back, but it is being continuously developed well so that it is becoming more and more common among im- agesetters. Many companies are making internal drum imagesetters, and of course external drum imagesetters are also there. One of the external drum imagesetter manufacturers claims they are faster having multiple channels of light. And the other technology is the virtual drum, which are also in the market although I have only read about them. The various factors of our concern in look- ing for an imagesetter are the quality, cost,


throughput and versatility. The quality of the imagesetter can be decided by comparing the separations and the halftones whether it is free from any streaks, or any fog in the shadow areas, or any pattern in the half- tone areas. Some manufacturers say my imageset ter has a spot size 5.6 microns. Let it be 5.6 mi- crons, the issue is whether it produces a clear re- sult. And the second quality issue is repeatability, when I expose the first colour with a register mark, and when I expose the fourth colour, is it capa- ble of giving the same register mark in the same position? This is given in the order of mills and we have to look into it. Then the speed. When we think about the speed, it is not the engine record- ing speed only, we have to look at the internal spinner that rotates at 20,000 RPM or 25,000 RPM. O.K. that is also an issue but that is not the only one. It is one of many because the recording effi- ciency comes from the RIP also. Unless the RIP feeds data to the mechanism it cannot be record- ed. Most of the imagesetter engines are faster than the RIPs currently, at least practically. How fast does the RIP feed the data to be recorded? Most of the time the engine waits for the RIP to send the data. That is generally the status although there may be hardware RIP’s which are faster in order to match the recorder speed. The speed or the rate at which the laser beam deflects, film travel, rotation of the cylinder, movement of the carriage, and the number of beams recording, decides the throughput. And the versatility of the machine is the film handling, whether it can handle day light film and the carrying format of exposures and having more than one storage bin from record- ing that is whether the machine has one input cassettes or can it have 2 or 3. And when it comes to the repeatability and the registration, whether we have a punch, that the punch can be matched with the plate making punch also. At the time of recording whether the film is held in the punch. Those are the factors to be looked into.


The quality of performance of an imagesetter is a function of both the RIP and processor. The RIP is characterised by the data input and output and the networking connected to it. Ear- lier we talked about the optical interface, and 100 base T, and 10 base T, and spooling from more than one client, that is printing from all clients by placing the output file in a spooler. As far as processing power is concerned, people speak of Risc processors, Pentiums, and Spare 1000, thus the processing power and the memory, the amount of RAM and whether it is a multi-tasking RIP and whether it has buffers for supporting the engine and whether it supports two CPUs or multiple


“How fast does the RIP feed the data to be recorded? Most of the time the engine waits for the RIP to send the data. That is generally the status although there may be hardware RIPS which are faster in order to match the recorder speed. The speed or the rate at which the laser beam deflects, film travel, rotation of the cylinder, movement of the carriage, and the number of beams recording, decides the throughput.”


CPUs to do parallel processing on the same job. Those are the factors to be looked into when we consider a RIP and RIP performance. 


Imagesetter engines consist of two subsystems, one is media handling and the other is the opti- cal system. The media handling is mainly decid- ed by whether it is a capstan or drum. Again, when we look into the imagesetter, the issue is what is acceptable, what is our requirement, and that is the one deciding factor, quite apart from what gives us almost ultimate results, higher accuracy and all. For black and white printing on newsprint do you need an external drum? Do you need an internal drum or a capstan type? Your decision depends on your requirement. Just as in the case of charge couple devices (CCDs) a lot of devel- opment and R&D has gone into capstan type im- agesetters. People claim that their capstan imag esetters will record colour. They are calling these colour capstan imagesetters. What is the way to test for colour capability? Repeatability. Expose a job which is lengthy enough or expose one co- lour of a job just after the film is loaded and after some metres of the film is exposed, expose the other colour of the same job and check for the registration in the capstan type. If that suffices, that itself is good enough.


The light source and the exposing system are also parts of the engine. The light sources nor- mally used in imagesetters are argon ion lasers, helium-neon lasers, gas laser tubes, laser diodes, infra-red diodes and light emitting diodes (LEDs). Argon ion laser for example, is used by the Pu- rup Maestro and the Linotronic 930, helium-neon laser tube is used in Lino 560, Laser diode are used in Avantra, Dolev, Hercules, and FTR 3065. In- frared diode is used in Purup Maestro and other machines, LEDs are used by Screen DTR 3075.


What are the sources of film? What is the film cost, for instance if I choose an imagesetter with

an infra-red laser diode? And what is the availability of that particular film. Those are the factors that should be looked at


The recording light is then sized to the re quired spot by different types of lenses and mir- rors More and more lenses and mirrors will in- vite over time, more dust accumulation and scratches. Those problems will be there and should be looked into as a maintenance and even quality issue.


Some of the designers of the imagesetters keep the laser tube on the head itself. Some may keep it stationary, take it on a fibre optic cable and then expose it or put it on the spindle. Those are the design variations which we look into for the quality of the imagesetter.


What is the advantage of the capstan type? Capstan type imagesetters have continuous images, you can record any length of film and they are lower in cost. Advantage of drum imagesetters: As already enumerated, they are repeatability, reg istration and accuracy. Then what is the advantage of external drum and internal drum? In the ex- temal drum, the optic movement is simpler in comparison with internal drum recording, less complex, it has a simple lead screw movement like in our high end scanners. Another feature is that it can have more than one beam. In an internal drum, a single beam will revolve and again the internal cavity decides whether it exposes 180 degrees or 270 degrees of the internal drum which determines the speed of output. In external drums, an important feature is the number of beams, there may be 80 channels or 120 channels, stich a recorder with a movement of 1000 RPM is equis- alent to or much faster than a 25,000 RPM inter- nal drum recorder. And another claim is the close ness of the laser beam towards the film source: which is a plus point for external drums. But now again, when I see a result from an internal drum and an external drum for the same job, if I don’t find a difference does it really matter? Depend ing on what type of quality jobs we want to run, with what accuracy the results are required, the decision on what imagesetter is to be made. In- ternal drum has easier film loading and equally good repeatability compared to external drum.




The RIP is an input part of the imagesetter where it accepts any PDL (Page Description Language) file from a front end system and feeds the imag ing engine. PostScript is one of the standard page description languages. The PostScript code gen- crated from the different kinds of front end ms


“From any software, we can make the 5% patch or even 1 to 95%, record it, and see that we are getting the same percentages there. If not, we have to do calibration and any RIP will give some kind of utility to do that.”


like a Mac laser writer driver makes the PostScript file and this is accepted by the RIP and this generates a specific bitmap, that is rendering and rasterization for the particular engine. The PostScript RIP has two functions, one is interpreting the code and the second is rendering the bit- map. The RIP can be of software or hardware. Normally, earlier hardware RIP’s are faster in com- parison with the same company’s software RIP’s. At first a company’s hardware RIP is faster, but simultaneously they will develop a software RIP which in a short while overtakes the hardware RIP in terms of speed. Then they develop another hardware RIP (with perhaps newer components] which is faster. This is a chain-like thing, which is going on, as we involve the number of developments in processor like Sparc, Spare 1000, again software RIP scores over the hardware RIP


The network between RIP and engine. Here comes the connectivity between the RIP and en- gine. Definitely fibre optic cable is better, how- ever here it is not for a network. What will happen to the signal loss as it travels through a cop- per wire? But as this slide from an optic fibre ca- ble book says, a throughput of 100 MB/second is achieved with a CAT 5 cable which is laid, but there is the restriction that it should not run more than 100 metres. There is the question of cost, but what at present is standardised and fairly recom- mended is going for a CAT 5 standard with the 100 base T for networking an in-house network.


There are certain things I would like to show from the user’s point of view about RIP’s and engine. Again the calibration of the imagesetters whether it gives a correct  dot 5% to 95%. People say, “In my machine with Photoshop I set this per- centage but when it records on the film it is not there. Then they do some calibrations on the imagesetters. From any software, we can make the 5% patch or even 1 to 95%, record it, and see that we are getting the same percentages there. If not, we have to do calibration and any RIP will give some kind of utility to do that. I have an exam- ple, which is created in a separate software. This is one of the Lynotype machines, which gives a Mac utility form. You measure the percentages of  


“So the question is that when the D-Max is shifted from or increased from 4.0 to 5.0, you can obtain all the necessary dot percentage values on your strip from 5% to 95% but what happens to the text?”


each of the first steps, if this is not according to the list, you expose it without any entry on the normal value. Measure the values and enter them back. You record it, the RIP adjusts further, for the value, and it sees that it adjusts further for the optics and the processor and the chemistry. It’s a customer specific standardization, it has to go with a kind of film, the chemistry and the processor.


I want to know if any of the participants have come across a problem with this kind of calibration of imagesetters. Here is an example of another imagesetter, a proof which has calibration.

 I would like to put one question which I do not know. See, for example, when we do the calibration for the percentages and density, we are able to get the 5% to 95%. We set certain densities the film, for example, 3 or 3.5 or 4 what is rec- ommended. I am setting the density at 4, and I can still calibrate that 5% to 95% what I get. If set a density 5, still I can measure the value and get the 5% to 95%. What will happen to the text, when I change my solid density from 4 to 5? You can see on the top of the lettering, the thickness of the lettering, when I change the density setting whether the screen gradation or the calibration text for only the halftones or the ASCII text also. And what will happen when you record this as a positive and as a negative, will the width of the lettering remain the same? Anybody who knows can answer, it’s open for discussion.



Burjor Poonawala Mr. Lakshmivenkataraman has posed the question after his talk. He is certainly very knowledgeable on imagesetters, but yet he has a question. So the question is that when the D-Max is shifted from or increased from 4.0 to 5.0, you can obtain all the necessary dot percentage values on your strip from 5% to 95% but what happens to the text? Should the text remain the same or should it not. Because in his practical experi ence it seems the text tends to go a little thicker. Right? It tends to go a little thicker. So have any of you had similar experiences?


Lakshmivenkataraman The positive and negative. should be, theoretically speaking, the same when I expose a positive or negative. There is a change but still it is not known to me whether this is due to the chemistry or the processor. But for all that, I can attribute it to a RIP also, that is one of my observations. In one of my machines a D-Max changes, I mean, that with different manufactur- ers of machines and different kinds of imagesetters, there is a difference, not all the imagesetters behave the same way between positive and negative. That’s one observation.


The second thing is the thickness of the lettering variations when we adjust the calibration to the imagesetters. It does vary. But it varies about 0.02 mm around that. Under a microscope, it varies about 1 to 2 divisions. Probably many of us have observed it but just observe this when we expose positive and negative. Do you find any difference, it is in the output result [perhaps in the printed result).


Naresh Khanna I have often expressed to many people, that there are difficulties in calibrations.

Recently, I was trying to calibrate an imagesetter, or at least trying to check the output with the same kind of method but I also tried to output a 1% dot, 2% dot, 4% dot and 6% dot and I could not achieve a 1% dot at all, a 2% dot at all, I had a broken dot at 4%, I had some grey dots at 6% and then the amazing thing was that I had a jump, that is after the 5% dot, 6% dot, I had a jump to something like 12% instead of an 8% dot. In high light areas, it wasn’t a smooth transition, that was one thing. The second observation, I think is quite correct, that we are used to the fact that the peo- ple have calibrated imagesetters for positives. If we dare to sometimes, ask for a negative, then we normally get back that imagesetter negative with a gray background. Not a black, dense back- ground. The printer definitely complains. That is the normal thing. So  when you are describing a situation in which the imagesetter is able to out- put jet black, negative and positive without change of chemistry or film or adjustment of D-Max it is, I think, interesting. Some imagesetters can do it. Not getting a 1% or 2%, obviously was a problem in the adjustment of the imagesetter, I think.


Burjor Poonawala Sort of expressing an opinion on what was just said, usually positives carry less silver than negatives and it might have something to do with the processor, because you are process- ing at the same speed still and you are process- ing far more film than when you are processing a positive and perhaps that might have some ef- fect on the density of the black, the opacity of the film, but anyway, I would suggest that whether it is at 5.0 DMax or 4.0 which is the result you pre- fer? If you think that you are closer to what you want at 4.0, you continue with 4.0. And if you find that at 5.0, you get a better text and that is what 


  “…we are used to the fact that the people have calibrated imagesetters for positives. If we dare to sometimes, ask for a negative, then we normally get back that imagesetter negative with a gray background. Not a black, dense background.”


you like, you continue with 5.0. Because it is a question of splitting hairs, I think, because we are going too far, too deep into the whole thing, some- times it doesn’t really matter. It is a question of standardization, so standardize on what you think. In my opinion what you think is what you want and this is it-whether it is 4.0 or 5.0 or something in between. Whatever is your need, and that should be the right thing.


 Address of Niels Ohrgaard on Modern Workflow


This next presentation is little bit related to what we do, so I think, I would ask you to just disregard the name, and think of it as a presentation of what is possible on the prepress market today and what will become possible in the very short period of time to come. Some of the discussions, earlier on were about the trends and the discussion I refer to here is about LAN (Local Area Network), and WAN (Wide Area Network). Because, if we are talking about standards, why is there not a stan- dard for prepress for output devices let’s say. I am sure we all know that there is a standard today called IT8. This is the standard whereby we de- fine colours for input scanning, transparency and reflection scanning has a definition today. Now LAN, we were talking about LAN, the definition, or the standards that people tried to achieve was called OSI, as also mentioned a 7-layer module. This 7-layer module, actually described various things about how to handle interfacing between different equipment and how the inter-connec- tivity should be started on the low level by describ ing the physical interface, then going up in the layer module, describing the protocols, then go- ing even further up, describing the headers be- fore data information sent, etc. But we have to look at it from different angles because what did it ac- tually give us? We have been talking about OSI for 15 years or more and I think the problem with that standard actually was that it was not user-driv en. It was a standard set up by international or- ganisations, trying to drive users in one direction. I can only mention that OSI at a point in time was actually going to takeover completely for TCPIP. TCPIP was dead and gone and was not going to be used any more. That is not the case. TCPIP is well, thriving, a lot of development is still going on with the TCPIP protocols. So rather than look- ing at it from the standardisation point of view, standards should be driven from end users. And into networks, what was the drive from end users? The drive was: I should be able to sit in India, in my office and working on my digital platform- not a computer-just a terminal. This digital ter- minal should be able to-via the network-Lo- cal and Wide Area Network-connect directly to my digital host in Copenhagen. But more than that, it should be able to communicate with my IBM host placed in Australia-in Sydney-wher- ever. What the network actually gives you is total inter-connectivity, that is what it should give you. I think that is where OSI actually went wrong be- cause they were pushing for a lot of issues that did not originally give end users a great benefit. In- ter-connectivity was a benefit that most users were looking for with their networks and the inter-con- nectivity of IBM terminals, digital terminals, Wang terminals, whatever, to any computer host linked to the network.


I think that’s one of the problems we are fac- ing in the prepress industry today, there is not enough drive from the users, I think, it is difficult to get a focussed drive from end users in prepress because from output, let’s say imagesetters, we had discussions about formats etc. and there are so many different print machines or presses installed throughout the world today. So it is difficult to come up with a set of standards in a very short period of time, that will actually give all of us a set of specific standards, we can work from.


What we want to be is that in the ’90s, in the late 90’s we would like to be the world’s biggest prepress vendor and we are striving to become that, but at the least, our primary aim is to be a digital image capture and imaging company. I think, that today we are manufacturing modern professional high end colour reproduction equip ment, especially solutions as we would like to call it, and we are placed at mid range price point. I think, as I mentioned, the name that is in the pre- sentation is not the real important thing. The important thing is that where we are headed and where we are going and what is actually possible to do with various equipments we have in the market today. You remember the other slide I had where we saw that the flatbed scanner is taking the quality of colour down before it expanded and went up further in terms of quality of colour. The pioneers those people who have been on the leading edge who have invested in desktop pub- lishing few years ago-they are actually now pro- fessionals in the market. I think, the discussions we are having, whether this imagesetter or scan- ner is better, faster or having high quality, it’s all very interesting. But there comes a point in time where the pioneers in the business-I am talking world-wide, I am not talking just locally-what they are looking for is not better or higher quality prod- ucts, what they are looking for is all the addition- al equipment to make their workflow higher, to make their productivity better.


From time to time, we hear of some very good new software products coming to the market that can actually help with certain applications. We have heard about the Copix today which definitely will help in many situations, but in the way that when you are looking at the manufacturer, you should look at the innovation from the manufacturer as well, because it gives a good indication of where that manufacturer is going to be in the years to come. We heard earlier something about image- setters and I am sure most of you have experience with imagesetters and on-line processors. I know it is an invention from the devil himself, there are always some problems because of the long path of film as we were told here. So why not take that processor and build it straight into the imageset ters so you actually have a device which is like a big laser printer, you have the imagesetter, you …there comes a point in time where the pioneers in the business-I am talking world-wide, I am not talking just locally- what they are looking for is not better or higher quality products, what they are looking for is all the additional equipment to make their workflow higher, to make their productivity better”


have the conveyer system and the processor built into one. That is actually available in the market today. The mature states, those people who have had the equipment for some time, they are look ing more at expanding in terms of applications, and the workflow. They are looking at the initial expenses and expenditure for equipment, what is the cost of running and of course, how can I make more money with the investment I originally made. The core operational equipments, there are some processes, nobody can avoid today in pre- press. We are talking about input capture, on the other side, we are talking about output imaging. furthermore, we are talking about scanning soft- ware, how do we keep control of the colour, how do we make a proper RGB to CMYK conversion, how do we scan in CIE-LAB in all the various stan- dard file formats that are available in the market, how do we handle our PostScript, what RIP are we using, what screening technology, and of course, one of the very relevant issues today and for the years to come is the input workflow. How do we treat all the information that we are actu ally getting from our digital camera or getting from our scanner and how do we handle it from our output workflow point of view, And the solu tion that we are striving to get at, is actually an image server concept.


If we’re looking at the output imaging, what are the trends that we experience around the world. We can talk about output or we can talk about input and output when it comes to the first remark up here. There is a great tendency that customers or users of digital prepress equipment, they are going for the one-stop shops. They are not prepared to go to a lot of different vendors, and get one product from one side and another product from the other side and then have to go through all the hassle if the product does not work etc.


On the other hand, we have also seen that in general, buyers of today are much more aware that they have to invest in equipment that is truly non- proprietary, meaning that if today they need to expand their business, they need to get the additional scanning capacity, the investment they have made in imagesetters, for example, has to be on an open platform so they can buy any scanner not necessarily from the same vendor, if they are happy with the vendor of course, they buy a scanner from him. If for any other reason another vendor has amore likely product or whatever, they should be in a position to buy from anybody in the market.


In-plant production is coming up very fast, we are seeing the larger corporations, they actually take in-house, not only position scanning, they do the scanning themselves. Some of them are do- ing short runs in-house. Some of them are even having imagesetters in-house. We think that the printers today and I would estimate that world- wide 50% or even more of our customers today, are printers-they are trying to trim their fixed expenses and we can also say that they have done it quite successfully over the past years. We see a lot of investment from that market. The purchase being made at the moment is primarily in the four up market whereas we believe that a two up im- agesetter either as a backup or dedicated to mag- azine work, is a very flexible solution as well. What is coming up is short-runs very heavily, we have seen all the Indigos, we have seen the Xeikons coming in the market and furthermore, in Europe there is great trend towards a technology called polyester plates. A polyester plate is the plate that most imagesetters can run and they expose directly on the polyester plate instead of going to films and plates we take the polyester plate and put it directly onto the press. Now this is very good answer to the Indigos, etc., if customers already have the presses in-house. Those are good for upto 20,000 or more copies. We are also seeing that the drum cost is coming down. The quality of the capstan has gone up. I think there is no reason to discuss as we have already been discussing this topic, whether capstan or drum imagesetter is actually better but what we are seeing is that the market is becoming much more competitive which is to your benefit because the price of a drum-based imagesetter internal drum is approximately the same as a high end capstan device. The technol- ogy trend is that the spinners in internal drum imagesetters are becoming much much faster and of course PCs today-if I buy a PC today, I know that I am going to be cheated because tomorrow there’s going to be a PC double as fast at half the price. But that’s actually good because we men- tioned before, this is an upgrade path. Again, speed is comparative-what speed you need, what 


“A polyester plate is the plate that most imagesetters can run and they expose directly on the polyester plate instead of going to films and plates we take the polyester plate and put it directly onto the press. Now this is very good answer to the Indigos, etc., if customers already have the presses in-house. Those are good for upto 20,000 or more copies “


RIP platform do you need-if I can tell you that, today with the existing equipment, on a 200 MHz based PC you can do a 100 line ruling, a 750 mil- limetre x 550 millimetre full flat, 100% graphics and you can do between 7 and 8 separations an hour in that RIP-based on a PC platform. I think. there’s not really any discussion of why should you go to a Sun Sparc or why should you go to an Sil- icon Graphics platform if you can achieve that speed, because the real question is, what about the rest of your installation? What about your network? What about your workflow? Can that actually cope with that kind of speed?


So we have touched it a little bit earlier to- day, but what is coming-and some of these things are already here from some vendors in the mar- ket- and that is that you can have in-RIP trap- ping, it means on the platform that is driving your RIP, inside the RIP you can have trapping, you already have it with some vendors. Some very high price performance RIPs, some very high-speed imagesetters, more and more focus on polyester, and there is additional software like imposition- ing that we have also seen, being driven into the RIP itself. So a RIP is much more than just a ren- dering process for outputting your data through the laser and the imagesetter.


