The color components
of the printed page: the inks

What is a good copy transparency
for reproduction?

Compressing the tonal range to 1.9

How to compress the original tonal range?

Controlling illumination without the hint of a glare

The photographic light source

Calibrating a film emulsion

Sizing up the grain structure

What brand of copy transparency can be used?

What if the copy transparency is off?

Is the painting upside down?

Color proofing: a communication tool

The press is on the run

The 5000 degree Kelvin story

Proper use of the grey scale

Leave nothing to chance

COLOR REPRODUCTION
The color of original artwork on the printed page.

Reproducing paintings and other two-dimensional original color artwork on the printed page poses many challenges for those involved in every step of the process. Artists, designers, photographers, pre-press and press professionals all play important roles along the way, but they sometimes do so without a shared vocabulary or thorough understanding of reproduction techniques. Such a wide range of variables are involved during the reproduction of a painting in print that decisions regarding every stage of the process must be made by people who are truly knowledgeable about the effects and the content alterations caused by transfer into another medium.

This bulletin has been prepared to establish some common technical and communications grounds and to provide guidelines to help to understand the process.

No magic formula can replace a thorough assessment of the original art itself, since every medium "translates" differently into print. Similarly, a look at the printed page can help us to understand both its limitations and its possibilities. From this perspective, a strategy can be developed to remain inside the window of performance of the process. This is called calibrating to the printing requirement.

The color components of the printed page: the inks
Cyan, magenta, yellow and black (C.M.Y.K.) are the key inks in modern, full-color, offset lithography. These inks work mainly as filters, using the whiteness of paper as a reflecting surface. By combining and superimposing a matrix of micro-spots of ink, these ink films or filters record pictorial content and render shifts in density, contrast and hue into a precise, but limited, color spectrum. The inks are made primarily of ground color pigments whose purity and vehicle is variable and subject to some degree of color contamination. An ink set (C.M.Y.K.) differs from batch to batch and from printer to printer, and behaves differently from press to press and from one paper to another. For this reason, each batch has to be tested to determine its particular characteristics. One such test, an ink gamut test, allows a printer to evaluate the range of color of an ink set on a specific paper substrate.

Understandably, the infinite possibility of the artist's palette cannot be duplicated through a C.M.Y.K. ink set. Some hues of the original work of art may not be reproducible through a specific ink set. Sometimes a fifth or sixth color is necessary. After all, even a Pantone® color swatch book has nine different pigments to secure its full color range.

Professional photographers are used to this limitation, since each film batch comes with the manufacturer's recommended adjustment to color/sensitivity of the emulsion. The Fuji Company, for example, indicates all the necessary emulsion characteristics on the outside of their packaging — a courtesy that we wish others would extend! Unfortunately, the ink industry has some catching up to do. For the most part, basic technical information, such as batch numbers, hue shift of an ink set, is rarely provided. Just as it is not possible to fabricate a color film emulsion with absolute consistency, an ink set suffers to some degree. These variations can and should be monitored to ensure adequate color management.

In their very valuable newsletters, Info 30 and Info 31, Sinar®, the prestigious Swiss photographic equipment specialists, suggest that ink black is far from photographic black. They confirm that a transparency, designed for printing, should have a density range of close to 1.8. Some presses, mainly sheet-fed, may vary up to a highest possible density of 1.9 (paper white measuring 0). To simplify discussion, we will use the 1.9 reference in this bulletin.

Following Kodak's guidelines and recommendations, a color bar in your transparency — in combination with a control grey scale such as Q14 Kodak, a neutral, 14-step grey scale — enables compensation for some of the idiosyncrasies of film emulsion and ink sets to secure closer color interpretation.

Back to top

What is a good copy transparency for reproduction?
The well exposed and processed transparency has a highlight placement in "A" patch (Q14 Kodak scale) of .25 (+/- .05) and a density range that is close to 1.9. In other words, it is not a great looking 3.5 density range, brilliant, saturated, rich and contrasty transparency. Sorry!