Actions. If you’re looking at what is it that you are producing today, I am sure that if you go through your books you’ll see that your print run has become much smaller, on the other hand, you’re probably doing a lot more diversified jobs. People are getting more and more free time, people are reading more and more different magazines- they want to read golf magazines, sports car magazines, etc. So you see lot of magazines coming out, all higher and higher quality and colour but in small runs. You need a quick turn- around and you need the tools to get all that. It also means in general, what we have seen at least that there are smaller margins on the traditional repro services and again it is a benefit to have one-stop shopping for the customer or in sourc ing. I think that the product changes are much 


“If you’re scanning in the colour target, there are various ways of actually making the scanner, see the colour as it is defined in the CIE colour space. The look-up table will then-when the scanner is scanning red colour like this-build a profile for it and whenever it is then measured in LCH, it will come out with defined colour values for that colour.”


faster than the technology in general. There could be many reasons for that and again I would like to believe that a lot of the product development that is going on today, is because of user require- ments. There are products being put on the mar ket in prepress that has a very short time of life and are being re-engineered and being devel- oped into a product that most users are actually coming back and asking for. I ended off my pre sentation before by showing this (slide) and we went very quickly through it. Again this is de scribing, more or less the input workflow and the whole matter is really, the flexibility of the system. Again, the freedom to choose your plat- form-Macintosh or PC, software should be able to run on both platforms. You should be able to have previews on all the different workstations, one preview can be distributed to all of the work- stations. So your workflow is becoming much much easier. You could also visualise that you have a big magazine or newspaper publication where you have different departments-you have a sports department, you have a fashion depart ment, you have an editorial department- and you can actually scan-in a batch of images and from each department you can pick up those images that you would like to, on your desk, crop and set up exactly to your liking and over the network, just transfer it back to the server and the scan would be done and linked with the low res that you have already used for your QuarkX- press package. The colour concept, the device characterisation and calibration is not something that we should go into, but we mentioned be fore that there is a standard of colours. It means that it is possible today by an IT8 target to make a scanner, a colourmetric device, what is being done is a set of look-up tables is being built- whether this is a UMAX scanner, or a Scanview scanner or a Linotype-Hell. If you’re scanning in the colour target, there are various ways of actually making the scanner, see the colour as it is defined in the CIE colour space. The look-up table will then-when the scanner is scanning red colour like this-build a profile for it and whenever it is then measured in LCH, it will come out with defined colour values for that colour. I think we can take questions a little bit later re garding this… What it should give the whole idea of all this-the colour concepts and colour computers-is not to try to make things more difficult, quite contrary. What it should give is that if we’re working with a defined colour with- in a confined colour space, it should be easier for the end user to actually use the various soft- ware packages, because what it gives, first of all is, that if I do a scan, which is IT8 calibrated and take it into another colour package like Photo- shop, or Agfa’s Fototune or whatever, we know that the colours I have scanned on my calibrated scanner, would be represented correctly in oth- er software products that are also working with the IT8 calibration. So there’s conformity, con- sistency from one software package to the other. Furthermore, we are working with LCH values rather than CMYK, because some of us, we would understand how to read values let’s say, on skin- tones. However, even trained operators have prob lems because skintones all around the world are not the same. So they might also make mistakes in skintone scanning in CMYK values. The rea- son for the LCH is that it is a linear model and it actually is much easier for novices to get a read- ing of the various values of colours on the com- puter monitor. Now one big step forward is ac- tually going to be made, I believe it should be ready in January, and that is, we can build any ICC conformance profile. A lot of people are taking about ICC profiles and I think really it is going to be major step in the right direction in a sense that we are going to be better in control- ling colours all the way through the whole pro- cess. So on the back end, the output server-this is pretty much what is possible already, you have your workstations where you are placing your documents, your Quark or page layout program, you’re sending it over to a Windows NT cluster, you can do in-RIP back strapping, you can do OPI, colour separation, which is going to be in- credibly important again, when the ICC profile is going to be made. Again as we heard before, it is probably going to be much more viable to scan an RGB or an CIE-LAB because conversion from RGB to CMYK might take place in the RIP itself, not at the scanner side. And of course, you can also use third party OPI servers and further more, many vendors are offering something which is a multi-RIP- the first available RIP will do the job that is coming from the page layout station. Just to finish off, by giving a broader scheme


of events or the whole workflow, we can look at this slide, which is looking a little bit complicat- ed maybe, but we can take it from the very left hand upper comer here, and it could be that somebody’s actually calling or sending a job over. We’re doing it on the Mac, Windows, UNIX, it doesn’t matter the hardware platform is not the most important thing. You can receive the jobs, you can do the colour correction, you can actu ally ‘do jobs. ‘Do jobs’ means put them in for scanning. You have a colour UGS operator in our world with our software. We are talking about automatic software as UGS, means User Guid ance System. It is a software that allows you to do all scanning automatically where the operator would not have to do the corrections himself, but allows software based on a set of specified specifications to do the corrections. On this mon- itor here you do the preview, you can do proper size, do all the work you have to do and the whole idea of the concept is that you have a Scan- flow server in the middle which is a fault toler- ant data-base and in the future it should be pos sible to build-in a lot of other applications in the server such as accounting, costing and all sorts of different things, data-base applications are nor mally used for. In principle what we are trying to get at is that there is one way to increase work- flow and to increase productivity and that is to do processes in parallel. The more processes you can actually do at the same time, the faster you can get the job done and that is a very short way of trying to describe what we are actually aiming at. Like with the Scanflow service system, that I was also describing before, if you can do a scan at the same time as you are probing and setting up, at the same time as you’re doing a preview etc., of course, the final time will be much short- er than if you’re doing it in serial, have to mount a very large drum with a lot of images, have to wait for the scanner to finish all the jobs before you can do other jobs. That’s the main idea of the data or the lanmate as we call it. I know this is a lot of information and I hope at least some of you have got some ideas of where we are go- ing and where we are trying to see the trends in the market and I would be very thankful, if you have any questions, and we can discuss that lat- er. Thank you very much.


Burjor Poonawala introduces Sunil Khullar

Thank you very much Niels. That was a lovely pre- sentation. Niels talked at length on workflow and I am sure, a lot of you might have questions and we can allow a few questions after Sunil Khullar has made his presentation. I don’t know if all of you need an introduction to Sunil Khullar, he is a very well known personality in our country. But just in case, somebody, some people are not very familiar with his background, I will briefly intro- duce him.


He is an engineering graduate from IIT, Bom- bay. He did his post-graduation in Visual Commu- nication. He worked at Godrej on the Indian lan- guage terminal development. He has worked at VSS Bombay, to develop the first Indian script PC based WYSIWYG pagination on Ventura. He start- ed ‘Digital Studio’ for software development, font digitisation, and consultancy. He has developed ‘vp-fx’, a PostScript effects plug-in for Ventura. He was involved in system integration for colour pre- press and newspaper systems-I presume that’s consultancy and currently, he is Director of Sum- mit Data Products (Bombay) Pvt. Ltd. His areas of interest are cross-platform networking, scanners, PostScript, PostScript-based systems integration and consultancy. He is going to talk to us on the death of PostScript. It’s a very interesting subject. We’ve been talking about PostScript, and suddenly we have a speaker here to talk about its death and passing away. Let’s hear what he has to say.


Sunil Khullar on The Death of PostScript


Though this presentation is in digital format, I really cannot assure you better slides than what you saw in the analog format. Digital doesn’t al- ways mean better. In this brief presentation, we’ll be looking at topics like, first of all is it possible for such a powerful product to die off? And why would it die? And should it die what is the mean- ing to us, as users of PostScript, who can kill PostScript. I am sorry I am using words like death and kill, it’s purely in marketing terminology and not being a marketing person, I’d like to use those terms, and finally, how will Adobe react to all this?


So can it die? Well, how many of us remem- ber these programs-it could be WordStar in In- dia and it could be Word Perfect in other coun- tries, or for example, closer to publishing, who remembers Ventura or Ready Set Go? Maybe to- morrow, who will remember Quark the way it has 


“PostScript started as a device independent language, that was its USP. Today it is no more device independent. It has become so device dependent that if you take a file which was created for X-output device taken on Y-output device, chances are it will bomb. It’s a very flexible language, and as a result of that there is no structure. It is very complex-there are about more than 500 key words in the language.”


been developing as compared to PageMaker. Why would PostScript die? There are some inherent weaknesses of PostScript-we’ll go into that lat- er and we will see some new alternatives which are coming which are threatening PostScript and new media-here I am referring to the electron- ic media. I won’t be talking much about the elec- tronic media because I think in the next two days we’ll be hearing a lot from other presenters, but you will be able to take some hints from the (pre- sentation). By weakness I mean two things, one is limitation of features in PostScript and second- ly, problems of PostScript as a language.


Limitations of features


One of the primary limitations is that the PostScript allows only linear transforms. Rectan- gular things can be modified into a linear kind of transformation. But a simple, perspective kind of transformation is not possible in PostScript; it is possible in any small video-daily on MTV we see graphics where that is happening in real time, but in PostScript, that’s not possible.


Spatial screens, screens-we can have spot screens, we can have dot screens, we can also have screens which vary according to the space, like if you want wavy screens or circular screens, which we used to get in the older cameras. Gray screens are limited to only 256. 256 is a wide range for a band especially when you have high contrast vi- gnettes it’s no problem, but when you are display- ing low-contrast vignettes-in fact, at the bottom on my monitor, it shows the banding. When you have specially vignettes from say 40% gray to 45% gray over larger area, then you see those bands are visible. When you have 0% gray to 100% it’s not a limitation, but in smaller contrast, low con- trast areas it’s a limitation. There are no vignettes in PostScript-PostScript language doesn’t sup port any vignettes. To develop a vignette or put a vignette in an ellipse for instance, you have to define crude patches of gray level and create vi- gnettes and this is one of the reasons why RIP manufacturers even if they try to improve the qual ity of vignettes, there is a limitation, because ulti- mately vignette is not through a command, but through this kind of crude trick. But the biggest limitation is transparency. This is what a lot of us ers of PostScript across various fora have been crib bing about, that one object when it combines over a second object, with masks it can either be fully transparent or fully opaque. There is nothing like transparency where I would like one background merging into another, one object merging into another object. Transparency is a feature which is a serious limitation of PostScript.


And then, there are problems of language PostScript started as a device independent lan guage, that was its USP. Today it is no more de vice independent. It has become so device depen dent that if you take a file which was created for X-output device taken on Youtput device, chances are it will bomb. It’s a very flexible language, and as a result of that there is no structure. It is very complex-there are about more than 500 key words in the language. It becomes very difficult for a person writing programs to understand the language since there are so many key words. It is very fatal-from a one GB PostScript file or a one MB PostScript file, corruption of a single byte can cause the whole page to drop out or not print. In any other PDL language, you might have a small corruption, as resulting from one byte of corrup tion. But in PostScript one character or one byte can cause the whole job to not print. And no late stage editing -at the eleventh hour, if in a job you want to make a small typographic correction, that’s not possible. Again the biggest problem of the language, is its unpredictability. I don’t know how many millions of dollars must have been spent in output which was fine the last time we printed it, a small correction was made and now it’s total- ly different. The unpredictability actually comes because of all the above reasons.


So, should it die? I mean what does it mean to us? Number one, Niels just mentioned that development or products should be customer related or consumer related and it should be guid ed by our needs. So that is not happening and perhaps if PostScript is threatened, Adobe will be more customer-oriented. And the need for an industry standard-PostScript though it’s an in- dustry standard, but it is actually a proprietary tool of Adobe. Adobe develops it the way its feel like. So briefly, PostScript, like other products, is vul- nerable to change. We are going through changing times, new technologies, new media. A change, however painful, would be for the better.


Now the big question: who can kill PostScript? Word Perfect, Lotus 125 and dBase, these were killed by a very powerful company called Microsoft. But there are no Microsofts… or are there any Microsofts in publishing? Well, according to me, there are two Microsofts in publishing-one for the low end laser printer market and one for the imagesetter market.


First the low end. This company is the largest PostScript OEM, so a big customer for Ado- be, it is also the largest non-PostScript OEM, so also a big competitor to Adobe. This company is Hewlett Packard. HP is the biggest Adobe PostScript vendor, and I have been told that out of Adobe’s revenues that come from PostScript licenses. Adobe are the market, So it’s a big OEM for Adobe. But recently HP announced that its future products won’t have Adobe PostScript. We don’t know the details but this announcement was made a couple of months back and which means a big loss to Adobe. And simulta- neously there was another announcement that HP has tied up with a company called Xionics.


Now who is Xionics? In fact I had never heard about this company Xionics and it’s BIOS and it was the first company to come out with a hard- ware PostScript for HP Printers and at one time we had read stories about them coming out with a BIOS compatible for the Mac operating system, though it never developed. So it’s a very strong clone technology manufacturer and Xionics is to laser printers what Harlequin is to imagesetters. They have all the high end laser printers and high end colour output devices as their OEMs, some 50 OEMs. And then they have some better tech- nology for colour. They call it enhanced colour screening, display list technology, super multigrey. They are also open to developing PostScript. Many of the products have got a switch to have a stan- dard PostScript and enhanced PostScript. That was a brief introduction to Xionics HP also has its own page description language called PCL which has been, and which I think is a potential threat to PostScript. But what is the difference between the two? What’s so good about PostScript that today PCL can’t do? And HP’s new PDL, that is PCL6, can it change that? What is so great about PostScript? What is it that we cannot get, in the low end of the market, in laser printers or ink-jet printers? What does it mean to have PostScript on your laser printer or what’s additional that’s there 


“So with PCL6 coming, and becoming object-oriented, some competition can be expected. A second new feature of PCL6 is that it is highly compressible. It is said that what previously took about 1000 bytes to describe, let’s say a rectangle, now takes only 19 bytes. So it will be very compact, and together it means that there are possibilities that this can lead to a language which can pose a severe challenge to PostScript”


in PostScript, that’s not there other PDLs like PCL or SKP? Well the biggest is the way to define and modify screens and through screens we mean colour separations. And only if we can define the screens properly, the angles, the frequency, we’ll be able to get colour separations. So the biggest difference is screens. Second is encapsulation. PostScript is not only popular for its features but also popular as a way to transport graphics from one device to another device, from one platform to another platform. So EPS file-that’s what I mean by encapsulation-is a feature which can be limited only to PostScript environment. And then there are other smaller features like linear transformations, which are not there in PCL. But the new PCL is said to be object-oriented. Now, object-oriented was another one of the USP’s of PostScript when it started. It was an object-oriented device independent language. So with PCL6.com- ing, and becoming object-oriented, some competition can be expected. A second new feature of PCL6 is that it is highly compressible. It is said that what previously took about 1000 bytes to describe. let’s say a rectangle, now takes only 19 bytes. So it will be very compact, and together it means that there are possibilities that this can lead to a lan- which can pose a severe challenge to PostScript. That was the low end laser printer, ink- jet printers, those kind of devices


What about high end? In high end we have a different company which will challenge PostScript and they already have an alternative ready for PostScript. Secondly, they have the back- ing of the industry and this challenge is nothing but Adobe’s new PDF-Acrobat PDF. So what is the PDF threat? What are the PDF advantages? Today what is it again that PostScript can do that PDF can’t? What is the difference between the two? And how can Acrobat 3.0, the new PDF format, change that. The biggest difference between PostScript and PDF is that PDF is page indepen- dent. You have a 10-page job, I can just remove the 8th page and print it independently, it will  print. In the PostScript environment, you have to print all the things linearly because one page can affect another. It is much more compressed than PostScript, it’s reliable and predictable. It doesn’t require any preflighting. An Acrobat reader which is available free of cost in all the major platforms can preview PDF and editing at last stages is pos sible. What is it then PostScript can do that PDF can’t. Again, screens-currently there are no well it prints in screens, but we cannot define screens. Encapsulation and RIPing. If you have a PostScript file, you can output it. A PDF file, you cannot output it, you have to go through a secondary program like PostScript reader to cre ate again a PostScript file to output. And what about Acrobat 3.0? Acrobat 3.0 is a set of new tech- nologies which will incorporate the new PDF for- mat. It will be colour capable, you’ll be able to define screens in that, it will be encapsulable, so that takes care of the first two things… and it will be directly RIPable with some new technologies which I’ll be talking about, and it will preserve… totally, it will be workflow kind of thing; so all OPIs, commands and everything, which are currently removed when you convert to PDF, will be pre- served. So the worst scenario is, that in the low end, HP might improve PCL to eliminate PostScript from the low end of the market, or HP- Xionics might develop a new standard which they have hinted at. PDF can become a standard for prepress and Adobe at many fora has agreed to it that the future it sees for PDF, will make PostScript redundant.


That was one part of the story. Now how will Adobe fight back? How does Adobe see all these developments? Adobe is developing a number of new technologies which I think are aimed at fight- ing these eventualities. One is Printgear, the Am- ber technology which is Acrobat 3.0, the Supra and the recently announced PostScript Level 3. We will briefly go into these.


Printgear is aimed at the low-end of the mar- ket, essentially aimed at the PCI. market. It will have features-these features are not there in PCL so these are plus points for Printgear. N-up print- ing for labels and repetitive kinds of things, it will print water marks, you can blow up or mix pag- es, print posters, booklets, duplex printing, it will have more built-in-fonts-instead of 35, 78 stan- dard fonts, some other Adobe technologies, and it will have PostScript PCL emulation, so that ex- isting programs which don’t understand Printgear format will still print. And it’s already being im


So again Supra is ready for implementation. And that’s what the whole technology will do. It will take PDF files directly, it will take PostScript files, divide, a PostScript file which is, say 10 pages, into individual pages and then those individual pages will be available for editing, archiving, and doing all sort of things.”


plemented. The first product NEC Super Script laser printer which costs about 500 dollars, has been recently announced, so this technology is ready.


What about Amber? I just mentioned this, Acrobat 3.0 with screens and colour separations will be a part of Amber technology. Amber is also aimed at the Internet and the web market. So it will have technology that allows PDF files to be directly browseable from Netscape or any other browser. It will support plug-in technology, so you will be able to add small features to PDF and it will have EPS links for dynamic things like but- tons. Implementation is still going on, right now only demo versions are available. We still have to see Amber.


Supra Supra is what will make the new PDF RIPable. So with Supra you will able to take PDF and output to a Supra RIP directly, you don’t have to go through an Acrobat reader. It will have PDF management built-in, you’ll be able to view and edit and archive PDF files. Personalisation of graphic data, this I think, is aimed at direct to press market. It will have features like modularity and scalability, so vendors will be able to add small features without modifying the entire Supra, it will have some post press controls, again aimed at the direct to press market and it will be a true paral- lel processing environment where within one RIP you will able to output two multiple output devices, you’ll be able to print, say 10 pages of a single job to 10 output devices. So aimed at again digital presses and high speed small run presses. And its first implementation is also ready, this is from IBM Infoprint, it’s a 600 dpi high speed laser printer kind of device. So again Supra is ready for imple mentation. And that’s what the whole technolo gy will do. It will take PDF files directly, it will take PostScript files, divide, a PostScript file which is, say 10 pages, into individual pages and then those individual pages will be available for editing, ar- chiving, and doing all sort of things. And lastly the newly announced PostScript


Level 3. Not much is known about PostScript Level 

3, but according to Adobe, it will have four fea- tures-the first feature is enhanced image tech- nology. It will make things faster, easier, and im- plement optimal image quality which is very im- portant. We get a lot of PostScript errors as limit errors, whereas a job which is printed on one de- vice, doesn’t print on another device because of a limit check error. With optimal quality, they as- sure us, it won’t happen; it might print slightly cruder, not to be best of the output device abil- ity, but the job will go through, you won’t get an error. Again aimed at the Internet market and electronic media market, it will directly take not only PDF files, but the HTML files from the web.


Again, aimed at the new networking environ- ment market, you’ll be able to control PostScript Level 3 devices from the web. So basically the con- trol panel of the printer or a PostScript printer will look like a web page and you’ll be able to modify and query the status of the printer from the web page. And lastly, what they are talking about, is “Planet Ready” printing. I really don’t know what this means. But according to them it will mean better drivers and international font support among other things. This next level of PostScript is not really ready, around the middle of next year is when we’ll see this technology.


So the best scenario for Adobe’s PostScript is that Printgear might redefine new standards for low end printing and even pose a challenge to HP. PDF will become a standard for electronic media which there is a great chance of it becoming and PostScript might remain a choice for us prepress people. But I personally feel HP-PCL is very strong in laser printers, it will be very difficult to dislodge it. The HP-Xionics product is still to be seen. Print- gear is also too new to be commented on. PDF has definite advantages though it is not very pop- ular in web right now but it will become a very strong option on the web and for electronic work- flow it will become very popular. But still, till en- capsulation in PDF is defined and we are confi- dent about it, people will still use EPS as a mode of transportation of graphic data. And Supra and Amber will place PDF at par with PostScript. So PDF will become an alternative to PostScript. Ado- be is very clear about that. But whether it will make PostScript redundant is to be seen. Level 3 is still to be seen and lastly, we are sure what our needs are, and if a new technology comes in and replaces the old technology, one thing we’ll be sure of- that new technology is better than the old one. So if PostScript is dead….




Burjor Poonawala Thank you Sunil, I don’t know if you I’ve made out a case for the death of PostScript or its continuity. Both the sides of the argument were more or less equal, in my opin- ion anyway… We’ve completed the sessions pre- cisely on time, that is I think 5.30. The Expo is open and it’ll be open till 7 o’clock, so please go and have a look at what the vendors have to show you. I am sure you’ll see a lot and learn a lot.