According to the main structure of any color separation destined for the modern offset press, the "A" patch is the edge between specular and diffuse highlight. In this area, the color separation will carry a scant zero to three percent proportional C.M.Y. printing dot. Although the "A" patch is used as reference to clean a transparency of overall color contamination, grey balance should remain the prime directive of any separation set points.

On the opposite end of the scale, at the bottom of the shadow, the "B" patch is the 1.9 density area. It should be used for the setting of the maximum percentage of printing dots. Beyond this point, no shadow information is resolved.

The placement of the middle tone "M" patch in relation to the highlight "A" patch, also known as the  "A" to "M" range, will determine the main profile of the reproduction curve and the overall look of the printed image. This range may vary from .75 to .85, depending on the specific dot gain compensation. Other intermediary set points are used to further adjust the relative contrast of the separation in accordance with specific printing requirements.

A transparency has to be technically designed to meet the printing requirement. Photographically speaking, this means a transparency that has all its information contained in 4 F-stops (of aperture) from readable shadow to readable highlight. This requirement applies not only to the reproduction of original art but to all transparencies which are designed for the printed page. There is little room here for artistic license. A transparency may be stunning when projected, but that does not mean it is adequate for printing purposes.

Aiming for a brightness range within 4 F-stops gives plenty of exposure latitude, but a grossly underexposed transparency has compressed shadow detail that no color scanner software is designed to resolve. On the other hand, an overexposed transparency is missing details in the highlights, and where nothing is recorded, nothing can be done. The erroneous tendency is to underexpose a high-key original in order to get more detail. But this leads to a shift in overall density, lack of highlight contrast, and muddy reproduction every time. Although you should maintain the density range close to 1.9, you should also avoid shifting or altering the shape of the photographic contrast curve. A photographer familiar with the zone system of black and white photography should apply the same rigor to calibration by locating, as we said earlier, the range of illumination inside of 4 F-stops.

In a view camera, overexposure or, more often, underexposure may come from miscalibration of exposure metering. The compounding of bellows extension factor, correcting filters, and F-stop aperture setting can easily induce more than one and one-half stops of exposure error. A film plane probe metering can ensure consistently accurate exposure.

Professional photographers rely on color bars and grey scale. Not only do they measure their exposure with photo-densitometers, but they also verify, every time, the resulting transparency's exposure with transmission densitometers.

Back to top

Compressing the tonal range to 1.9: Who decides?
This tone compression will occur in the printing process whether you want it to or not. When the copy transparency exceeds the 4 F-stops of brightness by default, the decision of how to compress the tonal range most often becomes the responsibility of a color separation technician who is usually working at a disadvantage. In the first place, he/she does not have the original art at hand as a reference. Secondly, pre-press technicians rarely have the visual art training necessary to make the kinds of decisions and compromises implied. Because the professional photographer, the curator and/or art director have access to the original work of art, it is their visual literacy and decision-making abilities that must be engaged from the start.

A report on the copy transparency's performance versus the original work of art should be prepared by the photographer and made available to the color separation or pre-press specialist. Such a report should mention emulsion characteristics, color saturation deficiency, contamination of hues, the selection of highlight and shadow positioning in the transparency, etc.

Remember that a copy transparency tends to look spectacular, and for this reason it is often not properly assessed in side-by-side comparison with the original. Omitting this important step can result in multiple remakes and a lot of frustration. Note that this comparison must be made under 5000°K illumination to avoid any problem of metamerism.

How to compress the original tonal range?
Compressing the range to meet the printing requirement is achieved by controlling the contrast of illumination. I believe that double polarization techniques should be avoided altogether, especially if you are using an electronic flash system. Although not as disastrous as the dull varnish spray can, polarizing is a clumsy way to take care of an original surface reflection. Most of the time, polarization shifts the positioning of contrast and shadow tone separation beyond the software capability of the modern electronic color scanner. A simple densitometric evaluation of the transparency's tonal distribution and photographic curve should demonstrate if the contrast shift is acceptable. If a density zone is not readable or cannot be differentiated by a transmission densitometer, it will not be resolved by a color separation scanner.

Back to top

Controlling illumination without the hint of a glare.
When all else fails to take care of surface reflection, using a diffuse light tent technique will do the trick in most cases. A clever dosage of illumination sources allows the photographer to define directional surface characteristics of a painting, since a painting is rarely a flat subject. If the grey scale is subjected to an intensity of illumination different to that of the painting or if it has glare, then the transparency will be defective for proper calibration at the color separation stage.

The diffuse light tent is best all around and has the advantage of placing the range of reproduction closer to the 1.9 density goal with a proportional acceleration of the photographic contrast curve.

The photographic light source.
The type of illumination is a determining factor in contrast, color, and surface rendition of the work of art. Tungsten or flash systems work as long as they are in tune with the film emulsion to be used, but only the top-of-the-line professional flash systems have equalized spectrum range and constant power emission. These flash systems produce almost no heat, an advantage when working with easily damaged original works of art. Some, such as the Broncolor, come with a set of spectrometric flash tube filters to aid the photographer in tuning the light source color temperature. To avoid photographic cross-curve situations and the like, the flash system modeling light should remain inactinic to the speed index of exposure.

Calibrating a film emulsion.
Just as hues of ink differ from batch to batch, so does photographic emulsion. A routine test conducted with a professional photo lab will determine the proper compensating filter correction and exposure index of a specific emulsion batch used in a specific illumination context.

If a copy transparency grey scale and color bar shows obvious signs of poor exposure and/or color contamination, it would be incompetent to pretend that such a transparency looks anything like the original. On the other hand, if the photographer chooses to ignore these quality control devices in order to exercise artistic freedom over the interpretation of the original, he/she should make this intention clear prior to the color separation stage. In this case, one must question the use of any control device to appear in the transparency, in the first place. The scanning technician would, and appropriately so, follow the standard procedure of calibrating the separation set points to these devices in order to recover the colors of the original, and clearly not attempt to match the interface transparency.

Back to top

Sizing up the grain structure.
The finer the emulsion grain structure, the better. But a larger transparency may not necessarily produce a better reproduction. Electronic scanners gain resolution on a magnification mode that is contrary to optical enlargers, which makes the 80% to 300% reproduction size the ideal ratio and the 4"x5" transparency the best all-around format. When a reduction ratio is needed, a scanner should be used in a high resolution mode. This doubles the electronic sampling of the original by multiplying the reading passes. An 8"x10" copy transparency can be outstanding, but proper exposure tests are often not carried out simply because of the inherent cost of the photographic material. Yet, the initial cost of one sheet of film is nothing compared to the gymnastics that an inappropriate transparency will result in, further along in the process. Don't expect to fix a transparency with a scanner.

Grossly underexposed, poorly processed and/or high-speed transparencies often show grainy layers of emulsion that can't be referenced properly through the microscopic eye of a scanner. For example, the yellow layer, located at the back of the film emulsion, incorporates partially in its structure the grain image from the front layers of emulsion. The microscopic sampling of an area of the yellow layer by the scanner may cause difficult or erratic calibration of the software set points. In fact, the positioning movement of the scanner across this grainy area could be compared to trying to skateboard in a gravel pit. The resulting separation can turn out to be deficient in or contaminated with yellow. This problem can be alleviated by calibrating arbitrary set points, or by using a slight soft focus of the photo-multiplier reading head. If this sounds like guesswork, it is!