Naresh Khanna Twelve years ago, when PostScript first came, I think it really played a very big role in the implementation of non-Roman scripts on computers and side by side, inexpensive person- al computers came, and now 12 years later, Ado- be is talking about Planet Ready, and two byte characters. According to their own estimates, these products would not be out even in mid-1997, but end 1997, although the technology would be ready next year. Do you think that we’ve done as much as we should have with PostScript even for non-Roman languages in terms of development al- though the implementation of non-Roman lan- guages and scripts became much easier with PostScript? With PostScript and software such as Ikarus, and Fontographer and Publishers Type Foundary font implementation became much more tractable than when it used to cost 50,000 Rupees or 1 lakh or 500,000 Rupees to try out one type-face, and that also by sending it to England to Monotype or Linotype or somebody to have it digitized in the old technology. We suddenly had the hardware, software and end user environment where we could design our own scripts and be- gin the work of offering a variety of script type- faces. Do you think Planet Ready is a serious fea- ture or product from Adobe or is this one reason we should remain with PostScript Level 2 and may- be we should develop our commitment to it with script add ons just as RIP manufacturers are de- veloping alternatives like Delta technologies and various other workarounds to limitations of


“… as we are talking about device independent scanning, we are talking about imagesetters and scanners on a cross- connectivity platform, we believe that the user will benefit quite a lot by having a scanner that’s giving a consistent output, it doesn’t matter whether it’s let’s say from one vendor or another and he can always rely on the calibration he has done inside the RIP and with the imagesetter for the RGB to the CMYK conversion. That will be consistent because input colour from the scanner itself is always defined by the IT8 standard.”


PostScript? Should we not for the purpose of non- Roman languages pay some attention to the issues that PostScript does not address here, and do some R&D on our own? Maybe we should kill PostScript.


Sunil Khullar Number one, as far as I know, Plan- et Ready won’t introduce any new technology as far as fonts go. They are only talking about local customization, which means today HP when it’s selling its printer or for example, some laser print- er manufacturer when he’s selling a CPSI or in- ternal RIP and wants to, instead of giving 35 standard PostScript fonts, give 35 English Roman fonts and 2 Devanagiri fonts, it should be able to do it easily. That’s what they call ‘local customisation as I understand and of course today 2 byte is to- tally limited-very few people implement 2 byte and perhaps the licensing for 2 byte technologies is different. I think the licensing of 2 byte tech- nology will be easier. I am really not sure what new features they are going to introduce in Level 3 as far as font technology goes.


But your question was also addressed to us as have we taken advantage of PostScript seriously enough to develop fonts and exploit this technology. Yes and no… we have been serious as you yourself pointed out from the cost of fonts. To- day, fonts are available, though their quality may still be questioned. But a number of fonts are avail- able at down-to-earth prices which means some- one has paid some attention to it, though the is sues are there, of quality, of standardisation, of exchange of fonts. But we have paid attention to non-Roman scripts. I think more than other non- Roman scripts like Arabic or Hebrew or Cyrillic. Those are still very linear kinds of fonts. We have many more diverse languages and our fonts are much more difficult and within the limitation of 256 characters that Adobe PostScript allows us, and our requirement being larger than that, I think   surely a lot of attention has been paid to the best that can be done to Indian language fonts. As far as fonts are concerned they are still Type 1 fonts, I mean, no new font technology has been an nounced. Level 2 also as far as fonts were con cerned, remained the same; nothing new was an nounced and fortunately today we are not limit ed to Type 1 fonts. Truetype is an alternative though again, I think, quality is slightly an issue. TrueType is an alternative which Adobe is also tak ing seriously and more and more RIP manufac turers at least acknowledge the existence of Tru etype and they are included as directly RIP’able fonts


Naresh Khanna What about Unicode?


Sunil Khullar Unicode as defined by ISO and Mi- No, 2 byte technology has nothing to do with ISO 2 byte coding. It is just a technology which avoids the limitation of standard PostScript of 256 or 8 bit font code. That’s all. It’s a kind of a patch essentially meant for Chinese and Japa nese script requirements.


Yes, interesting things are happening on the Web front, I mean… In the print media, we can support complex technologies and people can pay attention, sort out issues of font clashes, we have lived in that environment. But the real challenge here today is the electronic media. You cannot ask millions of users around the world to have your kind of font, code or font type, so serious think ing is being given by various people to unify tech- nology for the Internet. I really don’t know what is happening. I am not the right person to talk about the electronic media. Maybe the future speaker, maybe Mr. Duggal, when he will be speak ing day after would be the right person to ask. Some initiative has been taken, how serious I re- ally don’t know.


Lakshmivenkataraman What is the idea of RGB to CMYK conversion being implemented in the RIP? When we define a device independent LAB value system and all, why should we make the conversion from RGB to CMYK in a RIP? RGB to CMYK.com version varies from software to software. It is just a mere look-up table which Photoshop imple ments in one fashion and ProPM Press, Monaco, and Kodak, everyone does in a different way with their own look up table method. If we implement the conversion in a RIP, then the standardisation will vary, will it not? What is the idea of putting  the RGB to CMYK conversion in the RIP?


Niels Ohrgaard It is quite true that there are some software packages in the market that are doing fantastic RGB to CMYK conversion and I might add that we have one of them, so, of course this is not just a drive to sell less of our software pack- ages to put the CMYK conversion on the RIP it- self. However, we believe that it is inevitable, that is the way it’s going to go, because as we are talk- ing about device independent scanning, we are talking about imagesetters and scanners on a cross- connectivity platform, we believe that the user will benefit quite a lot by having a scanner that’s giv- ing a consistent output, it doesn’t matter wheth- er it’s let’s say from one vendor or another and he can always rely on the calibration he has done inside the RIP and with the imagesetter for the RGB to the CMYK conversion. That will be con- sistent because input colour from the scanner it- self is always defined by the ITS standard.


So we believe that there are benefits in terms of workflow over the network as well, that since the RGB file is normally 30% smaller than a CMYK file, you’ll also save on the network itself while timing and transporting the files. Often if you’re build- ing an in-RIP OPI, you also have the high resolu- tion data on the RIP itself and therefore, there’ll be a further added incentive to actually do the RIPing inside the imagesetter.


Sunil Khullar I’ll add to that. One of the greatest advantages of having RGB to CMYK in a RIP is you can do colour separation from programs which cannot take CMYK data. For example, you have placed an RGB graphic in Microsoft Word which does not recognise the CMYK file, (and you can compose very decent kind of layouts in that using RGB graphics) and suppose you want to just output that, including the graphics as colour sep- arations, one of the ways is by using this RIP fea- ture. It is there in most of the new software RIPS and not only that, RGB to CMYK conversion and reverse conversion is also part of standard PostScript Level 2. For devices like laser printers, you have a colour laser printer, which is designed to take all kinds of data, RGB and CMYK, both. So if a colour laser printer or colour ink-jet gets RGB this is converted internally.

Indian Printer & Publication                                         April  1997

Proceedings of the third annual pre-press technology conference Part 4


Held at Bangalore on November 28, 29, and 30th 1996

Introduction of Veronique Sierens by Burjor Poonawala


Our first speaker today is a lady electronics engi- neer. This is the first time that I’ve seen a lady elec- tronics engineer and she must be pretty good because she’s been with Agfa since 1990. She has been with Agfa as a products sales and application trainer for EPS, that is, Electronic Prepress Sys- tems. As of January, 1997 she’s been responsible

 for EPS sales and applications’ support for Direct Export Group in Agfa. So I suppose India comes within her scope. And she’s been here for a few days, and I had the pleasure of meeting her in Bombay when she visited our workshop and I can assure you she is very knowledgeable and it will be a very interesting session.


Veronique Sierens on stochastic screening


Thank you very much Mr. Poonawala, Good morning everybody. I would like to start this first presentation about stochastic screening. Stochastic screening is a screening technology that has been introduced and commercialised for the printing industry almost four years ago. Since then we have learned a lot and most of our customers have been giving us a lot of feedback about this technology. That feedback is what I have included in my pre- sentation today.


So the content of the presentation will be explaining first of all a little bit the half toning principle, then go to the specific benefits about the stochastic screening, have a look at the impact on the production process and then look at some of the typical applications and one of those ap- plications will be Hi-Fi printing. Everybody knows the classic half toning principle. I do not have to re-explain that to you, you are the professionals. When we look at a conventional dot, that is the one in the middle that you see and we look at a digital conventional one, we see that it resembles it a lot and it has been built-up by regrouping la- ser dots that will be with imagesetters as we know it today in the market. This conventional technol- ogy, the digital one that we know, has a few prob- lems. Some of the problems that on a day-by-day basis you are struggling with and I am sure that you will confirm this to me, are the typical jumps that you can see in production skintones, for ex ample are very critical in that manner. When we doing process colour work, moiré is always a problem. we have a need to compro mise-we have to compromise between the dot shape and the line ruling and recorder resolution, and sometimes those decisions are not so easy to make. Research patterns are sometimes visual and very disturbing and then of course, there is the subject moiré and I don’t have to explain that to you as India is a major country where textiles are very important, which have to be handled with lot of care when printing those subjects. But, of course, the printing as it is today is of high quali ty also in India because you have certain workarounds for all these day-by-day problems and workarounds are there You can use eliptical dots to create less jumps in the tone reproduction. You can vary the angle of a specific colour in order to avoid moiré. You can increase the resolution on your imagesetter in order to create a better qual iny dot, increase the screen ruling to see less pat tems, less rosettes, or vary the screen ruling to avoid the subject moiré. These are all things you can do today. But we have a new solution for you and that new solution is the stochastic screening technology.


The way that stochastic screening is built-up, is what you can see here in this slide, instead of grouping together the laser dots like the conven- tional dot is built-up, we spread the dots in a spe cific area so we random the position of the elec tronic laser dots on film. That’s how we built-up the stochastic dot. If you look at a tone curve and you compare the way that the tones are varying with conventional screening, what you also called AM or amplitude screening, we see that we change the size of the dots, as we go along in tones. With stochastic screening, we never change the size of the dot. The basic dot size stays the same all over, what we do change is the amount of the dots so in a highlight area you will see less of those dots than in a darker area. That’s how we create the stochastic screening which is also known under the name of FM screening or frequency modu lated screening. This is another slide showing you the difference between the conventional and sto- chastic screening. When we evaluate the halftones, we look at specifics like image quality aspects, we look at patterns, artifacts, noise, grayness. That’s how we can judge if reproduction is good enough. We look at the tone rendering, we also look at process characteristics. We want to know, how we can implement it in the work-flow, what about the reproduction, the plate-making. Can we print it? Those are a couple of questions that people had few years ago. And of course, how do we imple- ment it in the digital workflow. Do we need spe cific computers? Do we need more hardware? Do we need lot of memory? Is it compatible with ex- isting workflows?


First of all, let me explain to you that the heart of this screening technology, its all around the micro dot size. With stochastic screening we don’t talk anymore about the line ruling like we do with conventional screening. When we talk about the fineness of the stochastic screen, we talk about the size of the dots and there are multiple-size dots available. We can have a 15 micro metre dot, that will be used for showcase colour, we can have a 20 micro metre, which will be more for commer cial colour printing and even a 30 micro metre dot, which can be used in newsprint and flexo. The size of those dots we create by addressing dif ferent resolutions on the imagesetter. And of course, it has to be a balance between image qual- ity, text line art definition, press capabili ties and ease of use.


Now, when I talk to you in micro metres that doesn’t ring a bell, but let’s compare the size of the stochastic dots with the size of the conventional dots: When we look at a 50% dot at 150 lines per inch, that is about 135 micro metres, that’s the big dot that you see on the slide. The crystal ras ter dots are the yellow ones, so you can see that a 21 micro metre stochastic dot size is smaller than a 3% conventional dot at 150 lines per inch. So that means that in order to keep that on the plate we have to be very careful and somebody who can- not keep a conventional 150 line per inch dot on plate will have difficulties in implementing this technology.


We modulate the grayness, and this is how it looks enlarged with stochastic screening. Now some of the claims of our customers are the fol lowing: they say that with the stochastic screening technology you really get this continuous tone appearance. It really looks like the original pho- to-realistic effect. There are no artifacts any more, which were classical for conventional screening. There is an enhanced detail reproduction, that is of course linked to the small dot size and en- hanced tone reproduction.


Even the productivity in the prepress environ- ment can be improved by implementing this tech-


“When we look at the tone reproduction, we do not have to make a trade-off any more between the line ruling, the resolution and numbers of gray levels. We have the full tone range available at all times,””


nology and also on the press, life becomes much easier.


When we talk about photo-realistic image quality, we actually mean that with the stochastic screening technology you get a quality that is com- parable to gravure-like continuous tone. We could at least state that with the stochastic screening tech- nology you get gravure quality on an offset press. There is absolutely no moiré, also when you want to print very difficult colours with this technolo gy, it is not an issue any more and there are no colour shifts in the neutrals, no subject moiré and of course, it opens the door for hi-fi colour print- ing. In stochastic screening, we do not speak about line ruling and when we talk about four colour printing, we do not speak about angles any more for the colours. There are no angles, everything is random. So it is much more easier to print with more than four colours and that is what hi-fi print- ing is about. It’s the printing process with more than four process colours. And then, of course, there are those typical rosettes that you can see when printing conventional four colour process Those are becoming invisible. Also spot colours are rendered much better. The following slides will show you some comparisons between conven- tional screens and stochastic screens. On the slide you can really see that there are no screen pat- terns any more visible on the stochastic one, that’s on the right hand side. There is no moiré with stochastic screening, no visible rosettes with sto- chastic screening. And also line art reproduction, text reproduction, it’s much sharper with stochastic screening. During tea break, you will have the possibility to look at some of the customer sam- ples that I have brought with me and one of those samples is a print of jewellery and watches. Look at the watches and you will really see that every- thing that is on the watch is very clear. That has to do with line art reproduction sharpness of sto- chastic screening. So we have an extremely sharp detailed rendering potential with stochastic screen- ing. Actually with stochastic screening, you can get the same quality as you do with a very fine con- ventional line screening. You will get quality like a 300 lines per inch conventional screen. You can change the input scanning resolution, when go ing to the stochastic screening technique, you don’t need the two-time over-sampling any more. You can keep it the same, if you wish you can scan with lower resolution, and if you wish you can also reduce the unsharp masking. But as I say you can still continue to work the same as you have been till today. If you do wish to reduce the oversam pling at the scanner, the immediate result will be that your file size is reduced and that means of course that transfer of your files over the net will be much faster. As I said, we will give you a quali ty that is comparable with 300 lines per inch con ventional screen but at a lower recorder resolu tion. That means you will be able to obtain the same quality at a faster speed. This slide shows you a comparison between a detailed rendering on stochastic and conventional.


When we look at the tone reproduction, we do not have to make any trade-off any more be tween the line ruling, the resolution and numbers of gray levels. We have the full tone range avail able at all times. Also the transitions in the tone rendering are much more smooth. We don’t have these dot corner link-up effects any more and we don’t have any plugging in shadow areas. Also in printing we can obtain much more brilliant co- lours, much more saturated colours and that of course gives us less paper show-through.


This slide shows you the tone rendering in the pink, on the right you will see the tone ren- dering area for the conventional screening. The green area shows you the extra tone rendering that you can obtain with this technology.


When we look at the prepress requirements, as I said you don’t need to change anything in specific. We can work with existing workflows, we can work with existing unsharp masking, gray balance and under-colour removal settings. It may or it may not be compatible with standard film and chemistry. Density is going to be of major impor tance for film. Density of 4 is going to be crucial and with our materials we even recommend to go higher than 4. It does require a change in dot gain compensation. When you print a crystal raster, you will automatically see that the dot gain on the press is much higher than with a conventional screen. So this has to be compensated for and we will do this in the prepress area. As I say, the dots are very small, so some care will have to be taken in the stage of plate exposure and process. And also proofing guidelines have to be adapted. The in- cremental dot gain compared to conventional printing, will be different for negative and posi ave printing. In the negative printing, we see an erage of 8 to 20% more dot gain while in the positive printing process, an average of 1 to 6% more dot gain. This will depend from press to press So on a Heidelberg machine or on a Ko mori machine, you can see different extra dot gains


There is a combination of recorder, plate and press gain. There we have to take the same care as we do when we want to print fine convention- al screens. And if you want to create any differ- ences in print quality, we will look at smaller dot sizes. So somebody who starts with a 30 micro metre stochastic dot and who wants to have even finer quality will have to go to 14 or 21 micro metre dot. This curve gives you an idea of the extra dot gain. So the blue line is the dot gain in printing with conventional screening and the red line gives you the dot gain with stochastic screening, both for negative and positive printing.


How do we compensate for this dot gain? We do it for you. Everything is implemented in this screening technology. It can be performed by the user and a calibration can be performed during the scanning process or during the recorder pro- cess. But our experience is that every calibration gives a loss of gray levels, so to guarantee that you always have at all times the full range of gray lew els, we implement this correction for you in this screening technology. So the Agfa crystal raster which is Agfa’s stochastic screening technology. has internal built-in pre-compensation for the press dot gain.


When we look at prepress guidelines, as I said, of course we need to have proper care for film and plate preparation. The density of the film will be very important. We have to work in a dust-free environment and of course, we have to work with uniform illumination, good vacuum and no scat ter faults. The same goes for plates. Dot etching will not be possible any more and contacting is not impossible but is not recommended.


What are the factors for success with this tech- nology? Of course, you need to have a high qual ity imagesetter with high precision and repeatabil ity and an imagesetter which has a good, controlled imaging density. Then of course, there is the sto- chastic algorithm and you need to have properly controlled prepress environments. This technol Ogy is compatible with existing offset procedures. You can work with standard plates, inks and press es. There is no need for anything specific. You can even work with waterless presses. Of course, when I say standard plates, it is important that the plate can keep that small dot. So the resolution of the printing plate is important. Paper: you can print this technology on various qualities of paper, high quality paper, but also on low quality paper, re- cycled paper and newspaper.


The feedback that we get from the printers, they say that the make ready is much faster. When starting off the press, the press comes much fast- er in colour and there is less waste. Also the ink- water balance is much easier to keep on level.


Press registration is easier to handle with sto- chastic screening technology. With difficult images, even being a little bit out of registration, you see no moiré and patterns coming up. You don’t see any rainbow effects in the neutral areas, Printers can print with their normal ink densities, but if they wish to obtain this high saturated glossy cov er effect, they can print with a higher ink-densi- ty. And there is less paper dusting, less blanket cleaning necessary. The stochastic screening tech- nology has been used for a couple of years now in commercial printing. And we have seen from our customers, it has been used for brochures, annual reports, all kinds of collateral and adver- tisements, products and mail order catalogues, posters, books, greeting cards, all kinds of maga zines, newspapers, even the yellow pages. We have even had some success in flexo printing and some customers even use it in silk screen printing. Of course, in silk screen printing, the dots are too small to handle, so customers who use this tech- nology today will blow up the dots with camera before they go to the silk screen printing. From all these applications, I have brought samples with me and during tea break, feel free to come and look at the samples and see for yourself the re- sults of this technology.


Burjor Poonawala introduces TN Rajan


Mr. Rajan is also going to speak to you on the same subject but more from the user point of view. And I wish to introduce Mr. Rajan to you, I’ll just go through his bio-data. Mr. Rajan is an ex-service- man from the Indian Air Force. After meritori- ous service in the Air Force he migrated to the United States. Now he possesses abundant expe- rience of two decades in the world of printing in the USA after graduating in Printing Technolo- gy from there. A much coveted graduation in pho- tography from the New York Institute of Photog- raphy is another qualification. He is also a diplo- ma holder in Marketing Management from the Institute of Marketing Management. Armed with all these Mr T.M. Rajan returned to India and opened Process Colour in 1989, thus becoming the pioneer in colour scanning in the state of Kerala. Mr. Rajan is a member of the Governing Council of the All India Federation of Master Print- ers, Vice President Kerala Master Printers Association, and Chairman of Academy of Printing Technology, Kerala.


TN Rajan on FM screening


Good morning, friends. Let me tell you in a few seconds about process colour. We are based in Cochin. We are the first people who have these facilities of FM screening in India. A few years back- I think it was somewhere in 1993 in Bir- mingham, England at the Ipex exhibition, I was there. Agfa introduced their Crystal Raster screen- ing there. It was a big hit. It was a very fascinating print quality at that time and when I saw that it was very good, I decided then that I should have that facility in Process Colour. After three years when I was ready to buy an imagesetter, I was look- ing for different brands. Finally I decided to buy the machine from ScanGraphics along with High Fidelity screening. High Fidelity screening is a brand name for Scan Graphics’ F.M. screening. We tested this screening facility.


As you know, Kerala is not a big place for the printing industry, it is a small state, we don’t have sophisticated printing equipments like in Bombay, Bangalore or other places and most of the print- ers have only single colour Indian made presses. Only Malayala Manorama, or Mathrabhumi, those publishers have better facilities, otherwise there are hardly any facilities at all. This FM screening… people were afraid to touch it. When I said, why don’t you try this. No, no, sir, it is not possible, our machine is not ready for this. It’s not made for this. Finally one fellow, one small printer, a single colour pressman came forward to try this. His name is Mr. Ravindran, of Modern Graphics in Cochin. He tried, the first print he has made in his press is this one, I will be passing this to you. You might have seen this, it came in last year in Coramandel, as an advertisement for Process Co- lour. This image is taken from a Kodak Photo C.D. That same image is reproduced in conventional screen. You have seen Bombay Printers’ magazine last issue and Coramandel last issue. That same picture has come out in Coramandel but not on FM screen, in conventional screen. I will pass it over, you can see this and compare the quality dif ference in details. You don’t need any sophisticated equipment for this. Our regular press is O.K


What is High-Fidelity screening? High fidel- ity screening is the brand name of ScanGraphics frequency modulated screening. What are the differences between High Fidelity screening and other Frequency Modulated screening processes? High Fidelity (Hi-Fi) screening encompasses


a number of processes which bring considerable improvement to the quality of frequency modu- lated screening. Many critics of frequency modu lated screening justifiably complained that smooth surfaces, for example, metallic surfaces, porcelain, the sky, etc. very often appear too rough, jagged, patchy or cloudy. This is because of the calcula tion procedure of the frequency modulation pro- cess, which either distributes the screen dots in a purely random fashion even where the screening pattern can be manually corrected. There is no possibility to improve the dot distribution between the bordering details with similar tonal values. High Fidelity screening utilises a calculation pro- cedure especially developed by ScanGraphics which calculates the spaces between the screen dot representing the tonal values, at an extremely high rate of precision, thus allowing smooth surfaces to really appear smooth and then changes the position of the dots at a precise degree of random distribution.