Back to top

What brand of copy transparency can be used?
Film emulsions were designed to reproduce skin tones, not to copy artwork or paintings. For marketing purposes, American manufacturers gear their product to render redder skin tones — giving that tanned, all-American look — while Japanese products render slick, cooler tones. You might come to the conclusion that a painting being reproduced can only be photographed with a tint of cultural color balance, and you wouldn't be all that wrong. For example, an emulsion can resolve two different hues, such as green and brown, by lessening or exaggerating their differences. It may shift their relative brightness, and it may also react disproportionately from shadow to highlight. The photographer should choose the emulsion that is the most compatible with the color hues to be recorded. Even so, this is no guarantee that the hues recorded on the transparency will, in turn, be readable by a scanner photo-multiplier.

By changing the type of film and light source (from daylight to tungsten or vice versa) or its color temperature, you may either solve or confuse the closely related hues. In a similar manner, the photo-multiplier of a specific scanner may be blind to a transparency hue which is easily distinguished by the human eye. Hues that are missing in the copy transparency will not appear miraculously at the separation stage, with or without the help of grey scale and color bars. Color compensating filters, which are mainly designed to secure a film emulsion grey balance, may be used to some extent to help the reading of a specific hue, but at the risk of affecting other hues or creating some overall contamination.

It has been my experience to photograph 10 paintings in exactly the same setting and under the same conditions, only to find that one of them refuses to fall into place. For whatever reason, the combination of painting pigmentations, light context, and photographic emulsion is obviously incompatible for that one painting. The combination then has to be revised, keeping in mind that an elegant solution is not always at hand.

Back to top
 
What if the copy transparency is off?
A very old transparency (and sometimes three years is old enough), improper exposure, unrecorded or contaminated hues due to poor illumination or processing — all of these can lead to problems. Unfortunately, a fine transparency is not always available and electronic color scanners are not intended to correct photographic cross-curve situations and the like. The procedure will be expensive, tedious and, at best, will give average results.

Before it hits the scanner, a transparency should be scrutinized by a technician who, with checklist in hand, ensures that the transparency meets the conditions on the customer request form, and carefully plans the procedure to follow, anticipating where any problems might arise. There is plenty of high-tech instrumentation available to carry out this professional analysis — including the telephone and email. The infamous "If it's not good, we'll redo it," is evidence of poor problem assessment and lack of technical expertise. After all, time is money.

Back to top

Is the painting upside down?
When this happens in an art gallery, everyone laughs — except the artist, of course. But when it happens in the printing industry, everyone is looking for the incompetent individual who doesn't know up from down or left from right. The best time to stop this embarrassing problem from occurring is at the photographic stage by adding to the color bar and grey scale an arrow sign with the word "up" written next to it.

How about care of the painting being reproduced?
It is the author's belief that unless you are dealing with a professional photographer specially trained in copying a work of art, your painting is not in safe hands. Pre-press technicians simply do not have the expertise to handle original artwork, regardless of how careful they may be. It is recommended that a conservator accompany the work of art on every occasion that it is necessary to use it as immediate reference. Always determine if insurance will cover the cost of damages and whose insurance will pay.

Back to top
 
Color proofing: a communication tool.
Various proofing systems are available to test a color separation and simulate its result on the printed page. Since the color proof is somewhat of a contract with the client, it becomes important to exercise some control to ensure proper communication. The color hues of these proofing systems are adjustable, to some degree, to a printer's ink set. Color separation houses present a proof set which is tailored to meet the average printing requirement. In many cases, the individual printer must supply to the client an in-house proof which is more closely adjusted to his/her ink set and press performance. A color proof can be made with a high degree of repeat consistency; however, during a press run, the appearance of the printed page is known to vary. If one is not happy with the color proof or has some hesitation about it after examining it under 5000°K illumination, the matter should be addressed and rectified at this stage. It has been my experience that, if the color proof is not quite right, the headache will not "get better on the press", even after 50 thousand printed copies.

Back to top
 
The press is on the run.
Even though more than 150 variables are at play, the main factors influencing the "look" of a reproduction are the grey balance, the contrast, and the shift of the middle tone of an image. This middle-tone shift is caused mainly by dot gain, a factor which varies from press to press and even, to some degree, during a single press run. Dot gain makes it difficult to preview the results out of conventional color proofing systems. Dot gain or deterioration of the dot structure is aggravated on porous, uncoated stock. Similarly, the peripheral surface augmentation of a single dot is proportionately larger on a finer screen ruling. The finer the screen, the more dot gain compensation has to be built in at the color separation stage. In most cases a 150 lines per inch screen on a coated paper will produce a quality reproduction; finer screen rulings are only better if well under control.

A scanner technician can make a 400 lpi separation at the flip of a switch, but it is doubtful that such a separation was designed with printing in mind. The modern offset press uses rubber blankets to transfer the ink to the paper — the soft part of the process. Conventional proofing systems are not designed to resolve a film separation that is finer than 200 lpi, nor to emulate the fine dot behavior of direct to plate. For that reason, they would be of no use in simulating the printed page for such a fine screen ruling.

As a rule of thumb, a color proof which shows a lighter middle tone appearance is more forgiving on press. In fact, an anemic and muddy look will result from having to reduce drastically the amount of ink (solid ink density or S.I.D.) in order to compensate for a middle tone that climbs towards the highlight. It is therefore preferable to give the pressman some room to maneuver by building into the separation a lighter middle-tone set point. Although monitoring the S.I.D. is essential to maintain grey balance, monitoring contrast during the press run ensures a closer visual match. Changes in the lamination layering used in some proofing systems simulate a variation in the S.I.D., rather than a classic case of middle-tone climbing. Middle-tone gain does not automatically imply shadow gain.

Building expertise in color separation is only possible through the assessment of printing behavior. Scanner technicians, for unknown reasons, rarely follow their work to the printed page and are in fear of shadow gain. On the other hand, printers show their work to scanning technicians when and only when it does not work. Yet a large proportion of rejected printed products are rejected because of middle-tone and upper middle-tone dot gain — the anemic or muddy look. There is nothing wrong with dot area measurement and tailoring dot formation to the printing requirements rather than to the proofing systems, but when it comes to the communication contract, proofing is no signature of a perfect match.

Modern electronic scanners permit the pre-recording and analysis of a set of the printing requirement. This data is collected from previous printed sheets and fed into a computer software. The software then calculates a new set of values, so that the reproduction curve can be optimized to the printing context. A scanner operator can easily tailor the separation to each customer, press, etc. Although the technique is not difficult, it seems that few technicians have the expertise to carry it out. Instead, some rely on what can be educated guesswork to take aim at their target — the average printing job.

Back to top
 
The 5000 degree Kelvin story.
We are all familiar with the cacophony of sound produced by an orchestra as it strives to tune all its instruments to the same key. But, what order results when that magic frequency of 440 Hz is reached! A similar reference to set an illumination standard for printing has recently put the entire industry on the same pitch — the 5000 degree Kelvin calibration reference.

The 5000°K illumination standard provides a good context in which to compare, side by side, a proof and a printed sheet, since it enables the pressman to adjust the ink fountains to match the proof more closely and to maintain grey balance. Any demand on deviation in density of more than +/- .10 on any one ink film from a pre-established ink set standard density should call for a remake of the separation or plates. The rampant tendency to ask for "more red" at the press check is probably the best illustration of this, and in all cases, it is a shocking visual experience for the client when the work is taken away from the 5000°K to a different illumination context.

If an experienced pressman cannot match the proof, it's often because the color proof has been exposed or processed improperly, or because the hues of ink sets are incompatible with those of the proofing material. If the pressman does not know why it does not match, it is a sure indication of his or her lack of experience. But if you didn't like the color proof in the first place, expect the printed sheet to be a disaster.