The second major advantage of Hi-fi screen- ing is, it can immediately adapt to the requirement of plate exposure. Hi-fi screening is the only screening process which does not calculate for printing output but linearises instead. Possible screening frequencies i.e. for offset, newspaper offset, standard offset, high-end offset, and ultra high fidelity for the best quality of art paper.


Newspaper offset screening frequency approxi- mately 160 pixels per centimetre that is approxi- mately 200 dpi. Length of screen dot adjusts ap- proximately 62 microns. This is recommended for print-runs under high waste condition, as for ex- ample, newspaper rotary and very poor quality paper.


Standard offset For commonly used art paper. Screening frequency approximately 200 pixels that is around 600 dpi. Length of screen dots will be approximately 40 microns. This can be used for all offset applications.


Three quarter tones and shadows remain open even with poor paper quality and over inking. Low dot gains. There is dot gain but if you scan and prepare in advance you can control that in comparision to the conventional screens. That’s why said low dot gain. 35


High End screening frequency approximately 320 pixels that is 800 dpi length of screen dots approx- imately 30 microns, print-runs should be carried out only on high quality printing paper. We tried this. We were not too successful in this. You can print it, but it is not that good. Maybe because of our press conditions.


Ultra high frequency approximately 480 pixels, length of screen dots approximately 20 microns. This demands very high quality plate processing. And also very good quality plates are required, I don’t know if we have that kind of plate available here, and we have never tried this.


Film materials high fidelity films may not be en- larged or reduced with a camera. Intermediate copies should not be made. Film should not be bent, in duplicating or copying, we come across a problem, you will not be able to reproduce the dots exactly the same way so that is why it is said, not to be duplicated, but we can do it, you will lose some details in duplication. We have tried duplication but it was not that successful.


Plates Any good quality presensitised plate is rec- ommended. Hi-fi images require great care and cleanliness during plate exposure. While proof- ing, it is recommended to use the offset proofing method. Now I think Cromalin is also available for FM screening.


Printing ajustment of inks is much easier for Hi- fi screening as there is much more room for play. Inking is deeper so, therefore printing can be car- ried out with less colour density. No problem with colour registration. Very stable run-ons. More play during inking. Better distribution of ink and wa ter. Three quarter tones and shadows remain open even with poor paper quality and over inking. Low dot gains. There is dot gain but if you scan and prepare in advance you can control that in com- parision to the conventional screens. That’s why I said low dot gain.


This is a enlarged view of High Fidelity screen- ing. Hi-fi screening dots have the same size and differences in tonal values are represented by the distance between the dots themselves. The screen dot size can be exactly defined for all dots and adjusted to give the best possible result – as small as possible to show the maximum details and as large as possible to ensure the dots are reproduced correctly during printing. Tonal values are not subject to limitations as both positive and nega- tive screen dots can be as far apart as required. The screen dots which are only slightly larger than the smallest printable dots can be printed over the entire tonal range. The random distribution of screen dots reliably prevents the formation of moiré when printing any number of colours and also inhibits interference between structure con- tained within the image only. Only those moirés which have already occurred during scanning can be prevented. The moiré can be corrected by se- lecting a different scanning resolution.


Print media The Hi-fi screening programme of fers the facility to adjust screen frequency to meet a broad range of printing requirement, that is, dif- ferent kind of paper. Printing is relatively simple, once the plate has been successfully exposed and processed.


I have some samples which were recently printed with FM screening. Those people who are interested, you can take it and see that, especially in re-screening. This is a picture which was print- ed before, somewhere in Sivakasi. We re-screened it in FM, there is no moiré at all. You see this is just like a continuous tone printing. So I will pass it, there are few more left here if you are inter- ested, you can take it. Also I will pass this film. You can see the screen dots.




Sujit Patwardhan Are there any differences re- quired when you print with F.M. screenings, when you check the densities with a densitometer. The ordinary densitometer, can it be used in the same way as we use it today?


TN Rajan Yes, you can use it in the same way…


Veronique Sierens It’s true that we have seen that with very cheap densitometers, which have very small apertures, the readings are less exact than the high quality densitometers. So we may expe- rience different readings with different quality den- sitometers.


Sunil Khullar Mr. Rajan, How much of FM screen- ing are you doing in the normal course of busi- ness per month?


TM Rajan Not too much. There are few printers in Kerala, they use it, especially they use it where they have to reproduce from printed originals, some of the pictures where you notice a lot of moiré There they and use FM screening. This also can be combined with con- ventional screening. Both can be combined in the one film itself.


Sunil Khullar That’s true, but are your custom- ers coming to you and demanding now that they will use it?


TN Rajan Yes, They are not in large quantity, but there are people coming.


Burjor Poonawala One question I might address to the both the speakers is that, inspite of all these advantages in FM screening, why is it that we don’t see much of it in commercial printing from the publications we see from all over the world? I don’t think we have seen any publications printed with FM screening, even if they are high-end printing and high quality magazines or books or art books, there might be some, perhaps, but is there any reason, why it hasn’t caught on the way we would have imagined?


Veronique Sierens First of all I think that when this technology was introduced almost four years ago, a lot of people got scared, especially when they saw or heard about the way to handle the plate- making. That was the most critical issue at that time and lot of people got scared. Also I think that we have seen different suppliers coming into the market with this technology and the support that was given from the suppliers has been different, we in Agfa, we always have found it very impor- tant to give the necessary support when imple- menting this technology. That’s why we have been successful. We really take the customers by their hand and we guide them all the way through the printing. It is also true that this technology has not come for free. That is another reason why it may not have taken off, and also most of our custom- ers today, who are using this technology, they use it for specific niche of their applications; they don’t use for everything and, also if possible, they try un charge extra. So I think these are a couple of reasons why we have not seen this technology used on a day-by-day basis for all printing publications.


TN Rajan Commercially, one of the magazines we know, Vanitha, of the Malayala Manorama publish ing group, published that last year an ad with jew ellery. They were showing some diamond jewel- lery so they used this technology and it came out very well. I think Vanitha is printing around 2-3 lakhs copies. I think someone from Manorama may be here and they used the technology commer- cially for the magazine’s back cover.


Burjor Poonawala Perhaps, I am asking more questions because I am more ignorant about stochas tic screens and its technology. I have found in my experience that when we are doing our usual course of work, using 150, 175 line screens. The finest printable dot say about 2 or 3% is very difficult for us to hold on our proofing plate, leave alone our printing plate. We’ve tried our best, we’ve sort of varied the exposures, development and everything. But we have not been too success- ful in maintaining the highlight and the shadow end at the same time, we might get around to maintaining the very fine printable highlight dots but at the cost of losing some in the shadows. Now would we not face a similar problem when we use stochastic screening where the dots are uniform- ly very fine and they are finer than the finest dots in the highlight end of a fine screen ruling. So I am just worried that either we have an entire im age if we’ve printed it… sort of transferred the image on the plate successfully, or there will be no image at all. So can you enlighten on that feature?


Veronique Sierens Yes, there are couple of param- eters that we have to consider when we want to copy the film on plates. The density of the film is the first one that I would like to highlight. It is true that if today you copy even a conventional screen, you will see dot loss when going to the printing plate. That’s why the density of the film is of major importance. If you have a density on your film of 3.8, you can get away with it, when printing it on plate. But I can assure you that with stochastic screening technology with a density of 3.8, it will be a complete failure when going to the plate. That’s why a density of 4 or even high- er-4.3, 4.4, 4.5 is very important. But there are other parameters that will influence the plate copy. Vacuum is another one that is very important. We recommend double vacuum time, when copying the stochastic screening onto the plates. The con- dition of the vacuum frame, the time is important and then of course, there is exposure time. Ex- posure time will be shorter than for convention- al screen. We used the UGRA wedge, to check out the exposure time for the stochastic screening and to give you an example, for the 21 micrometre Crystal Raster, we recommend an exposure time with the UGRA wedge so that the 8 micrometre lines are reproduced on the plates and that the 6 micrometre lines are broken away. So that means very short exposure time. For a conventional screen, using the UGRA wedge, you will proba bly expose at around 12 microns solid, 10 microns broken. So the exposure time has to be controlled and very short. And then of course, there is the plate itself. If the plate cannot hold a 6 microme- tre line, you will not be able to keep 21 micrometre on your plate. So these are a couple of the param- eters that have to be controlled in a good way.


Lakshmivenkataraman I have got a doubt regard- ing the colour correction for the FM screening. We have UCR and GCR methods for colour correction. So can we use any one of the methods for F.M. screening or which is recommended?


Veronique Sierens You can continue to work as you have been used to in terms of colour correction, UCR, GCR, nothing has to be changed in partic- ular for stochastic screening. That will stay the same as it is.


Lakshmivenkataraman. And for the proofing also, proofing devices do they support FM screen- ing, like thermal wax and all those things?


Veronique Sierens Proofing is, of course, also a prob- lem again you need a proofing device, that can keep that very small dot, the traditional proofing systems, the chemical proofing systems, that are based on toner and powder systems cannot always keep the very small dots, that’s why the proof press is the best proofing system for the stochastic screen- ing technology.


Lakshmivenkataraman… Any electronic printer is available for FM screening support?


Veronique Siere Not really, no.


Lakshmivenkataraman And regarding the screen angle, when you are using for duotone and tri- tone, the same method can be used?


Veronique Sieres The same method can be used. There is no angling, everything is random, so if you use it for duotone, black and white, four co- lour, it’s all the same technique.


Lakshmivenkataraman… And what should be the resolution of a TIFF or EPS file to compare to con- ventional screening and should it be the same? The resolution of the file… like a Photoshop file or something.


Veronique Sierens As I said in my presentation, you can use the same factor two sampling method for the scanning. You can go lower in scanning reso- lution, that means you can go down to 1.5 sam- pling. So if today you scan 2 times 150 becoming a 300 dpi file, you could go down to 1.5 times 150 and that will give you a small file size. You do not have to go down. You can go down…. We even have customers who go higher, because going higher, you can capture detail that you can still reproduce with stochastic screening which you will not be able to reproduce with the conventional screens. But 1.5 would be the lowest.


Sunil Khullar Madam, just in your last statement, when you say 1.5 and 2, what LPI are you refer- ring? What is LPI in stochastic? Is there any con- cept of LPI in stochastic screening?


Veronique Sierens I am referring to the normal scan- ning formula. If you want to output 100% image for 150 LPI (lines per inch) you will scan it 2 times 150, that is 300 dpi for the scanning resolution, that is what I was referring to.


Sunil Khullar But, Madam, in conventional Screen- ing, we have something like 133 or 150 output. But in FM screening what is the analogy?


Veronique Sterens In FM screening, you will specify the fineness of the stochastic screening by the size of the micro dots. So you talk about 30 micrometre dot or a 40 micrometre dot or a 20 micrometre dot.


Sunil Khullar When you are using a 20 micrometre dot, what is the equivalent in conventional

Screening? Veronique Sierens It will be a equivalent to 250 lines per inch… Sunil Khullar And how is that arrived? TN Rajan Equivalent to a 133 line screen is… 304 is a resolution (pixels per inch).


Sunil Khullar I am still not clear about this, because at one stage we are talking about random frequen- cy modulated way of generation, and on one side it is frequency modulated and LPI is the frequency.


Participant If you just analyse what is amplitude modulation and frequency modulation, you will be able to answer this question. Now in half tone printing, the size of the ink dot is varied. We are varying the dot size of the ink that is the ink size may be a small dot and the ink size dot increases, that is amplitude modulation whereas in the fre- quency modulation the dot size is constant but the frequency with which they are placed is varied, that is the difference whereas in amplitude modula tion or half tone regular printing, the number of lines per inch is fixed that is 150 lines per inch is fixed, but the dots vary, depending upon the… your continuous tone image, whereas in frequency modulation the dot size remains constant but the frequency with which we place these lines, varies. In some areas it may be 150 LPI, in some areas it may be 120 LPI, in some areas it remains 90 LPI, depending upon the density of the continuous tone. So that is the difference between amplitude modulation and frequency modulation. So it’s exactly… the amplitude modulation is LPI technology, the frequency modulation is stochastic screening, if I am correct.


Naresh Khanna Mr Rajan commented that on smooth or flat areas and surfaces, metallic surfaces, etc., frequency modulation, stochastic screen- ing, tends to create a rough look visually. Is that actually a problem or is there some way deal with that in stochastic screening?


TN Rajan Some critics claim that smooth surfaces do not appear smooth, but from our experi ence, in ScanGraphics’s hi-fi screening, smooth surfaces appear as smooth itself.

Participant This question is to Veronique. Dur- ing her talk, she said that because of use of FM screen, there will be less dusting on the print- ing machine and also less frequent cleaning of the blanket. Basically what we know is, paper is the main factor which contributes to this dust- ing. Could you please explain how the screen- ing can control this?


Veronique Sierens I am not a printer, so I don’t re- ally know the reason why this is happening? This is just a testimonial from our customers who ex- perience that they have less cleaning to do. Un- fortunately what is the reason exactly I cannot give you this answer.


Participant Just one more question, I am tempted to ask because an important component of this screening is the cost which was also mentioned as a deterrent why so many people are not using it so widely. I don’t want exact figures but would Mr Rajan be able to say that as compared to the conventional colour separation output, how much more you think it will cost. I mean from your own house, are you just doing a stochastic screening or are you also doing colour separation and how much more do you charge for stochastic screening?


TN Rajan For stochastic, we have put up a pric- ing at 15% more than the regular screening.


Participant With the printing process would the cost be more?


Mr. Rajan No, I don’t think so… printing pro- cess in conventional and both you know, we have to use the same….There is no need for increasing the price for that. When you print a higher value screen… 200 line, the same care you give it for this also. Ink usage will be less comparing to, that is… the theory say, I don’t know. We are not tired that way.


Lakshmivenkataraman To get a 256 gray level in normal printing, we adopt a resolution of say for example 150 lines per inch screen, we set the imagesetter resolution into 25-40. Does FM screening any way relate to that, anyway can it with less resolution of the imagesetter, will give the same quality, by choosing for example 30 microns for a newspaper printing, instead of using 1270 lines per inch, can I go for a lesser resolution?


Veronique Sierens With our Agfa screening we can give you the same quality at a resolution of 1800 dpi on the imagesetter.


Freddy Poonawala While implementing stochas tie, is it important that one creates a profile of screening with the press that you’re ultimately going to print with, or can one can safely go ahead and go by values that industry uses of dot gain?


Veronique Sierens If you output the film with sto- chastic screening, you have to know where it’s going to be printed in order to guarantee the quality. Every press has its own behaviour in terms of dot gain and there is not a standard dot gain with stochastic screening.


Freddy Poonawala That’s probably one of the reason why it is not picked up. Because prepress houses and presses are not as good as well..


Veronique Sieren Yes. It’s easy to implement in a printing house who has their own prepress installation but selling it to colour house who is working with multiple printers is more difficult. That’s why if you put it in a colour house the relation with printer is of highest importance. Correct


Burjor Poonawala One last question that I might ask again is that what bothers me most and what bothers most prepress people is dot gain while printing that is a major problem facing printers and it is always a dispute whether the prepress person is at fault or printer is at fault especially with finer screen rulings. We find that perhaps the experience is that in middle tones the dot gain is now whether it’s a middle tone or high- light or shadow, in the stochastic screening the dot size does not change and therefore, why is it that the stochastic screen has a greater dot gain than a conventional screen?


Veronique Sieren Yes, I understand your question. This has to do with the perimeter and that is the circumference of the dot. If you look at a 50% con- ventional dot, you can calculate the circumference. of the dot, if you look at a 50% stochastic dot, you will have all the small dots and if you calculate the total of the circumferences of all the small dots, that’s the perimeter value, and this value is much greater.

Indian Printer & Publication                                         May  1997


Proceedings of the third annual pre-press technology conference Part 5


Held at Bangalore on 28, 29, and 30 November 1996

Introduction of Mr. Satish Sharma by Naresh Khanna


On Digital Cameras we have a number of speak- ers, including some experienced users who’ve come here as participants. Mr. Satish Sharma of Associated Press is the first speaker. He is an elec- tronics engineer who has been with Associated Press for 14 years. He is the Technical Service Manager of Associated Press for Asia and has three years experience working with digital still cameras.

Satish Sharma on Digital Cameras


Good Morning, I have been asked to talk about Digital Cameras. The Associated Press is the pio- neer in Digital Cameras, we are the first to come out with Digital Cameras for news photography. As far as the news industry is concerned, the LC 2000, was produced with the cooperation of Kodak. We are using Kodak’s electronics in it, there’s nothing much to talk about it but if peo- ple want to have a live demonstration, we can take pictures and show them how you can capture the images on laptops and you can transmit. Associ- ated Press is using this camera for all the major assignments now. We are one of the major news agencies in the world, we are using this Digital Camera now for almost all the important assign- ments-for the Olympics, this camera was used on a very daily basis and 90% of the pictures from the Olympics were produced through this Digi- tal Camera and got printed in most of our news- papers in this country. There are many newspapers in India using this camera now we have The Hindu, The Times of India, Bombay, who are using this camera. The Press Information Bureau (PIB) of the Government of India bought this camera, and it was used for the Prime Minister’s visit to Italy. When the Prime Minister was in Rome, the PIB used this camera and took pictures with it.


As far as the electronics in this camera is con- cerned, we are using the Kodak Imager CCD, which is 25 millimetre x 16.4 millimetre, the pic- ture size is about 3.5 MB, every frame is of about 3.5 MB, I mean the data in one picture.


The best thing is, I can invite people to come to my stand and they can really have a feel of the camera, and we can let them try it out. It’s basi- cally a Nikon body, and we are using a Kodak back in this camera. There is a 105 MB PCMCIA card, which stores about 80-100 pictures in one card. This camera also has a voice caption-after you have taken a picture, you can speak your caption into your camera, it gets recorded and when you go to your laptop and scan the frame, you listen to the caption. It’s very easy to work with.


The newspapers can receive the digital pic- ture service in their production room. Basical- ly, what they need to do is install a dish and As- sociated Press provides all the required equip- ment so that the picture service can be received. It is a very high-speed picture service, one co- lour picture takes about 45 seconds to come, and we move about 100-150 pictures on this network. I’ve also installed a dish and the pic- ture service is coming at the stand. You might have seen The Times of India, Bangalore, they’re receiving that picture service and they’re print- ing colour pictures. If you look at the picture of the Chinese Prime Minister who has come to Delhi, the picture of the Indian Prime Min- ister and the Prime Minister of China, is on the front page today. That picture has been taken in Delhi and moved through the satellite. Rep- resentatives from the Press are welcome to come to our stand and have a look at the picture feed. It’s going to be very nice, the picture will be coming very fast, much before the deadlines in the newspapers.


Thank you very much.


Niels Orghaard on Digital Cameras


Good Morning, my name is Niels Orghaard, I am sales director for Scan View, which is a Danish com- pany manufacturing pre-press products as men- tioned.


We have had a lot to talk about, we have dis- cussed trends and technologies and it might have puzzled you that we did’nt talk about the Digital Cameras at all, because we from the pre-press busi- ness certainly know, we must face it, that this is the trend. This is a very serious trend that we see worldwide. There is a lot of work going on with Digital Cameras, we see new products coming out almost daily and I think that the discussion we should have today is actually what is Digital Pho- tography and how can we benefit from it. Is it something that we can use at all and see, how in the pre-press industry, that we can actually imple- ment this new technology. You can see on the first slide here, that you don’t see the name Scan View, and again as I tried to show you, we try to stay a little bit on the leading edge with the development of new technologies and new products and in fact, ColorCrisp is a 100% owned subsidiary of Scan- View.


Digital Cameras have been around for quite some time; in fact over the last four, five, six years, we have been able to get Digital Cameras of some sort. Now what we did four years back, is that at the time we were developing imagesetters, we were developing new scanners, that is our bread and butter. We are a pre-press company. We also know that in the future we cannot rely on our business with our scanners and pre-press equipment only. We have to look into what is the trend of today and the trend as I mentioned, is Digital Cameras. So we set up another company. Actu- ally, the founder of Scan View, Mr. Samilahaff, who invented the desktop drum scanner, he set up a company called Phase-I. Some of you might have heard about that and that’s actually implement- ing one of the two digital back scanning technol- ogies. Now, after a period of time when he launched the products we are still selling, we found that the technology was not the one that was going to be used in the future. So another company was founded, again by Mr. Samilahaff, which is fully owned by Scanview and that com- pany’s name is ColorCrisp.


I will come back to why it is beneficial for a company like Scan View, to have another compa- ny name when it comes to Digital Cameras. First of all, let’s see what is Digital Photography.


In a very basic way, you could look upon it as a new way to generate some revenue and save money in the pre-press process. There are a lot of things involved when you’re actually a photog- rapher. The name ‘photographer’ is the reason why we have two different companies, because when you’re selling into this segment with Digi- tal Cameras, you’ll find that there is a very very big merge, photographers speak another language than we normally do. So we had to get people into our company that understood the requirements from photographers, which can be quite differ- ent and the language we talk and they talk is not necessarily the same. We had to have people who understood what they actually wanted. Actually right now, let me try to ask you, is there anyone here in the audience who has a Digital Camera already? May I ask what kind of the Digital Cam- era are you using? I’ll come back to that later. There’s one right here as well which is a Kodak. It means that there is some familiarity with the Digital Camera as such but the discussion we are going to have today is also the various types of cameras and various purposes of Digital Photog raphy


Why bother at all with Digital Photography? There is a great deal of time-saving involved. You’re not using any time used or any film han- dling, you don’t have to send it out or you don’t have to get it developed or exposed, you don’t have to transport it, you don’t have to use time or money to do the scans and this is where we in the pre- press business should be a little bit aware of this technology. There’s no Polaroid cost and I can tell you, if you are doing a catalogue, if you are doing a 100 page catalogue, a 50 page catalogue, high end, high quality, your cost for Polaroid will be enormous. No film cost, of course. You don’t have to have it processed and you don’t have to have it scanned.


Well, anyway, let’s have a look at how the pro- cess is done, whether we are talking convention- al or the digital way. A very simple way of describ ing it is that you actually take the shots, of course, you have the film, you transport it down to wherever you have it processed, maybe you have it in-house already. You do the processing, you have it transported back and then you send it out or do it in-house. Another trend we’ve seen is that photographers worldwide, they are evaluating the technologies today, either they are going to buy a scanner themselves, working on the Macintosh gives them the greater creativity in the work they doing or they are going straight to Digital Pho- tography.


As I mentioned before the Polaroid cost, you are taking a Polaroid shot when you have a very big setup. In a studio, when you do a shot, when you have a lot of things, for example, when you’re doing a shot of furnitures, etc., you would make


Polaroid to check your lighting, you would make


Polaroid to check that everything is actually with-


in the framework, you are thinking you’re shoot-


ing. Now, when you are doing it digital, you’re just taking the preview, immediately when you take the shot with the Camera, you will have on your monitor, a preview, you can see the set up is cor rect, you can see the lighting is correct, you can actually check the lighting on the monitor itself, you can change the lighting or you can change it with the histogram as we used to do in Photoshop, you can brighten or darken the image and that’s it, after that you save it and you have your RGB file.


Furthermore, another benefit that Digital Photography gives the photographers is that they no longer just supply let’s say, an RGB data, after they get it scanned somewhere, they can actually control themselves a CMYK data. They can deliv er the file in RGB or CMYK as they wish and they can use it in a standard software like Color Quar tet-as we know, it handles flatbed scanners, drum scanners as well as Digital Cameras


The best way to actually find out whether the Digital Camera is feasible for you is to set up a spreadsheet and calculate the cost The cost locally is: what does it cost to have the film pro- cessed etc. etc. all these numbers mentioned up here. It would very quickly show how soon you can actually have a return on your investment, when you buy the Digital Camera. Of course, you can achieve your results faster than you normally do, it also gives a photographer a lot more flexi- bility. He can use a lot more time being creative, rather than having to do with all the film devel- oping and processing.


Now, before we go in and talk about how the Digital Photography actually works, I think we should discuss the various kind of cameras in the market. I think interestingly enough, today on the world market there are two segments equally large; there’s two segments buying Digital Cameras in the high end of the market. One is, or 50% of the market goes straight into the photographers and 50% of the market today goes to pre-press hous es like many of yourselves. We have seen a great merge in the world, we have seen a lot of pub- lishing houses, they are setting up in-house stu- dios and today I might add, the majority of high end Digital Cameras sold, are going to publish ing houses or printers, that are doing catalogues.


So, when we are looking at the various cameras in the market, you can find a camera for 1000 US dollars. A camera for 1000 US dollars is actually meant for consumers-you’re using it instead of having a standard camera when you are going on vacation, you take your shots, you can take them back and then you can look at it on your moni tor, or you can have them processed at various places and actually get them into analog form. A price of a 1000 dollars does not give you a very high quality. So, to do professional jobs, we have to go up into another segment, which would come in the cameras we have here today, which are very professional devices, and we have seen a lot of pho- tographers who’re working for newspapers espe cially, they have a lot of benefit of capturing the images on digital format rather than on analog format. The next step, the high step today, is in the price bracket of around 30,000 US dollars for a Digital Camera. Now, this technology enables you to print in very high quality and I will tell you little later, what the quality levels that we can achieve the Digital Cameras, actually are. There are two technologies, basically there only exist two different technologies for Digital Photography. One is the scanning back. This is a linear array and it is exactly like you would see it on one of the older 3-pass scanners, that was on the mar- ket. And the other one is the array CCD chip and we should try to have a look at how they differ.


First of all you have an black and white CCD chip. They are normally in the size of 2048 x 2048, when we talk about high-end of the mar ket. We have the 4000 pixel element that can capture the data. The advantages of this chip here is that it can work with standard stroke lighting. It is a important factor that you don’t go out and invest in special tungsten lighting. etc. It is very fast, it would act for a photogra pher like a standard film, an ASA 100 film. It’s colour accurate and you can get one-shot black and white images. Now, of course, there are some things that are not so good about the black and white array CCD chip, still it’s relatively expensive and you can only shoot black and white here and the file size is limited to a 12 MB file size.


If you go into the colour-coded array, again the advantage is you don’t have to invest in any special lighting system. They are very fast. You can have a one-shot colour capture. We see them on the market. There is a leaf camera in the market which is called cat-slide and the reason I mentioned this is that, that is probably one of the first cameras that has been sold quite well into the segment that I am talking about-the high-end segment especially for catalogue works-and again we get a pretty good dynam- ic range upto 10 F stops. We have to recognise the fact that when we’re taking Digital Photog raphy, probably one of the reasons why it has’nt really taken off in all countries yet is that, the investment we have to make is quite high and that is again because of the CCD technology. The CCD is byfar in all cameras in the market, the most expensive device in the camera. An- other thing that we will see and we’ll talk about the quality how much we can actually print from the Digital Camera-and that is that we have to, when we are talking about one-shot in colour, interpolate upto 12 MB file size and that will give you some colour artifacts, if you try to enlarge the images too much.


The other technology in the market is lin- ear array and like I mentioned before this is the Phase-I camera, they’re using this technology. What you can do is you can generate some enor- mous files. A 150 MB files is no problem at all. The technology implemented is not very expen- sive and the camera as such is much less expen- sive. You’ll find that this technology can only be used for studios, it can only be used in-house, it can only be used for stills. One shot of any given furniture, or whatever you try to make a shot of, can take upto 15 minutes. So you can imagine the importance light has, the light has to be completely stable, otherwise you will have colour artifact in the image when you’re finished shooting. So they have to buy very special, very expensive lights to do their shots with this cam- era. The benefit is definitely that you can gen- erate very large file sizes.


Now the Carnival 2000, we are going to mention because again it differs in technology from the other cameras. We’re using a colour array chip and we’re using it a little bit differ- ent than the other vendors. As you can see here, what was shown on the screen was, that this cam- era has the capability of doing something we call one-shot and a multiple shot. On the market you will find that as, I mentioned the cat-slide from leaf is a one-shot camera that allows you to take images of moving objects. On the other hand, when you need to take images which are a 12 MB file through file size not interpolated, you need to do it with a multi-shot camera and again leaf has a camera called the DCP 2. That is two different cameras, why would you need a multi- shot camera? What it actually does is it takes three shots, it filters the colours through, it takes red, green and blue shots and then it makes some cal- culations and builds up plains and then you have your shots. This shot is a multi-shot and what it will give you is that, if you’re printing at 175 lines, you will be able to output that or enlarge that upto an A-3 size. That is one of the limitations we have even with high-end Digital Cameras, that if you’re shooting in multiple shot mode the maximum output size at the 175 lines would be A-3. If you try to enlarge it further than that you will start to see some colour artifact and you’ll start to see some coroness in certain areas.


Now with the Carnival 2000, this new tech- nology that has been implemented here, you’re actually able to do one-shot and multiple-shot in the same camera. (We’ll just try to see you can find it, it’s a little bit easier to show when you have some- thing to show it from)..


Again, we’re taking shots here of moving objects. Everybody working in a studio, they will know that when they’re doing a catalogue, they have to do something called fashion-shots. It mod- els that are showing fabric, etc. you need to be able to do moving objects at a relatively high quality. When we’re taking shots in a single shot-mode, the maximum output size in 175 lines is A-4, because it is an interpolated file upto 12 MB and again you’ll see the colour artifact. This is a known fact, all cameras have the same problem, there’s ways to work around it and we can do everything we want in photoshop, etc. and all other softwares. But I am talking about the basic features of the cameras.


Now a colour-coded pixel element has some benefits over the other technologies like the black and white CCD elements I mentioned. The black and white, what they would do is they would put some filters in front of the CCD chips, what we believe is that when you’re involving any mechan- ical movement it will actually generate some vi- brations. So we have chosen to go for colour-coded CCD array. And how it works, and how can this camera do both multiple shot and one-shot is be- cause we have utilised the new technology called “Pixo” technology, that actually allows us to move the CCD element, one pixel row at a time. It’s a very controlled movement, just one pixel row at a and thereby, we take four shots, we don’t take three shots, we take green, red, green, blue and again it’s a matter of how you actually get the best possible data for your file. Because the hu- man eye is very perceptive to a green information that’s why we sample two times.


I think, that was a basic introduction to where we are with the digital photography today, what can be done and again, all the questions are wel- come. Thank you very much.


Patrick Anderson on Digital Photography


Good morning. I am totally unprepared for this. I came to find out what you knew about photog- raphy, not what I know about photography. I used my first Digital Camera almost 20 years ago. I work as a consultant in photography and digital imag- ing, basically because I’ve been a photographer for more than 30 years. I was lucky enough to grab hold of a Kodak DC 3 to cover Miss World, which meant that I was able to transmit pictures of Miss World, elephants and security permitting. 13 minutes after she was crowned. If it had’nt been for the elephants, I would’ve got the crowning, if it had’nt been for the security, the picture would’ve moved 3 minutes after she was crowned.


It has great advantages. The biggest problem in India with Digital Photography, I find is a thing called Indian telephone lines. I’ve actually spent 12 hours moving a small digital file from London to India, due to your telephone line problems. Basically, I haven’t got a lot to say, but maybe I can answer some of your questions, because I’ve used it. I worked as an Agency Photographer for Reuters and some of the other agencies, I still work as a photographer.


My feeling is the photographers have got to get more into your field which is the pre-press, and you’ve got to get more into their field. In other words, the lines between the two have to be re- moved completely.


“My feeling is the photographers have to get more into (the pre-press field) and (the pre-press people) have to get more into (the photographer’s) field. In other words, the lines between the two have to be removed completely.”


My guru in the printing world is up here. In the year I have been at The Hindu, he’s taught me an awful lot. I never knew about printing. I now know that there are certain things I shouldn’t do with my images, and I try and tell the photogra phers they shouldn’t do, because he can do it a lot better. So I’ll let RPL talk to you. He knows more about it than I do.

R.P.Lakshmivenkataraman on Digital Cameras


I am going to add nothing new to what our pre- vious speaker, Mr. Ohrgaard has said. But I just prepared something from some of the magazines, courtesy of Agfa and the IFRA pamphlets. Actu- ally, the Digital Camera comes into picture, due to the development of CCD, like linear array and matrix array. Linear array helps in the develop- ment of scanners, whereas matrix CCD array de- veloped into the Digital Cameras. Normally, the arrays will be 2048 x 2048, but most of the con- ventional cameras that we have shown here use about 1024 x 1280 pixel elements. Why I am dis- cussing the pixel elements point is because when we pre-press people scan it, we look at the sam- ples, the file size, whether the Digital Camera pic- tures are suitable for printing and publishing, can it give a picture equivalent to a film. That is basi- cally what needs to be looked into.


Normally, a conventional film, whether pos- itive or negative, consists of about 16 million pix- els. But imagine the number of pixels the Digital Camera chip is having. We cannot underestimate the Digital Camera pictures as unuseable. The technology, the way the elements are designed, the way a picture is captured, how the light is in- tegrated between the elements of CCD, and the lens mechanism, make publishing it possible. IFRA has published on a news print, it is nearly an A-3 size and some of the digital photographs from Patrick and even wire photos from Reuters, are also published as full A-4 size in art as well as glased paper. Mainly the advantage of Digital Cameras, as everyone knows, is the time schedule.


This, as you can see in the slide here, is one of the Digital Camera pictures published in a news- print, 283 x 200 millimetre size so even though the pixels are smaller, still it is capable of produc- ing bigger pictures. The working of Digital Cam- eras has already been explained, I would show you some slides.


The principle is that when the light particles, the photons, strike on the CCD element, they re- lease negatively charged electrons. And each CCD element has an electrical contact to it. By apply ing a certain voltage to this contact, the area of the silicon below it becomes receptive to the freed electrons. That is the working of the CCD array. The released electrons are proportional to the light, and this electrical signal is quantised to send the digital information. That’s common to all CCDs. Normally, the CCD chips are, what in present-day cameras will be smaller than the 35 mm film size. So, the area which the photogra- pher captures will not be the same frame; there will be a small cropping which will be marked on the camera itself, that’s the area which is going to fall on the back of the CCD array. Normally, if you take pixels of about 1024 x 1280 mm, the file size is going to be about 1.2 MB. Imagine the file size you are printing in colour. Can this 1.2 MB file be printed? Yes, it can be printed. These are the experiment results and in the market there are different kinds of cameras available. AP (As sociated Press) has shown the Kodak body NC 2000, Kodak DCS 3C, Kodak DCS. Each one has a difference in elements, the number of elements and the size of the chips. Actually, the size of each CCD element also matters. The amount of light it captures, decides the quality of the pictures. That is a factor we should look into when we decide on a Digital Camera. For example, the size of a chip which is used in a NC 2000, will be about 16 micrometre x 16 microns, whereas in Nikkon E2 and ES it is 6 micron x 6 micron, still the num- ber of pixels is the same.


What I am saying is that the size of each ele ment varies. Variation in the size of the element, means that the amount of light capture also var ies. According to that, the colour depth will vary. It is about 8 bits and 12 bits. More bits will get more detailed information in the Digital Camera. When we use the 3 channel, the file size in case of 24 bit or RGB image is around 3.7 MB. That is what Mr. Satish Sharma has shown; in a cartridge of 105 MB it can capture about 70 to 80 files


But in these files, when you look on to the true optical resolution of a picture, it comes to around 1.22 MB only. But these files are again interpolated. As he was showing in a slide, you would have noticed it is not RGB, it is not RGB, RGB. He has taken human visual persevering things. Two green pixels will be there. He’s giv en more importance to the green lights, green pickup, which normally our eye is seeing. So the interpolation is also generated according to the perception of the colours there. So that is the way in which it defers from the normal calculation of resolutions. These image-captured files are trans- ferred to PC cards. PC cards, earlier called PCM- CIA, are cards which go on to all the laptops and powerbooks, and now they have been developed especially for Digital Cameras. These are known as PC cards 80A Type I and PC Cards 80A Type II and also go by the name of “flash memory cards”. They have shorter pixel storage, but fast- er memory, something like D RAM and other things The PC Card 80A Type I is about 15 MB and the Type II Card is about 40 MB. But these pictures are stored in a compressed form in the flash memory cards itself.


Normally, Type-I will take about 5 to 84 im- ages whereas Type-ll will take about 15 to 199 im- ages. The PCA cards Type-III, what Mr. Satish Sharma has shown, come in a different capacity. About 150 to 340 MBs, are normally used and they’ll capture around 70 to 250 images. It is not just the optical resolution that decides the size of the image, but the interplay of censor elements, the interpolation algorithm, and the quality of the 


“…it is not just the optical resolution that decides the size of the image, but the interplay of censor elements, the interpolation algorithm, and the quality of the lens system, all help to print bigger on newsprint and Digital Cameras.”


lens system, all help it to print bigger on newsprint and Digital Cameras. I have made a few slides for your visual interest, on other cameras and the mechanisms. This slide is about-courtesy of Agfa-how


our eye sees an image, how a conventional cam-


era sees it, how the Digital Camera sees and picks it up, with the lens system. The second one is the film that goes to the back of the retina, the brain senses it and the im-


age is recognised by the brain. The next one is about how the image falls and what the film consists of. The film consists of lay- ers like a blue layer, then there will be a yellow filter to prevent the blue light to fall on the back of the green and red layers. The slide at the bot- tom is the Digital Camera array chip, where the chip is placed and how the gate is connected to the all CCD elements. That array is connected, through the SCSI board, to the hard disc. These are some of the cameras, this is a DCS 460, next one is from Agfa, I think it is a studio-cam, one is a still camera, one is an action camera with its PC


cards showing at the bottom. This next one is a Nikkon ES2 Camera, and the back side of the Camera, where they normal- ly put the flash memory cards.


The cameras shown in this slide are the Kodak EOS DCS-5 and the ES420. Of course, Kodak leads in the manufacture of this product, the body comes from Nikkon and Canon.


This next slide is just a comparison of the models, the format and the resolutions with the ISO rating comparisons.


Patrick has captured his images through the DCS EOS-3 camera in the Miss World Competi tion and we have used some of the pictures in our magazine also. They are quite good, but you can- not expect the same result as from a film. With the advent of colour management, you can expect a good reproduction.


Magnification is the only problem when we handle the Digital Camera pictures but we should be careful enough to know the size of the file and quality and only then magnify it. That’s the les son we have to learn. Thank you.




Anil Kumar from Mathrubhumi I am the Head of the Department of Electronic and Electronic sec tion there. I have a few samples with me, which are printed using colour in Digital Camera. We are using DCS 420 Camera with Nikkon N90 and we are using 600 DPI laser printers for taking the separations. A few samples are with us for circu- lation, and most of the pictures connected with Miss World Competition 1996 are taken with this camera and all other news photos are taken and scanned in by the flatbed scanners Arcas-2 and Microred-3 and UMAXPower-II and all the sepa rations are taken on 600 DPI laser printers. You can see this.


Naresh Khanna If the audience understood, they took pictures of the Miss World contest with the Digital Camera, transmitted them to Kerala from Bangalore and reproduced them using 600 DPI laser printer output. Is that correct? And did you use 4 outputs? And then shot those with the cam- era, is that what you did?


Anil Kumar Yes…?


Naresh Khanna… you took 4 laser printer out- puts, at different angles, you took a line shot of that on a conventional copy camera….


Anil Kumar No No, We have taken it on a polyes


ter master film, directly on to the plate. Naresh Khanna… 600 dots directly on the film?


Not in the laser printer?


Anil Kumar… Not on the laser printer, we used laser film, Nova Sharp film from Technova.


Naresh Khanna… That means on a laser print- er, they used, instead of using paper…


Anil Kumar We don’t use any imagesetters, we use only laser printers. This is only for news photos not for the advertisements. We use Brother HL 1236, or you can use Epson 5600 also. You should be very careful when you select the laser printers because in some laser printers you can having banding effect. So you have to see the dot percent- age, you have to output the screens, then you se lect the laser printers. HP I think, is having some banding effect, I could see during the evaluations


So that’s why I went for the laser printer. Naresh Khanna So this is colour reproduction with-


out a scanner, without an imagesetter…


Anil Kumar… That depends on the paper we use.


Anil Kumar (In response to a question) Yes, the


photos connected with Miss World everything is fitted on the Digital Camera and we even trans- mit this photograph through the ordinary P&T line, using the 2400 modem. It took 10 to 12 min- utes for the transmission to complete, with 96% efficiency of the transmission, with a throughput of 230-235 CPS


Burjor Poonawalla I am a bit confused. I should have thought that when you get an image from a Digital Camera, you can output separations directly without having to go into laser printers so where does the laser printer come in?


Lakshmivenkataraman I don’t have an imagesetter.


Burjor Poonawalla Oh, you don’t have an image- setter?


Naresh Khanna There is no conventional camera?


Burjor Poonawalla But then the purpose is self-de- feating. Now can you tell me, have you employed the same technology as far as the picture on top is concerned, or is that conventional photography? The picture on top has been done by conventional photography and the picture of Miss World…


Anil Kumar That has been taken from the origi nal photograph. That event we had missed and we have taken that photograph from a conven- tional camera.


Burjor Poonawalla


This is conventional?


Anil Kumar No, No, the last one.


Burjor Poonawalla This?


Anil Kumar Yes.


Burjor Poonawalla And how about this picture?


Anil Kumar That’s a conventional camera. We have taken the separations in the laser printer, only the last photograph is taken with the Dig- ital Camera, that is also from a photograph. We scanned the first photo. It is around 180, 170 DPV scan, then we took separations with 85 LPI.


Naresh Khanna I think that what is shown by Mathruboomi, is very interesting concept. The idea is that they have done away with the con- ventional camera, the transparency, the scanner and the imagesetter and have been able to trans- mit on a fairly low-speed modem over conven- tional voice grade telephone lines and have used these photographs in the newspaper. I think, later on perhaps we can get back to discussing the merits and demerits of this approach, the quality issues, etc. At this time we would like to address questions concerning Digital Photog- raphy. You have in fact shown a total alterna- tive approach to pre-press, which quite interest- ing. We will get back to it later.


Participant Sir, this is about the Digital Cam- era and the conventional camera. Suppose, I buy a conventional camera by paying Rs. 1000, and get a snapshot camera. I load it with a black and white film, I take black and white pictures. I load it with a colour roll film, I will take colour pic- tures. If I buy a Digital Camera, suppose I want to take black and white pictures, can I take black and white pictures? Suppose I want to take co- lour pictures, can I take a colour pictures? Or should I have two different cameras-one to take black and white pictures and one to take colour pictures?


Patrick Anderson I would like to answer that for I two reasons, one, first of all, let me clear up a thing I have been asked in the past two days dozens of times. You cannot put film in a Digital Camera. That’s the first thing. Secondly, you can with any Digital Camera, shoot colour or black and white, basically it’s going to shoot colour unless it’s a lower-end Digital Camera, but you can then take it out of the camera as a gray scale therefore it’s black and white. One of the ad- vantages I see, of the Digital Cameras, is some- thing I’ve argued with many pre-press people for many years. A colour negative has a high- er dynamic range, than transparency. Most pre- press people will only work, or prefer to work, with a colour transparency. My feeling is that if people worked with colour negatives, which has a dynamic range of about 9, and as Neils said, the Digital Camera has a dynamic range of 10, you’d actually be able to use far more of the images that were shot.