Back to top

The infamous grey scale and color bar
Maintaining grey balance is fundamental to the management of four-color reproduction. The printed grey scale is placed alongside the scale used for the copy transparency in order to match their densities and color characteristics. The color bar should show no sign of color contamination, if it was made with the same hues of ink to be used on the printed sheet. When grey scale and color bar are adjusted to the scanner's electronic scale, the color of the original work of art should fall into place. Remember that densitometers measure densities, not color; it is to be hoped that spectro-densitometers will soon gain acceptance in the trade and simplify the grey scale story.

Now, the grey scales match, but not the painting or artwork? Oops!! This happens all too frequently, and it's mainly due to the color deficiency or contamination of the transparency or a scanner photo-multiplier that refuses to record some particular hues, as we discussed previously. If the scale has been neutralized and the color bar is free of contamination, you have half the answer. The other half lies in color manipulation of specific hues and the team effort of an experienced scanner technician and the art director. In their evaluation they must take into account that the human eye is the ideal instrument for comparing color side by side, but its color memory is unreliable. One cannot trust one's memory of "how a painting looks" as reliable criteria for adjusting color reproduction.

Leave nothing to chance.
Printing is neither an art nor a science, but a technology that demands an understanding of all its subtleties — its physics, chemistry and mechanics. To achieve optimal results, those who are attempting to reproduce original art on the printed page must have an in-depth knowledge of both the print medium, the medium to be reproduced, and the interfaces between the two. I hope that the information contained in these pages will help anyone who is serious about providing a quality message through the printed page.

Nelson Vigneault, CEO, CleanPix Corp.

Back to top

Tuning the Vocabulary.
Grey Balance: is the relative percentage of all combined color (C.M.Y.K.) necessary to render a neutral tone for each step of a grey scale or original. Failure to maintain grey balance causes overall color contamination, which is further aggravated when the printed page is seen under different light sources.

Contrast: is determined by the gamma angle of a particular point or zone of a photographic or reproduction curve. In layman's terms, it is how quickly the visual density accelerates or decelerates from one step of the grey scale to the next. The density change would be steep when contrast is high and flat when contrast is low.

Printing Requirement: is the percentage of each of the C.M.Y.K. printing dots allocated to each step of a scale (or original) so the printed page will match or simulate closely the original on a given interactive system of film, plate, press, paper and ink set. Any variation of the above changes the printing requirement, and the separation has to be adjusted accordingly.

Tonal Reproduction: is the equalizing of contrast profile distribution throughout the density range of a scale (or original). Tonal reproduction is achieved by accentuating or reducing relative density shifts between the various zones of an original — i.e. shadow to 3/4 tones, 3/4 tones to mid-tones, mid-tones to 1/4 tones, 1/4 tones to highlight.

Further references:
Southworth, Miles. Color Separation Techniques, 2nd ed. Livonia, NY: Graphic Arts Publishing, 1979.
HQ, "Blau," 7/1987. Heidelberg, West Germany: Heidelberger Druckmaschinen, 1987.
Info 30,  "Measuring Photography for Offset Reproduction." Pittsburgh, PA: Sinar Ltd. Schaffhausen/QuadGraphics/Black Box Corporation, 1987.
Info 31,  "Contrast Control in Practice." Feuerthalen, Switzerland: Sinar AG Schaffhausen, 1988.
Kodak Color Films, 8th ed. Rochester, NY: Eastman Kodak Company, 1980.
Kodak Color Separation Guide and Gray Scale, Q14, 35.6 cm. Cat. no. C80-152-7666.
Paperworks Colour Bar. Calgary, Alberta: Paperworks Press Limited, 1987.
Paperworks Colour Guide. Calgary, Alberta: Paperworks Press Limited, 1987.
The Quality Control Scanner. Livonia, NY: Graphic Arts Publishing. (available by subscription)
White Space, vol. 1 no. 1. Chicago, Ill: Consolidated Papers Inc., 1984.
White Space, vol. 1 no. 2. Chicago, Ill: Consolidated Papers Inc., 1984.
White Space, vol. 1 no. 3. Chicago, Ill: Consolidated Papers Inc., 1984.
 
Back to top