Participant No, my idea is, if I buy one camera, I can take pictures in black and white as well as colour depending upon the roll I choose, so the consumer is interfaced with choosing a roll- whether he wants black and white or colour. Suppose I got a Digital Camera now…


Burjor Poonawalla I think your question is quite


understood. The thing is you can have a colour


Digital Camera and produce both the images,


black and white and colour. Patrick Anderson Let me give another advantage of a Digital Camera. During Miss World, basi- cally I was shooting colour. I could shoot one frame at 200 ASA, the light changed, I did’nt want to use a flash, the next frame I can shoot at 1600 ASA, I can then go back to 400 or 200 ASA, in other words, every single image is an individual picture, one image could be basically a black and white file, and the next one could be a colour file. So every time you shoot, you shoot what you want to shoot for the optimum results. The other advantage of the Digital Cam- era I see, is if you’re using conventional lenses you had for your other cameras, if I put on a 200 millimetre lens to a standard Digital Cam- era, like the Kodak camera, it’s actually equivalent, in 35 millimetre terms, to 380 millime- tres. In other words, I don’t need spent so much money on very long lenses to get the same effect. The only disadvantage is if you’re used to using wide angle lenses, you cannot really use a full wide angle lens because of the restrictions in the frame area. A 20 millimetre lens is about the same as a 50 millimetre lens on a 35 milli- metre camera. And if anybody wants to see it, you can look at the Canon I am using, I use my standard Canon lenses on this Digital Cam- era. For most of Miss World, I was actually shooting on a 35 to 350 millimetre zoom which gives me the equivalent range of about 60 millimetres to somewhere in the region of 500, 600 millimetres. And I could cover everything in one lens.

Indian Printer & Publication                                         June  1997


Proceedings of the third annual pre-press technology conference Part 6 


Held at Bangalore on 28, 29, and 30 November 1996

Burjor Poonawalla introduces Mike Webb

Fin Australia. I can identify myself with him RIENDS, MR. MIKE WEBD REPRESENTS DUPONT rather well because he also seems to have started as an apprentice, doing his photolithography, and reached his journeyman status after several years, worked in several companies and moved through the ranks from scanner operator to prepress man- ager. So we perhaps might be able to ask him a few questions on scanning and operating also. It was in 1976 that he joined DuPont, UK as a technical representative and moved on to the position of Proofing Specialist in 1979. In 1983, he migrated to Australia and rejoined DuPont where he still is. His current position with DuPont is Proofing Sales Manager (Asia) and in that capacity he has the responsibility of selling and supporting proof- ing businesses throughout the Asia-Pacific. He is also a sitting Committee Member on the Austra- lian Graphic Arts Council and is called upon to address various industry bodies from time to time on subjects related to proofing and standards with- in the printing industry, both in Australia and in the Asia-Pacific region.

I think we can look forward to his presen- tation because I know for a fact that we have lot of problems in the area as far as pre-press and printers are concerned and of course advertising agencies.

Mike Webb on Digital Proofing

LET ME JUST SAYA COUPLE OF THINGS BEFORE I start. I actually cover all of Asia including India in my job, I don’t think that I am just held at home in Australia, that’s why I am here. And just to prove that there’s more in life than printing, did you know that at lunch India was 302 for 7 wickets?

Probably the most vibrant area of the confer- ence seems to be the questions, so bearing in mind that it’s running a little late, I’ll try to answer as many as I can while I am speaking, if there are any others, obviously I’d be pleased to hear them.

Coming back to digital proofing, in the ear- ly stages of its life if you like, the dominating trend in the industry right now is undoubtedly the desk- top, some say revolution, rather than evolution. But it’s happening, it has been for some time- changing from the mechanical process that it was, into the digital process that it is becoming.

First we saw the picture images move from a camera on to a scanner. Then we saw the type- setting part of it move from phototypesetting to digital PostScript setting. And now we have image manipulation and colour management via computer, obviously that’s the third step. I know we need to look at it. Well, Okay if we have a com- plete digital image signal now, should we really still convert it to film and make plates and print, or should we try to do something with that digi- tal image and see what we have before output- That’s really what I want to talk to you about to- day.


If we look at pre-press, many people say that in the future, pre-press will be a filmless society. I think we should turn that around and say that there will be a transition, and that pre-press will become a less filmed society. It will be a transition as scan- ners were a transition, computers were a transi- tion. And so if that’s true, if it is a transition, we have a very difficult task ahead. Because that means that there won’t be one day, when analog proof- ing cuts off and digital proofing comes in. The two will have to live side by side, and so one of the most important things we need to look at- you as consumers, whether you make the proofs, whether you okay the proofs, whether you print to the proofs, or whether you sign-off on them, what you need to do is, make sure that when you have an analog or a digital proof, they make sense to you.

Earlier on in the week we spent a lot of time talking about standardisation. The only way for that communication to take place accurately from the computer, the scanner, or the desktop to the final printed result via a proof, is with standardis- ation. So if we think about emerging technolo- gies like computer-to-plate which is still emerging, then I think, we also have to realise that digital proofing is still emerging. There are a lot of ques- tions, lot of wondering, and concern about qual- ity, reliability, cost of course, is it going to work for us, does it work? How much will it be? So, as a salesman, which I am, although I want to talk to you today as an overall technical person, these are the sort of questions we’d asked: What is it? When can we get it? Who sells it? I can’t tell you that in this session, because we have to keep it nice- ly technical and informative, and I am happy to do that. When can you get it? Let’s see if we can answer those questions for you.

First of all, what is it? What is digital proof- ing? Most people’s concept of that is, it’s a plug- in printer that gives you a result from your digi- tal information. That’s probably the simplest def- inition of what it actually is. Let’s look at some of the options of what you have. Before we do that, let me say one thing, that a speaker yesterday said that famous phrase which people use all the time: “You get what you pay for”, and in digital proof- ing it is absolutely true. When you look at what you’re buying, you really need to look hard at what you want out of it. Because depending on the quality, is the price you will pay. Unlike some of the items we talked about yesterday, it’s not a case of if I pay less it will take longer. It’s not the speed of the RIP, it’s the quality of output that counts and in my side of the business, we have what we call a price position statement. If you pay low dol- lars, you get low quality, if you pay high dollars, you get high quality and anything in-between, so that’s a decision you all have to make. I know it’s a tough one, but let’s have a look at what is avail- able from a technology point of view and then talk a little bit about each one.

Basically, you can split it into five. There are many technologies but approximately five tech- nologies will give you an end result which is a dig- ital proof.

Number one is electro-photography, or an- other word for that is xerography, the office copier that you have. That works via light passing through a lens onto a drum which is electronically charged. The light dissipates the charge, the toner in the machine then adheres to the drum and is fused to the paper by heat. That’s why when you take a photocopy out of the machine quickly, it’s still very warm. That’s fine for black and white, but it doesn’t transfer well in colour. When I give you comments like this, I am trying to give you a com- ment from a neutral point of view, that is, from a quality point of view. My opinion of a proof, not a picture. A proof is a standard piece of artwork that you rely and print to, and the customer can look at and know that that’s the end result. Now a lot of things have to go into that-it has to have print characteristics, it has to have the right co

“So if we think about emerging technologies like computer-to-plate which is still emerging, then think, we also have to realise that digital proofing is still emerging lour, it has to have the right dot gain if you’re talk- ing about dots, it has to have the right hue error- there’s many things that go into the standard proof.

So for electro-photography, even when we look at the colour side of it, and even when we use lasers to do it, we come into the more sophis ticated colour copiers. In my opinion, we’re still seeing pictures, pictures that probably don’t show a dot formation- and many of these technolo gies don’t. We’ll talk about that in a minute-they don’t show a direct relationship to things like ink colour and of course, if there’s no dot formation, don’t show anything to do with dot gain.

Look at the second one, it’s called wax trans- fer. There are a couple of these in the marketplace, and basically the way that works is that the colour, the original colour, the CMYK colour, is impound- ed in wax ribbons. It is a very special printing head which contains tint elements, and those tint ele- ments will heat depending on the information that comes to them, which of course, is the binary dig ital information that comes from your system and you’ll either heat or not, and if you’d heat, it melts wax, which is transferred into paper, which is fused on to the paper, and so produces the image for you. Some of the benefits of that are once again the price, they’re usually not expensive items, the drawbacks are that the resolution is usually not ideal because melted wax, which is fused again, does not give the sort of resolution that you’re looking for in a proof which is eventually going to be printed to.

Dye sublimation is probably one of the best known in this country-by the way, the comments I am making now refer to the global market, I wouldn’t presume to know the Indian market, but I give you my information, and you can take it as you will and apply it to a marketplace. But I sus pect that the best known digital proof that is avail able right now would come into that category- dye sublimation.

Dye sublimation really is very similar to wax transfer and that is that, it’s a very special element, except that it’s a long bar element rather than a tint head. And that element heats special dyes, high resolution dyes which are built into a ribbon and the digital information once again creates in the bar, a heating requirement which transfers, but it doesn’t melt. The reason it’s called dye subl mation-the word sublimation is a scientific term -sublimation is the transition from a solid to a gas, without going through liquid. So if you take something simple like water, at low temperatures it’s ice, medium temperature, water, high temper ature, steam. These special dyes, which are spe cial chemicals, do not melt-they go from solid to gas. So that gas is then trapped in the special receptor and creates the image. It can create a very high resolution image this way and so the dye sub limation has a slightly higher quality than wax transfer, because it uses high resolution inks which are obviously a little more expensive.

Then we come on to the DOD. The DOD stands for Drop on Demand-inkjet drop on de mand, and I have to specify Drop on Demand as opposed to continuous flow. And I will tell you exactly about the two of them. Drop on Demand is the typical laser type printer, and that is, that it only sprays ink onto a paper when it’s asked to. The ink sits in the printhead, and when the digi tal image comes upto the printhead and says to the magenta or the sign of the yellow or the black, it’s your turn, it sprays a drop with that ink. The drawbacks to this, are that it can only spray large droplets of ink. So, resolution is a problem and once again, a Drop on Demand is in the medi- um end of digital proofing and gives you a medi- um result. By far the most efficient- and I have to say by far the most expensive-digital proof- ers that are available right now, are the continu- ous flow inkjet proofers, of which there are at least a couple in the marketplace. And the difference with them is that rather than just delivering a drop of ink instantaneously as they get the message from the digital signal, the ink is spraying all the time that they are printing. The digital image then tells the droplets, you will either go to the image, or you go down into the recycling drum and come back again. The benefits of that are huge. It means that with a continuous flow ink stream, you can create high precision proofing; but something even more than that. Imagine a situation: you put a proof-it may be a flatbed printing proof, it may be what I call a pre-press proof, it may be a dig tal proof- and there are six pictures on it, plus type, plus price, and you submit it and it comes back and they say, love it all except this picture here, please change this picture, so you change it and you resubmit the proof. Now what you’re really hoping more than anything, is that when you make the changes, the proof will show the picture’s okay now, but it will also show that the rest of them have not moved. There are very few proofing systems in the market that will do that, and as we move into the higher quality end of proofing and we look at standardisation, when you consider things like the export market dealing with Europe, maybe dealing with the US, that is the sort of quality that we would look for.

So repeatability and reliability are paramount with these types of proofers and you only really get that in the higher quality proofers. So I’d like you to consider that when you are thinking about looking at one of these proofers. Because there’s no doubt that, as I said before, you get what you pay for. So if it’s that easy and it’s simply a price choice and it isn’t, what problems can you expect to see? Let’s just have a look at a few of them. Ear- lier on, I said that some of them were low resolu tion. There is only one type of proofer in the marketplace right now that will give you a digital proof with dots with what you can look at as some what of a conventional halftone. And even that’s not correct, because the dots that you see on that particular proof, are not the dots that you’ll finally reproduce if you decide to go to film. It’s a dif ferent RIP, it’s a different signal. But all of them do not produce conventional halftone dots. It’s a mixure of continuous tone, pixel, and stochas tic, which is, as you all know, random dot screening.

So one of the biggest problems that I see, for anyone who buys a digital proofer, my problem is selling them and that’s a tough job; it’s not as tough as your job, because your job is to get ac ceptability in the marketplace. Somebody’s used to looking at a proof and then gets this new one and says, hey, this doesn’t look anything like what I expect, so it’s a marketing job too. Because our customers’ customers need to be convinced that these proofs really are good quality, and that’s why I say to you, when you’re thinking about buying something, don’t just look at the price, think about the quality you may get from it, because it could bite you, if you don’t make the right decision.

One of the things that people look for, for differentiation in this marketplace, in the global marketplace right now, when you only have to look at a conventional brand name, you only have to look at someone like Heidelberg and say: What’s your biggest selling line right now in printing presses? You know what they say? A 6-colour press

I hear in India, that most people want to see their proofs on the substrate it is printed on. have a problem with that, and I don’t understand it. Let me just tell you quickly why, I am sure you can probably tell me at question time or afterwards that I am wrong and I’ll be happy to hear that.”

Because people are looking even in offset, for special colours and in flexo, the whole point of flexo is special colours.

Almost all of these proofers give CMYK. So there’s two approaches that we need to make- either the conversion from CMYK, which can be done via the application software, or a migration from digital proofers that we have now, to a new- er type will give us a wider colour gamut with more colours, in other words. Those just aren’t avail- able right now. Most digital proofers that offer spe cial colours, often offer them via the application software. So if you’re in the special colour mar- ket, just look out.

I hear in India, that most people want to see their proofs on the substrate it is printed on. I have a problem with that, and I don’t understand it. Let me just tell you quickly why, I am sure you can probably tell me at question time or afterwards that I am wrong and I’ll be happy to hear that.

Let me just ask you, when you get the proof you get at the moment-whether it’s from a flat- bed press, whether it’s from pre-press, whether it’s from the few digital proofers that are available in the marketplace right now-can you honestly tell me that every time you receive that proof, it is on the paper which you’ll eventually print on, made with the inks you’ll eventually print with, printed with the plates that will go on the printing press to make the job printed by the operator, that will print the final job. It doesn’t. So the argument going around that the proof is needed on the sub- strate I am going to print on, doesn’t really hold water, although I can understand it is partly an 1 emotional argument, particularly when you think about newsprint; if I want to see a job on news- print, why do you give it to me on nice, white, shin paper. So substrate, because many of these tech– nologies require a special substrate when there’s intense heat involved, when there’s ink involved that’s a water-based ink. To guarantee that repeat- ability that I said is the most important, you need to have, generally, special papers, now just be care- ful here, because there are some manufacturers who may claim: you buy my products, you can use anything. It is not true, and that’s including us, okay, so be careful of that.

If I were you, and someone said that, I’d say show me, here’s a group of papers, print a proof for me on all these papers and I’ll be happy. What I say, and what other people should say is, you are paying a lot of cash for this proofer, a lot of cash. Are you telling me that after making that invest- ment, you’re going to print a proof and save your- self a few cents? You’re not just buying the proof- er-I hope I don’t sound like a salesman, but when you buy a proofer, when you buy a car, when you buy anything, you don’t only buy the object; when you buy a digital proofer, you’re buying repeat- ability, you’re buying quality. You’re buying the proof that will be accepted by your client. So I ask you to look very carefully, about what you’re buy ing when you do take the decision, as indeed you would on any digital equipment that you’re look ing at at this seminar. It’s available now, the tech- nology is still, in my opinion, in it’s infancy as in- deed is computer-to-plate, but, it’s definitely com- ing. I would estimate that globally, no more than 5-10% of the proofs are made digitally. But it will grow, partly because technology is driving you. When computer-to-plate comes in, there is no film. How will you print the proof then? So it’s a min- efield. You need to get all the information you can, check it out thoroughly, deal with a reputable dealer, consider the problems you may see, have a marketing plan in place to consider things like the screen approach, 4-colour, as opposed to spe- cial colour, and it should work well for you.


Burjor Poonawala I think Mike has given us a very frank and candid presentation on proofing with- out holding back any information whether it is DuPont or anybody else. He has tried to restrict his talk only to the pros and cons of the various proofing systems available in the market today, and like he said everything comes for a price, so once you decide what price you can afford and what quality you want, you have to make up your own equations and make your own decisions and like he said your decisions should turn out right be- cause if they don’t, it will turn out to be an ex- pensive mistake. In this area we as pre-press peo- ple do face a lot of problems and one of the great- est problems is clients’ acceptability. Client edu- cation, I think in this regard, is absolutely neces- sary because a client, not being an expert on pre- press, proofing, printing or what have you, can’t really make up his mind so he prefers to play it safe and follow the traditional methods. Now, we need to find ways to get round this problem. The floor is now open to questions

Mike Webb I would like to make a comment on that last statement and that is acceptability. As I said in my talk, there’s no doubt that acceptabili- ty is the biggest hurdle overcome. There are several ways around that and each of them require a good understanding with your client, a change in thinking, because there are many areas to over- come around acceptability and it’s not just, well, there’s no halftone for you. Consider for instance when you put in a digital proofing machine, that it will output a proof for you, and if that proof costs you $20, two proofs would cost you $40, and three proofs would cost you $60. So if the client should quite often say to you: I’d like 25 proofs of this job, it’s going to be tough. But the experience I’ve seen from people in Asia so far and in the other countries that are putting digital proofering is amazing. Suddenly, it’s okay for one or two proofs, it’s just that they have to pay more. So I know it’s a concern right now, I know it’s a concern when people say: There’s no dots on it, I don’t like that. Let me just tell you how I would handle it. The way I would handle it is this, imagine this was a proof, what do you do with a proof, do you feel it, smell it, taste it? No, you look at it. Suppose I have an original and a digital proof here, suppose I have a conventional proof from a flatbed proof press and a digital proof here, the most critical thing about high quality digital proofing is that they are linked to a sophisticated and reliable co- lour management system. That means that they can be controlled to look like anything-that dig- ital signal can be output to look like newspaper, a flexo press, a gravure press, an offset press. So this one here, would be a stochastic screened dig- ital proof. This one here would be a convention- al halftone flatbed proof and they match and from where you’re sitting or even closer, you have got no chance at all of seeing which one is stochastic and which one is conventional. So if you assume that the way we handle a proof is to look at it, from a usual, normal, visual distance and we cannot see if it’s stochastic or halftone, what does it matter? Now these are just arguments I can tell you, that I use. You may say, Oh no, that won’t work. But there are ways to persuade and this is part of the marketing of the proofing systems-there are ways to persuade people to say. Hey, these are okay. these are good quality. Anyway, that’s just a com- ment to what some of you said. Acceptability is the biggest hurdle. It needs to be worked on, there are ways around it and arguments around it, oth- erwise, a digital proofer wouldn’t work, and in fact there are many in other countries.

Participant What is the best technology available for a FM screened digital proofer?

Mike Webb There is no doubt in my mind-prob ably people think I am biased-but there’s no doubt in my mind that the continuous flowing inkjet is the best available in the market right now. It is repeatable, it is reliable and it usually is ac companied with very sophisticated colour man- agement which is essential, otherwise, as I tried to make a distinction earlier on, the difference between a picture and a proof is that a proof has applied standards, it is a print quality that you can hand to a printer and he can print without strug- gling and it is recognised as a standard globally. Others, I would call them pictures; they show the job, they show the position of the picture, they show that the price is right, you can read the type, but you can’t okay it for colour. Only a proof is okay for colour, has print quality and is reliable.

Burjor Poonawala Whilst we are here in this forum, I would like to stress a point or two, and never mind advertising agencies having a problem ac- cepting digital proofs. We find in our business that even technology-oriented people such as printers also tend to reject digital proofs. They all want to see a halftone dot. They still want to see a set of progressive proofs. No digital proof or pre-press non-conventional proof can produce progressives Progressives were required in the days when one colour was printed at a time on a single colour machine. It was a guideline for the printer to fol low, because there was no densitometer to mea sure the amount of ink, they had to judge every thing visually and therefore progressive proofs were required. Today, in this modern age of dig ital technology, I think, all the pre-press houses are still producing progressive proofs, which no-body looks at and which nobody uses, nobody even tries to ‘taste’, or ‘feel’ anything, they are just go- ing into the dustbin. But we find that we are al- ways called upon to make progressive proofs, which nobody eventually looks at. Now I think, this is a trend we have to reverse and we should say that alright, one set of positives, one proof, whether it’s conventional or digital that suffices because 99% 4-colour printing, is done on a 4- colour machine. Nobody these days uses single or even two-colour machines and therefore, pro- gressive proofs are no longer required. So a digi. tal proof or a conventional proof what have you, a single proof is good enough to go along with a set of 4-colour positives.

Naresh Khanna I want to thank Burjor and Mike Webb and I think this session on digital proofing is perhaps the most historic session of this con- ference. Because, I don’t think ever before in our country, there has been a serious, technical or general discussion of proofing at all. The concept of proofing, which is the single most important step in quality and quality control. Unfortunate- ly, 99% of the printing in this subcontinent is done without any proof at all although this is the sim- plest visual methodology for establishing quality. There is nothing, that can be done in our indus- ury without simply understanding the concept of proofing. So we have jumped from having no dis- cussion of proofing, except that which is carried out over the years by people trying to sell their proofing technologies, to a discussion of digital proofing. This also shows the kind of leap-frog- ging that we are doing, very often without mas- tering the basics or going through the steps. We are forced by technology to look at things which are seemingly not realisable or viable.

Introduction of P.V. Manohar by Naresh Khanna

THE NEXT SESSION IS ON COLOUR MANAGEMENT. We have two speakers, Vir Bhanu and PV. Manohar. The first speaker is P.V. Manohar, who worked for The Hindu for eight years and for the last several years has worked with Afga in Bombay as a Technical Specialist. Mr. Manohar is also a Bachelor in Printing Engineering from Anna University, Madras and he has been the Market ing Manager for Agfa Graphic Products for the past two years.

P.V. Manohar on Colour Management

THE TOPIC THAT I WOULD BE SPEAKING on is Colour Management and Colour Standardization. In traditional repro, t the users were not aware of a colour management problem that needed to be solved. In traditional systems which were proprietary and closed, all the hardware and software that went into such con- figurations were selected by the vendor. So what the vendor would do is, he will identify an input and the output, and since he already knew the characteristics of the input and output, he will type to them and configure it as a system. So it was easy for the vendor as well as for the user to expect what colours they would be able to see, print or scan at every stage. So such a colour management prob lem did not exist in conventional systems that were supplied. Then came the Postscript, and now we live in a world of open systems. If you want a scan- ner, you have a big list to choose from and if you have a computer, you again have a list to choose from and then the same applies at the output end -you could choose the type of printer you want and the type of technology you want and because of the standards that exist today it is possible to tie all the systems and configure a workflow. You also have on the other hand, images being pro- duced on an increasing number of types of inputs -you have scanners today, you have digital cam- eras and you also have the CCD cameras and so on, so the number of ways by which to input are very many and you’ll also find that the destinations are also equally that many. You have colour print ers, digital printers, offset presses, and multime- dia and internet applications. The colour data is being moved between various market segments and the one-to-one situation that was existing ear- lier, is no longer tenable. When colour digital data is moved between the various components that form into the system, there is nothing in this op eration system which takes control of this and sees to it that the colour that you scan is exactly what it is going to look on the screen and finally the print that you get is going to look close to the screen. This colour management problem seems to be existing in the case of open PostScript sys tems and certain vendors have come up with some solutions and they have been refined over the years. The main function of the colour manage ment system is to see that colours remain constant and predictable throughout the whole process during scanning, when you are retouching as well as during outputting. The problems of colour management are said to be three-fold:

  • The first thing is differences in colour render- ing. Here, this refers to the fact that the way the colour is reproduced is different. Today we have an RGB model, we have a CMYK model, and the colours in the RGB model are differ- ent from those in the CMYK model due to the addition of ‘K’. And because there are differ- ences of colour rendering, that is going to be a problem for management of colour.
  • There are also differences in every reproduc- tion method. The CMYK that is done in USA is going to be different from the CMYK pro- cess that is used in Japan. So there are differ- ences in the CMYK process; every reproduction method has its own gamut and every reproduc tion device has its own gamut. For example, scanners from vendor A and B will have a dif ferent gamut-a gamut means the colours that the scanner can detect. The colours that a par- ticular scanner can detect would be different from the one supplied by another vendor. So this is the another problem which colour man- agement has to handle or manage.
  • The third thing is colours also change over a period of time. So what I actually get from a scanner today will not be same as I would get one year down the line. This is because of the way in which a scanner sees colour. The scanner sees colour through a light source, a filter- ing system and a detector -a sensor which could be a CCD or photo-multiplier tube. So all these three will finally tell you what the cor- rect colour is going to be like. So if the light source is going to age then there is going to be a difference in quality. The filter may also wear out and the CCD responses may also change over a period of time. So the change in time is also one of the problems which the colour management system must handle.
  • Now, we come to two situations wherein we have a scanner and a monitor. Both these use RGBs to create colours. They work on transmit- ted light and are based on the RGB models. As- suming I have a coloured surface which I have at the other end and put it on one of the scanners and scan it. When I scan it, because the scanner has got RGB filters, it’s going to give me RGB data and that RGB data when it is sent to the first screen, the monitor is going to display that red. The red that I am going to see on the screen is going to be quite different from the red of the colour sur- face, primarily because the red of the monitor is different from the red of the screen. If I send the same RGB specifications to another screen, it is quite probable that the red that is going to be seen on the second monitor is going to be different from that of the first monitor and also different from that of the original surface. And now we come to a problem where the colour we see on a scanner is very device-specific, it is only linked to that particular monitor. The colour that is com- ing out of the scanner has got to do only with the particular scanner that we are talking of. The true meaning of it is attached to the device where it is coming from. That is why we say that a scanner is device-specific-I mean it is device-dependent, a monitor is also device-dependent.

Assuming that I use the original colour sur face and put it on another scanner, then I am going to get a different RGB specification-may- be I’ll get red as 95, green as 5 and blue as 10. When I send that same data to the screen, the re- sult that I see on the screen is also going to vary. So what we can generally conclude is that these devices are device-dependent and the true mean- ing of colour is attached to the specific device which is used to detect or display.

Similarly when we come to the CMYK mod- el, assuming that a scan has been made and a CMYK conversion has been done, this CMYK spec- ification can either go to a digital printer or films can be made and it can go to a sheet-fed press and printed on coated paper, or it can also be used on a web press using newsprint. The colours that have come from same CMYK specifications, are going to be different depending upon the desti- nation. This again shows that the CMYK process I mean all the output devices-is again device-spe- cific and the colour that we see on these devices is tied to that particular device. So the problem that the colour management system should han- dle is, when it sends an RGB to a CMYK, it can- not just send it like that. What it has to do is, it has to transform and it has to adapt the colours for the particular output device it is sending, so that the colour faithfulness is maintained. So the job of colour management is to adapt and trans- form the colour values to suit the particular des- tination device.

Now we come to the colour spaces or colour gamuts. The colour gamut is the description of the range of colours that a particular device can detect, display or print. For example, here we see the RGB monitor has a bigger space than the CMYK space and when you are going to print some colours using the Pantone specifications, you get a different range. So what we see here is that the range of colours that for example, a monitor can display, is different from the range of colours the CMYK device can print and that Pantone-speci- fied colours can print. Since these gamuts do not match, one of the jobs of the colour management system is also to see that the ranges are com- pressed, so that gamut matching or gamut map- ping is done. So this is also one of the functions of a colour management system-gamut transfor- The main function of the colour management system is to see that colours remain constant and predictable throughout the whole process during scanning, when you are retouching as well as during outputting.”

Changes in time is the third problem we spoke about Assuming that a original surface is scanned and the specifications that we get from a scanner is a red on red and after a year, we get totally dif ferent specifications and these specifications could lead you to the light source ageing or the filter having wom out. So these are the changes in time and colour management must ensure that changes. in time are also accommodated in the calibration process. To accommodate this the colour man agement system must give you the tools for you to calibrate the devices, so that they are in align- ment and the faithfulness of colours is maintained.

Characterisation: One of the first steps, the colour management system should take is to char acterise the device. Characterising a device means, if you talk about scanner, this is the red, blue and green that the scanner can see as the maximum, so it is to identify the characteristics of a device and have it in the form of tag or profile, that it can be used when a transformation is done So one of the first functions is to identify the charac- teristics of the device. The characterisation is a sin gle action process. Characterisation generally can be done by the use of hardware and software. Char- acterisation is also supplied by a vendor at times, for example, if you buy a scanner you also get what is called an IT8 target. This IT8 target is a colour patch, is a colour piece available in re- flective and transmissive forms and they are rep resentative of all the colours that the human eye can see. So along with these original targets we also get what are called ‘digital files’. These are the files which have the spectrophotometric val- ues for all the patches that the IT8 carries. So the way it is done is you put these ITS targets on the scanner and then use the colour management software to scan it. you scan it, the scanner is going to see the colour patches on the IT8 tar- get in the specific way and the colour management system is going to compare the digital values that have been supplied by the factory and those that have come from the scanner and there is going to be a relationship built. So this is how charac terisation of a scanner is done and is a typical sin- gle action, meaning thereby that a scanner can vary over a period time, as I told you because of the light source or the filters being worn out. So the scanner needs to be calibrated. We put these ITS targets, scan it and make the characterisation What I am trying to say is that the characteris tion and calibration in the case of a scanner, is a single action process; the characterisation is no different from calibration. It is a typical single ac tion process. The result of this characterisation is a mathematical transformation chart. You have a chart wherein it details as to how the scanner sees colours and that information is given to the co- lour management system. This transformation chart can exist either in the peripheral equipment or in the workstation.

Assuming the first case, where this informa- tion is residing in the workstation and the colour X is what I want to finally print and this transfor mation chart is present in the workstations. When in an application you say that I want to print a colour ‘X’ then this disc command goes into the transformation chart. When the transformation chart sees that the colour ‘X’ is to be printed on an output device such as the printer, it knows that the gamut of the device is different and actually to get the colour ‘X’, I have to transfer it into a command and say print ‘Y because the gamuts of the devices are different. So actually the trans- formation chart would give the operating systems software the information to say that to get the co- lour ‘X’, you have to convert the command and say to printer, print the colour ‘Y, so that finally what you get is colour ‘X’ that you desired. So this is a workflow wherein this transformation chart is present in the workstation.

This transformation chart can also exist with the printer. So the workstation is going to give you only a command saying that print ‘X’ and it is the job of printer controller software to look at the command and say if I want to print ‘X’, then I have to convert it to a command and say print ‘Y. So these are the two types of workflows and there are colour management systems available in the mar ket which use either of these, or in certain cases both approaches are also implemented.

Now, after having known how to character ise the particular input device, we have seen in detail only the scanners. We’ll go to the monitor and the output device later. Now we will highlight the limitations of characterisation. Characteris tion as we saw, is supplied in most cases by the manufacturers. So if Agfa scanners are coming out of the factory, for every 100 scanners they will pick up a scanner and then characterise the particu lar scanner and that particular characterisation is going to be valid for the 100 scanners that come Out That’s how characterisation is done in a batch process. But you also know that in the manufac turing process, there is a variation from scanner to scanner and that is not yet accommodated. The local conditions also have to be taken care of. There are changes in relative humidity and tem peranure, and all these can have an impact on the colour that you are going to scan. So ideally char acterisation is something that should be done by the user and it should be done as many times as it is warranted. So this is first limitation of char- acterisation. Characterisation should not be some- thing that you should get from the vendor, but you should be having the tools for you to do it at your own site for a particular device that you are using. We have discussed the local variables and we’ve also seen that the changes in time have to be accommodated. So the tools for characterisa tion should be at the command of the user.

Now we come to another aspect of colour management called calibration. Assuming that the scanner behaves as the manufacturer intended it to, or according to a manufacturer’s specifications, then there is no need for calibration. But we saw that there can be changes because of local con- ditions, because of manufacturing variations and also as the scanner ages, and therefore you need to calibrate your equipment to always have it within the limits of the characterisation. So you need to know what is calibration.

Calibration of an imagesetter: The imageset ter is the output device whose behaviour can be linear or non-linear and this has to be taken in connection with the photo-chemical process as well as the activity of the developer. So as far as imag esetter characterisation is concerned, you will ac tually be sending a tone file which represents all the dot percentages. Then you will process it in standardised conditions, read the dot percentag es and then decide whether the output system needs to be calibrated or not. So this is how out- put devices are calibrated. In the case of monitors, calibration is not that easy as in the case of scan- ners. There are simple as well as professional ap- proaches available in the market. But one thing that has to be remembered is that the characteri sation that most monitor vendors supply, can be used but once you have used the brightness of the contrast of the screen, the characterisation is no longer valid. So there are other approaches such as the Radius Preswview as well as the Barco callbrator which primarily do not depend upon the characterisation but they do calibration. They continuously calibrate their monitors. We know that in the case of the scanner char- acterisation and calibration is a single step pro- cess and when we come to monitors there are var ious approaches used and the general approach is because they users are going to adjust the bright- ness of contrast, the monitor tags or the monitor specifications supplied by the vendor can no long er be used and so generally calibration is the sort of approach used when you are working with monitors. Another approach that is used in the market is characterisation and calibration used as two totally different processes. Characterisation is done at the workstation level, where the viewpoint is that the device will behave according to the char acterisation data, and the calibration is done at the output end, where it will fine tune the gener al characterisation. So various approaches have been used by various vendors that are available in the market and this sums it up.

So the basic function of the colour manage ment system is to make the colours uniform and predictable through all the phases of colour re- production. The other function is to provide in- dependence from the application. The colour management systems were originally available, tied up with a particular application. So in case you were using Photoshop, you could use only Efi-co- Jour and the same Efi-colour system may not be available for other applications, say PageMaker and so on. So actually it was more application-spe cific, when it was originally available. It was more of a proprietary approach. Later, these colour management systems were implemented at the op- erating system level, so that it became a standard operating feature and all the colour commands coming from various applications were answered at the operating system level. Output simulations is also among the functions of the colour man- agement system. You should be able to see on the screen what the final output device is going to look like. You should also be able to get an indication from the monitor as to what are the colours that fit within the gamut, what are the colours that fit out of the gamut and what is the gamut compres sion that is taking place; certain softwares also give you gamut warning or a gamut alarm wherein you get such indications. And then there is the uni- form file transfer. What that means is, suppose we are working in a work-group and the same file is 

“What the CIE system does is, it standardises the three components of colour as experienced by the human nervous system. The three components of colour are the source, the object and the observer. These three things were standardised by the CIE.””

being accessed by various members in the work group. The colours that each of them see on their screen should also be consistent. So colour man- agement supports this as a standard function and some colour management systems also provide colour separation as a standard feature. So the RGB to CMYK is something that is done as part of the colour management system software.

To make all this work, the colour manage ment system will need to specify colour in a de- vice-independent way. The colour specification should not be tied to a particular device, it should be independent of peculiarities of the devices. The CIE system is one such system, which is the inter national standard for colour and illumination. What the CIE system does is, it standardises the three components of colour as experienced by the human nervous system. The three components of colour are the source, the object and the observ er. These three things were standardised by the CIE. What they did is they picked up the data about the colour perception of various observers with normal colour vision and used this colour data to define what is known as a ‘standard observer They also standardised from a few light sources. So once they standardise the observer as well as the light sources, then they use some mathemati cal magic and arrive at three primary numbers or three primary sources, ‘X, Y, Z, which when com- bined at various levels, will be able to produce all the colours that the human eye can see, and this was called the CIE-XYZ system, which in fact forms the basis of most of the colour management sys- tems that are available in the market. So once the colour model, that is the CIE system has been iden- tified, then you have to go about characterising the devices in a workflow. So you have to charac- and calibrate your monitor, your scanner and your output devices. Thereafter, once you’ve characterised you can go about doing the colour transformation wherein you would want a scan- ner to be connected and transformed to moni tor space and the monitor space to an output de vice. The third function is colour adjustment or colour transformation. So this is the characterisation that I talked of, this is the ANSI ITS target and this target has g 264 colour patches and are the patches rep resentive of the colours that the human eye can see. The way it is arranged is on the vertical di rection you have the various hue angles and on the horizontal axis you have three lightness les els and for every lightness level, you have four chruma level So in all, these are 264 colour patch es and you also have some area where you can put up a original picture that you can see or you can also put up some colours that you want to track in your process. So some vacant area is also left for you to put in some colours. So this IT8 target is a standard which has been formed by the ITS commission in collaboration with Agfa, Kodak, Fuji and Konika so that there is compatibility of the film products that are supplied by these four people. These targets are available in reflective and transmissive form; reflective comes in 5×7 and transmissive comes in 4×5 transmissive and 35 mm transmissive. Along with these targets, you also get the digital file which will have the spectrophoto- metric values that correspond to this target. This is what a user should use to scan and compare the values that the scanner has seen with the digital files and create the characterisation.

So in the case of a monitor, the very basic approach is to use the characteristics that are sup plied by the monitor vendor, but then you will have very little visual control because you are also going to adjust the brightness and contrast norma and the approach is to go for a professional sys tem, that is from Radius, Rasterops or the Barco Calibrator, wherein you will have a combination of hardware and software, you are going to use a suction cup, the suction cup will attach to the monitor and then the video card is going to send in a few colours, and the suction cup is going to detect what the video card has sent and then call- brate the video card. So there are close loop tems that is also available in the market. So if you want total visual control, you should go in for such a product. Then you will also have to character ise your imagesetter. Since imagesetters produce black and white film, it is not possible for you characterise an imagesetter from a black and wh film, so you have to include the print process also So the tag for the imagesetter that you will see, is a CMYK. It is taken from the point of films to the point of printing and entire set is characterised so the output tag of an imagesetter would be a CMYK specification. I was saying that the scanners and monitors are device-dependent, and to make them device independent, you have to move through some standard colour spaces, and that was CIE. So when I move from scanner space to the device-indepen dent space, then there is a colour conversion in that process and then when I go to a particular monitor, then I also have to do a colour conver- sion. And these colour conversions are mathematics which will need quite a bit of computational power. To make things easy what certain colour management systems do is, instead of going, say from English to French, I go from English to Span- ish and then Spanish to French, what certain co- lour management vendors do is, once you know that you are going to use an Agfa scanner and an Apple monitor, why not tie it up, it is basic math- ematics. It is tied up, so that two tags or two pro- files are taken and formed as a link so that what in effect you see is only a single transformation process, that is going to take half the computation time that you would use in going through the CIE space. So this is one approach that is used by cer- tain colour management system solutions in the market.

I was telling you that the gamuts of the devic- es that are going to be used in the transformation are not the same. For example, the gamut of a CMYK process was different from that of the RGB monitor or an RGB scanner. So I have to match my gamuts. When I go from a scanner to a par- ticular monitor and the algorithms that are be ing used in this transformation process as well as the matching process are different. So one such approach is the calorimetric matching. There are two such algorithms, or at this moment I think there are three algorithms.

For the algorithm that is going to handle the gamut compression, there are various approach- es that are available. Calorimetric matching is the approach wherein the colour that falls, say, out- side is being compressed to the nearest possible colour that can be achieved and the colours that already fall within the gamut are untouched. This is calorimetric matching and this is used when you are printing some business graphics. The colours are brought down to the maximum saturation that is possible. The other approach is perceptual matching. In the case of perceptual matching, the colours that fall outside the gamut are brought inside, and the colours that were existing within the gamut are also compressed to a certain extent. This way the original colour differences that were existing in a scan are still maintained. This is the approach that is used in the case of scanned pic tures The quality of these algorithms is going to determine the quality and complexity of the co lour management system. This is a key factor. I have already told you that the colour management system was originally application-specific and that all the vendors were taking a proprietary approach and this could not take anybody very far. So cer tain standards have to fall in place, as Apple in 1993-94, came with a product known as ‘Col orSync ColorSync is analagous to what ATM was doing in the Mac environment. This ATM was something that was available as a system resource and it used to answer all the font calls that come and then display it properly on the monitor. The way this font handling was done on the Macintosh, was at the operating system level. So an applica tion developer, for example, a QuarkXpress or a PageMaker need not bother about fonts. You only have to incorporate the commands for the font call and thereafter once it says, I want an Helveti- ca 12 font, 12 point type, then that is answered by the operating system and it gets back the font that it wants. So font handing was an operating system function on the Macintosh. A similar ap proach was also followed for colour, and then they came up with ColorSync which gives you basic colour management functionality. This was also an operating system resource and it was able to answer all the colour commands coming from various applications. Photoshop can ask that I want this RGB to convert to this CMYK, and ColorSync would be able to handle this colour command. All that Photoshop needs to do is incorporate a few commands in the process of developing the application and then the operating system would take care of it. When colour management becomes an operating system feature, then the colours that you see in Illustrator are going to be same as you see in Photoshop as well as QuarkXpress. So that was a great step that really helped in standardis- ing the management of colour. But the way Ap ple implemented it, it was not really a full func- tionality. It was only the basic framework, mean- ing that the conversion for the algorithms that Apple implemented were without putting excess load on memory or RAM. At the time it was very basic and a simple transformation that was used. It also gave the other OEMs the opportunity to include their transformations, which were much more powerful and accurate to be used as OEMs. So what Agfa does or what any other vendor would 

“It is now a race and challenge for everybody. So whoever has got the better algorithms, is going to be preferred and their colour management system is going to be used.”

do is, they will have the colour transformations and then they will plug-in into the ColorSync. Of course, using these profiles that are supplied by third parties, would give you more accurate co- Jour.

Basically, ColorSync will give you colour cor respondence, so that you can translate colours from various colour spaces and then on the mon- itor screen it will also be able to give you an indi- cation as to what the final output is going to look like, as well as give you the gamut alarm, that will give you an indication as to what are the colours that are going to print and what are the colours that are going to fall out of the gamut. In terms of implementation, ColorSync is a memory sav ing feature, it does not load the operating system, and a simple mathematical transformation is used. With that simple transformation, you will not be able to get the quality that you will need for de manding repro and you have to go through cer tain third-party OEMs which used to plug-in into the ColorSync as profiles. This is the same situa- tion as in the case of fonts. So there could be hun- dreds of font developers, all of them could incor- porate the fonts and it is the users who have to decide as to which fonts they want to use depend- ing upon the quality of the fonts. So we are now in a similar situation, it is now a race and challenge for everybody. So whoever has got the better al- gorithms, is going to be preferred and their co- lour management system is going to be used.

Adobe supplied PostScript Level-2 with a mechanism which will support various colour spac es and these various colour spaces were device- independent as well as device-dependent. I told you that the reference model that is used is the CIE-XYZ. The CIE-ABC is a model supported by PostScript Level 2 and it is nothing but a trans- formation of the CIE-XYZ. From the CIE-ABC you can also apply some mathematics and go further into various derivatives. Such derivatives are also supported by PostScript Level-2 and the CIE-A is another model that is supported, which is an Ach- romatic version which you can use for black and white management also. PostScript Level-2, not only supports the device-independent models, but also the device-dependent models such as the  RGBs, the CMYKs and so on. Suppose an application, such as Photoshop, describes the particular picture in the form of a CIE space, and then that page is sent to a RIP. In the RIP there is an area called specification dic tionary which will convert any space that it sees Adobe calls it into an XYZ space. This XYZ space is then sent to a colour rendering dictionary- colour rendering dictionary. The job of a colour rendering dictionary is to convert this device-in- dependent data back to device-dependent data Because, now, you also have to send it to an out- put device which has got device-dependent char acteristics. The job of a colour rendering dictio nary is to convert device-independent data to de- vice-dependent data. An XYZ can be converted into an RGB and then sent into a film recorder. The XYZ data can be converted into CMYK for serving to an offset press, or the XYZ data be sent for the purpose it is intended to. Then the CMYK data has to be sent to the RIP and imagesetter combination, and the imagesetter has to be tak en with a film into account. They can behave in a linear and non-linear way, for which it has to be characterised and the transfer function for the characterisation would be a existing on the RIP. This CMYK data would go through the transfer function. The data gets linearised in this process and then it is exposed. This is the detailed expla nation of the workflow in a colour management system. 

We’ve talked about ColorSync and also said that ColorSync is not yet powerful or it cannot give you the quality for demanding colour and you also have some third parties who can plug in and plug- in into the operating system and also provide bet ter support which is called the application colour management system. It is not always possible for vendors to be able to give solutions for every soft- ware that is available in the market. For example, in the case of Agfa, there is a plug-in in the case of Photoshop and there is extension in the case of QuarkXpress. But for other applications, the real plug-in does not exist; for example, Illustra tor is used, that defines a colour, it can go through the ColorSync which can still reach the applica tion colour management system that has been supplied and make use of the powerful features and the better algorithms that are associated with it. Photoshop and QuarkXpress, for which plug- ins and extension exist, can go directly to the ap plication colour management system and make use of all the functionalities. So the basic frame work is supported by ColorSync. For you to get better quality, you always have to use a specific application for this purpose. This is a situation where we say that scanners are all RGB devices but an application colour management can give you the facility by which it can do some colour com puter functions. Colour computer function is the RGB to CMYK conversions. A scanner in combi- nation with application colour management sys- tem could convert the RGB of the scanner to a direct CMYK. Alternatively, this application colour management system can also convert it to XYZ data. All this data can be incorporated into an application and sent to the RIP. Depending upon whether it’s RGB, CMYK or XYZ, you have the various modules that are present in the RIP which would handle the colour conversion and then it will also go through the transfer function of the imagesetter and then it is exposed.

We were talking about colour rendering dic- tionaries which are going to give you the colour conversion from device-independent to device- dependent data and this colour rendering dictio- nary is the place where the other OEMs are go- ing to plug-in through their extensions, their bet ter algorithms.

If all the devices that we were using for ex- ample, a scanner, a monitor and an output device are working as per specifications, then there is no need for calibration. But we also saw that there are so many problems associated with colour man- agement and therefore you will also need to cali- brate the various components on a regular basis. So a colour management system must also have the tools for you to do regular calibration and when you are talking about calibration at the out- put end, you also have to take the photochemical process into account, you would have to know whether the film is behaving in a linear way or a non-linear way and whether the recorder is also responding in a linear or non-linear way. By a lin- ear or non-linear recorder, I mean that if a 50% digital value is sent, whether what we get from the film and imagesetter combination is 50% or more than that. If it is more then you have to calibrate it and then incorporate that particular tag into the RIP for it to be compensated.

So to sum up, the colour management sys- tem is not the answer to a total quality assurance system, but it forms a very important part in the set up of a quality assurance system. And even if you are going to invest very heavily in hardware and software and set up an elaborate colour man- agement system, it will fail if you don’t really set up your presses properly, calibrate then and run then within the specifications that you send. A lot of care has to be taken to standardise the print process-you should take a fingerprint of the process by printing a test form, identify the densities, identify the gray balance, end points, and also the font, find out what colour correc tions you need to do, and then get this feedback and the tone curves and then feed in this infor- mation and then use it in the scanning process, for you to really work in a standardised environ- ment.

So it is not only characterisation and calibra- tion, but a lot of care has to be taken in terms of monitoring and tracking the entire process. And in case you are going to change a plate or a blan– ket, you will have to identify as to the effect the change of plate or blanket will have got on the calibration and you might have to recalibrate that particular output.

Introduction of Vir Bhanu by Burjor Poonawala

I NEVER REALISED THAT COLOUR MANAGEMENT- although we know we need colour manage- ment-was so complex. The question that I would like to ask Manohar later is who is going to do it for us? Now, it is my pleasure to introduce the sec- ond speaker in this session and after his presen- tation, we will accept some questions. Mr. Vir Bhanu is publishing, Multimedia, Development Media Manager. He has 13 years experience in Graphic Arts, in India and abroad. He has worked with Apple India since 1994, and has been responsible for publishing, entertainment and New Me- dia markets for the SAARC region. He is a B.Tech. Engineer in printing and information systems from the Rochester Institute of Technology, New York. He has done his MBA from West Chester University in Pennsylvania and he holds a diplo- ma in printing from the Northern Institute of Printing Technology in Allahabad. He is a very young man and he has absorbed a lot of knowl- edge through his varied experiences and let us see what he has to present to us today.

THANK YOU, MR POONAWALLA, MR KHANNA and friends, good afternoon. I would also like to thank Manohar, for discussing some of the issues already. We can also discuss the prevailing situation in India in terms of colour management. What is the approach different processing houses have tak en so far? I have some experience with a few pro- cessing houses that I visited during the last one year and in fact before the presentation, I made a call to six processing houses, just to get an up- date including Pragati, Hyderabad and IPP. Del hi just to name a few. I called about six process houses and I got some information from them about what they are using right now. And then we would discuss what tools are available right now world-wide, and also a need for standardisation, which Manohar has very strongly advocated-that we no longer can have standards which are actu- ally not standards. We cannot work on systems which are proprietary


The evolution of colour-you know, we are experiencing that colour is becoming more and more important, we are seeing more colour on the newspapers, even on the local editions and in smaller cities as well. Also, colour is becoming very important on the Internet. Colour is not only an alog anymore, colour is becoming more and more digital. So, the importance of colour is increasing, the usage of colour is increasing and we are see ing more and more colour in our life. That is where probably the importance of colour lies- standards and how we communicate colour is becoming fundamental.

As Manohar discussed in his presentation, devices are different, and it is not only that we have different devices, but they talk different languag es-monitors speak RGB language, scanners also speak RGB language; on the other side we have printing machines, may be digital proofers, we have imagesetters, they speak a language which is CMYK, so the gamut is different in this case. And then from scanner to scanner we can have a dif- ference as well. So right now, maybe a good anal- ogy is a country like India, where somebody from Kashmir speaks Kashmiri and in Kerala people are speaking Malayalam and how do they communi- cate? It may not be the ideal way but English is a language that is becoming a common standard. whether it is a professional forum or whatever. So English is something where somebody thinks in his local thought-processing and comes back to his standard way of communication and the oth er person also understands that and they can com municate very well. So the need is basically to evolve a standard

I was talking about the evolution of colour and in the colour trends what we see is that quite a significant portion of colour is moving on alter- native media. That could be a CD-ROM, a mult media kiosk as I have seen at Bombay Airport, on the Internet. Another thing is that the people who create the alternative media, the content on the alternative media, they come actually from the creative community that traditionally has been associated with prepress, design, graphics and printers. The same people are doing this content development and their challenge is that now it is not only chasing colour on the traditional paper, but beyond that, on the alternative media as well. Now looking at the Indian scenario, my experi ence is, as I checked with a lot of processing houses, we all understand that there is a requirement for colour management, but we are all using propri etary colour management because standard co- lour management maybe in its infancy and nobody is seeing a solution and how to implement it. As Mr Poonawalla mentioned, we need someone who can implement it here. So in an in-house work- flow, you can have proprietary colour manage- ment. You have better control because you know all the devices, you know all the characteristics of your devices. However, what is happening is that the workflow is becoming more open in prepress or we are moving towards that. You are getting information from your advertising agency at times In the case of IPP, they told me that they are get- ting CDs from some of the multinational clients and those CDs are coming from advertising agen cies in New York or California. And then how do we make sure that the same colour which the cre ative artist perceived in Los Angeles or New York. is printed when it is printed in a printing press in So these are the issues we have. Now in the open workflow we do not have control on all the devices. From advertising agencies the job moves to a publisher and then maybe to process ing and then finally it is printed. Now, the advertising agency might have started with the same file but actually when you see at the end of the day if the sume advertisement has gone to some twelve editions all throughout the country in different newspapers, maybe in different magazines, the end result is different. Why? Because proba- bly we do not have standards. So open workflow requires standards. The current prepress situation here is that with the new workflow dynamics, where you do not have complete control the job is com- ing from your advertising agency in Delhi or Bom- bay and the processing house is somewhere else. The work-flow dynamics are not giving you enough control on the job. You have less control and you have to communicate again whatever you are doing, maybe from advertising agency to pro- cessing house, from processing house back to the advertising agency and at times to the customer. These are some of the complexities of open work- flow. I think that this is something we discussed and Manohar also put some light on that, that we have different devices and no prevailing colour management standard. This slide actually elabo- rates that a little better-we have scanners on this side, we have an imagesetter on the right hand side and maybe a digital proofer at the bottom. And the languages they are talking about, these languages are different because it’s RGB in the case of scanners and monitors and CMYK in the case of digital proofers as well as printing machines or imagesetters. So earlier, in the proprietary sys- tem, the same vendor was making translators be- cause they knew the characteristics of the devic- es. It was a turnkey system, however, in open systems that is not possible. We have a scanner from Agfa, maybe a scanner from probably two dozen other companies which are important ones, very well-established all throughout the world, which are making scanners and the same is the case with digital proofers and imagesetters as well. In a open system, it is total chaos and the solution to that is to look at the industry requirement. We need to communicate colour. It should be a common lan guage, it should be device-independent, it should also be vendor-independent so that you are not tied to a vendor solution. Also at the same time, if you have a solution that is standard but vendors are not supporting it, then again you are in a soup. So these are some of the industry requirements I have laid down, that’s where ColorSync comes in as a standard. Now, ColorSync is based on CIELAB CIELAB is a standard that was set, if I am correct, sometime in the 1950s to communi- cate colour. Now, as the communication of colour in the digital world is more complex with new devices, there was another committee which was formed, known as the International Colour Con- sortium (ICC). It was formed in 1993 and the char- ter of this committee was to communicate digital colour on CIELAB standards. This is the concept of the complete open system’s colour-that at the heart you have a standard CIELAB-based system which all the devices understand, and all the de- vices have their profile. That profile would tell the standard common language, what is the trait of this particular device. I think this is the heart of the whole thing. The challenge is to implement this. We all understand that in theory, if there is something that can work like this we are in busi- ness. However, the challenge is implementing it. Now, these are some of the members of the In- ternational Colour Consortium. Apple is a founder-member and took this concept to the open forum and became one of the members, so that it becomes the industry standard. There are a lot of companies who are part of this consor- tium, so there are a lot of vendors who are sup- porting this CIELAB-based ColorSync system.

We understand that yes, that is the theory. Now how do we implement it? Here are some of the tools. I think this is something important, there are a lot of different companies which are representing it. What Apple did was, it created a framework and Apple’s objective was not to go beyond that, because we all understand that dif ferent vendors have different devices. Apple cre- ated a basic framework which they could build

Now, the printers are smart. When they submit the final invoice to advertising agency, they will increase the yellow there and for the other advertisements they manipulate the best proofs and they will submit them to the advertising agency.””

upon. These are some of the third parties who are delivering ColorSync-based solutions as of today. It is a relatively mature technology, again, you know if I have to put it as Mike put in his numbers, percentage of usage, what people are using worldwide, a common CIELAB-based, ICC based colour management system. I think, prob ably that number is again between 5-10%, not more than that. In India, it is funny. It’s easy for a small design shop to implement it, where, maybe they have one scanner and one computer work. station. But the problem is that the people who are doing it, they don’t have all the pieces and people who have all the pieces, for them it is a little intimidating right now, I think maybe over a period of time we will see that the adaptability of this will increase in India. These are some of the solutions which are available right now. In fact, there is a brochure in the Apple stall, that gives you addresses, if not addresses maybe phone numbers of these vendors, and some of these vendors are represented in India.


So I think, we can start with this. Also, an- other thing which is important, unless vendors support it, you are not going to have a complete system. If reliability or usage of Photoshop is so important in the entire open workflow and if Photoshop does not support it, you are basically nowhere. Now the major programmes, they are supporting this ICC-based open standard right now, there is a plug-in available for Photoshop. I think you should note down this site address. There is an Internet site called www. ColorSync.apple.com From this site you can download a free plug-in for Photoshop, and I’ll tell you the functionality of that plug-in when 1 finish my presentation. Similarly, QuarkXpress has extensions, with these extensions you can support it, the address of those extensions is giv en in that brochure. Similarly, PageMaker is sup- porting this, Live Picture has been supporting it for quite some time. So major application ven- dors, major scanner and imagesetter vendors like Agfa and Linotype-Hell they are supporting this ICC-based open colour.

As Manohar said, 100 scanners are coming and for 100 scanners, they will one pro- file and they will give it to you. Yes, that will work. The percentage of accuracy might be any where between 90-95% plus or minus. There is a significant percentage of colour jobs we do in India where customers does not bother about the quality. The customer says give me the best reward for my money, give me the best price and that’s it. But there are a lot of customers- mean Burjor will agree with me on this-who are really conscious about quality, and they would pay you the premium to implement-the high quality standards in their final job. So, if you are addressing this kind of client base, then for you it becomes critical that every device that you have in your process house, is calibrated and for ev ery device, you have your own ICC profile. Actu ally, the whole summary is this-for every de- vice, if you can create a perfect ICC profile, you are in business, and depending upon the criti- cality of your quality, you can calibrate again and again, six months later you can calibrate again. And I am telling you there are tools available which I would talk about here again, and they are also listed in our brochure. This is a suction cup-based monitor calibrator. There are some of the vendors which I have listed down here. If you contact these vendors, they give you com- plete tools to implement this kind of colour man- agement system. So you test an uncalibrated mon- itor, you measure the response-what Manohar was saying ‘X’ Y, and then create a monitor profile, and tell the system to use that monitor profile. Similarly, you have to do the same with the scanner, using the IT8 target-which Mano- har was talking about- you can create a scan- ner profile and these are some of the vendors who are giving you tools to calibrate a scanner. Because they are really open systems tools, you could take Linotype-Hell Scanopen and in theo- ry you can actually calibrate a Agfa scanner. It should work. Some other company names are mentioned there-Colour Synergy, Colour Blind, Kodak, IPP, Logo, etc.

Similarly, you have to do the same with the printer. You measure the uncalibrated printer, you measure the response of the printer, and then you create a proof profile. I meet newspa per people often, and if you see the colour sup plements in newspapers like the Times of India, Brand Equity, etc., we have these problems-you have at times three or four different advertise- ments on the same double spread and maybe one advertisement requires more yellow and what do you do, if you put more yellow there, you are going to spoil the other three advertisements and it happens. Now, the printers are smart. When they submit the final invoice to advertising agen cy, they will increase the yellow there and for the other advertisements they manipulate the best proofs and they will submit them to the advertis ing agency. Otherwise, the advertising agency maybe would put a penalty and only pay 70%. It happens. You know some of the US printers, they say that they loose so much revenue because ad vertising agencies just don’t pay them 100% of the money. They say that we had given you an advertisement that was to go on 24 editions of the same newspaper everywhere like USA Today and the Chigago edition looks absolutely differ- ent than the Detroit edition. What is this? So I am not going to pay you. So you will get only 70% of the total money. 30% revenue lost, that’s a lot of money and I am serious, you know, we will have the same problem here. As you see the advertisement-spend increasing in India, you will have customers who would demand this kind of repeatability on every press. Now, again in theo- ry, it is very easy for me to say, however, to im- plement this is a real challenge and what it takes, is that you have to be convinced that this is the way to go, yes let me try this and only then you can implement this kind of a system, where what you can do once you have created profiles for a scanner, an imagesetter, and a monitor, you can do the same with a printing press. How- ever, there are a whole bunch of variables in- between- there are there are films, and if you change any of the variables including the temperature, that might lead us nowhere. How- ever, if you have most of the things constant-if you are are using the same plate, the same film and you take a fingerprint of the press using the same target, all you have to do is print a sheet on a press which is consistent and where the density of yellow, magenta, cyan is standard. And then you actually read all these 264 numbers and cre- ate a profile on your workstation, and then when you print an advertisement, you say that this ad- vertisement is for this press, for this printer, and you have the profile and then you have another profile of another press of another printer, and using this workflow, if you print on two different processes, or maybe twelve editions of twelve dif- ferent newspapers, we are likely to achieve bet- ter results than we would achieve otherwise. So that is the whole concept.

This is the workflow of the press that I just talked about. Test on an uncalibrated printing press, measure the response of the printing press and then create a printing press profile. I think what it requires is determination, that you be- lieve in this, this should work, let’s do it, or let’s start somewhere. I think here probably vendors have to support more customers in this particu- lar area and tell us what exactly you want, proba- bly we can do certain things-I think here mag- azines like Indian Printer & Publisher, Naresh Khan- na, and some of the printing associations, may- be the Printers’ Federation, we can take up this collective charter with the vendors, that let’s try a pilot project somewhere, all we need is your support and if you are convinced by the result only then you follow, you don’t have to follow otherwise.


Burjor Poonawalla Thank you very much for a splendid talk and I am sure there are questions.

Sunil Khullar This is for Manohar. One slight clar- ification, in one or two slides, especially in imag- esetter calibration, what you showed was CMYK as an input. Were you trying to tell us that we should calibrate an imagesetter separately for CMYK? Was that the intention?

P.V. Manohar No. Suppose you make up a page and CMYK data is part of the page and that is being sent to the imagesetter. On the imagesetter you also have a photographic material so you have to have it first calibrated. To have it calibrated, I mean, say, from a workstation, you send 50%, what you get out of the exposure as well as the devel- opment process should be 50%. So basic calibra- tion be done there. The result of the cali- bration of linearisation is a transfer-curve and that transfer-curve will be sitting on the RIP. The CMYK data that you send from the PageMaker will have to go through this transfer-curve before it is im- aged. That is what it means.

Sunil Khullar Should there be one curve or should there be four curves?

P.V. Manohar There could be as many curves as the resolutions that you are working, the dot shapes as well as the angles.

Sunil Khullar So for 133 dpi, round dot, do you recommend four different curves for four angles?

P.F. Manohar No, actually theoretically speaking, calibrations will differ depending upon the reso- lutions of the imagesetter, depending upon the dot shape, as well as the angles that are used. So basically you should work with a set wherein you take the most popular round dot and eleptical dot and work it at a few resolutions and then use that standard set. In practice the variations are mini- mum. Theoretically speaking, there is a difference but in practice, there is less.

Sunil Khullar I agree, as far as resolution goes, as far as dot shapes go and frequency goes, but an- gle is something any decent RIP should deliver 40% same in all the four angles.

P.V. Manohar Angle is something that we can- not talk about as a separate issue. Angle is to be discussed with respect to the screening technol- ogy that is being used. Depending upon the screening technology that is used, the angles are not always as you want. A best approximation is being used in certain implementations like Su percell screening as well as the Adobe screen- ing. So in certain cases it has been found that there have been slight variations. So calibration should help although in practice, you need not go into the details. You can still work with a few calibrations.

Sunil Khullar Most of the RIP calibrations also don’t allow you to specify angles in the calibration al- gorithm. They have only one.

P.V. Manohar Yes, that is why the background he hind that is, not all the calibration softwares sup port extensive features, but a few do.

Participant Which are the operating systems sup porting the ICC-based open colour system?

Vir Bhanu To answer your question, right now it’s only MacOS. Macintosh is the only operating 

We all understand that there is a requirement for colour management, but we are all using proprietary colour management because standard colour management maybe in its infancy and nobody is seeing a solution and how to implement it.””

system, that has OS-level support, However, what is encouraging is that other leading industry play- ers like SGI, Eastman Kodak and even Microsoft have joined this Consortium. I tell you it really take lot of effort before you can have a technolo- gy that can be commercially available. Apple has been working on this technology now for over five years, the first version of ColorSync 1.2 was launched about three years ago. The new ver sion 2.2 was launched about six to eight months ago and even today, in India, we do not have probably even a single processing house that implements ColorSync-based colour manage- ment solution fully, I do not know, I will be happy to know if somebody is implementing it. So it really takes a lot of time for a technol- ogy to become commercially available first and then for users to evaluate it and then imple- ment it in their places. So, although other in- dustry players have agreed to this standard, I think it will really take two or three years be fore you see this technology coming on Win- dows or other platforms. It can come little ear. lier as well.

Participant Now we all know it’s a very basic ques- tion, RGB and CMYK are complimentary colours. RGB+CMYK 1, that is what complimentary says. So, can I say, CMYK 1-RGB, is the formula cor- rect? This is a very basic question. Now people say red is a complimentary of cyan, blue is a com- plementary of yellow and green is a complimen- tary of majenta. Complimentary means if you add together two, you should get one. RGB and CMYK are complimentary, which means RGB + CMYK = 1. So if I want to convert from RGB to CMYK, does the simple formula hold good-RGB = 1- CMYK or is it the other way round–CMYK = 1- RGB2

P.V. Manohar RGB and CMY are complementa ry, but RGB to CMYK is not complimentary. The addition of ‘K’ to CMY would make a lot of dif- ference-so if you are talking CMY, it works the way you think, but the addition of ‘K’ would make a lot of difference.

error: Content is protected !!

Price Based Country test mode enabled for testing India. You should do tests on private browsing mode. Browse in private with Firefox, Chrome and Safari