What he also knew is that the conventional Process Color (Cyan Magenta Yellow Black, or CMYK) ink-on-paper color gamut is limited. You cannot reproduce every color that your eye can distinguish using CMYK inks on paper. This is not a matter of perception on the part of a scanner operator; it is not a matter of attitude on the part of a printing press operator; it is not a matter of cost and time as presented by a sales representative. It is a matter of science.
You cannot "match" a proof on press (unless it is a press proof made on the same press ... and you probably still can't do it). This, too, is a matter of science.
You cannot "match" what is on the computer screen with a proof nor on press. Again, it is a matter of science.
The degree to which a computer monitor, a digital proof, a layered film proof, a contract film proof, and a finished printed piece "match" is, to some extent, a matter of perception. It is also a matter of calibration and process control throughout the various operations in pre-press, at the supplier(s) to the pre-press house(s), at the paper manufacturer(s), at the ink manufacturer, and on press.
This does not mean that you can never get anything to look good. In fact, how good something looks is a matter of perception on the part of a scanner operator; it is a matter of attitude on the part of a printing press operator; it is a matter of cost and time as presented by a sales representative.
No matter what kind of proof you deal with, it does not show you what you will get on press; if calibrated properly, it represents what you can get on press.
Consider photographing a silver vase of red roses on a white cloth.
Light strikes the red roses. All of the light except red light is absorbed. A great deal of the red light is absorbed, also. Some red light is reflected back, but it is disbursed by the uneven texture of the rose petals. Thus the roses appear deep red, and not like a point source of light.
Light strikes the green leaves. All of the light except the green is absorbed. A great deal of the green is absorbed, also. Some green light is reflected back, but it is disbursed by the uneven texture of the rose petals. Thus the leaves appear to be deep green, and not like a point source of light
Light strikes the silver vase. Very little of the light is absorbed; a great deal of it is reflected. It is not disbursed, because silver is specular, or mirror-like. Therefore, you see the spot light reflected in the silver vase. It may be distorted, since the silver vase is curved. However, the vase appears to be a point source of light. Light reflected from other objects in the room hits the vase and is reflected, also, so an image of the roses and of the leaves and of the camera is seen in the vase.
Click
Now all of this reflected light hits the film in the camera and initiates a chemical change on the slide (transparency) film in the camera. The film is processed.
Choose Your Shot
You lay the photos out on a light table and choose the one that looks the best. You do not choose the one that most closely matches the actual studio set up. The transparencies you are looking at on your light table do not match the product at the photo shoot. Ektachrome is slightly bluer than Kodachrome; Fuji films are slightly more saturated. No film uses tincture of rose petal, so the red isn't the same red. It isn't. Really!
By the way, what color are the lights in your light table? A conventional flourescent bulb is bluer than an incandescent light bulb. A T48 energy saving flourescent is different from a conventional one. A "color correct" flourescent bulb is different still, and is similar to overcast outdoors. None provides light similar to sun light.
You are choosing the photo that looks best, not the photo that matches something. You are also making that decision based on relative color, not absolute color. The red rose will look red under slightly yellow or slightly blue light. The photo with the brightest red rose will appear to have the brightest red rose, regardless of the light. But if you had two light tables side by side, the red you would see on one would differ from the red you would see on the other.
Choose the one that looks best and take it to the scanner operator.
Scanners transmit light through the transparency; the light is selectively absorbed by the transparency, and the remaining light hits a light-sensitive photo multiplier or CCD array. In some way, the entire image is divided up into a grid of pixels. The light-sensitive thing assigns values in red, green and blue, the primary additive colors, to the light transmitted through the transparency at each pixel location.
The values of RGB are recorded as numbers in a computer. The precision with which the scanner can differentiate between shades of each component is determined by the hardware leading up to this point, and the bit depth of each pixel.
Pixel Bit Depth
Consider each pixel having 8 bits per color. In binary, there are 256 numbers available in an 8 bit word. The binary number 00000000 is zero; the binary number 11111111 is 255. Give an 8-bit word to each of red, green and blue and you get 24 bits per pixel ("24-bit color"). The total number of colors that can be created with 3 8-bit words is 256 X 256 X 256 = 16,777,216 (16.7 million colors).
Some scanners will record higher pixel bit depth, so that after color correcting or shadow enhancement or whatever other function is being performed by the scanner or its software, there is plenty of extra data from which a finished, 24-bit color can be defined for each pixel.
Some scanners record color in CieLAB space rather than RGB, which allows for controlled transformation to other color spaces. CieLAB space defines hue in two dimensions and luminosity as a third.
Monitors tend to be too bright and too blue. There are ways to calibrate a monitor to more accurately represent a printed piece. Your scanner workstation must have been so calibrated in order to do any good for soft proof (the image on the computer screen is referred to as a "soft proof"). Please, note the difference between a monitor that is calibrated to reproduce its reds, greens and blues according to its factory specification versus a monitor calibrated to look like print. The two are unrelated, although maintaining the former allows you to maintain the latter.
Without monitors calibrated to represented print, the only way to predict printed color at this point is by the numbers. A skilled scanner operator can read the CMYK values to which an area of the image will be translated, and know what that color will look like. Such an operator will know that there will be an overall cast of yellow or blue, or realize that the image is too dark or too bright. No matter how good the "soft proof" is, it helps to have a scanner operator who also knows what the numbers mean.
Now you have a computer file in CMYK, which has been projected by a computer screen calibrated to represent print by playing with how it selectively projects its red, green and blue light sources into your eye. The image you see looks good. It looks similar to the transparency. It does not match the transparency. It cannot. It does not match the original flowers in the original vase on the original white cloth. It cannot.
Down to the Bottom of this Document
There are color printers, there are really good color printers, there are digital proofers, and there are really good digital proofers (that was in order of the probability that they will accurately represent the finished printed piece, last being highest ... but still only a probability). Let's assume that you are using a really good digital proofer that has been calibrated to closely match film proofs and printed pieces.
A digital proof is ink or dye on paper. Light strikes the proof, is selectively absorbed by the ink or dyes, is reflected by the white paper under the ink or dye, goes back through the ink or dye and is selectively absorbed again, and lands in your eye. We are now dealing with reflected light, not transmitted light. We are dealing with CMYK, or the primary subtractive colors, not RGB, the additive transmissive colors of scanners and computer monitors. The digital proof cannot look like the image on the monitor. It cannot look like the transparency lying on a light table. It cannot look like the reflected light of the original photo shoot. You are now dealing with a severely compressed tonal range and colors that have been interpreted by the film, by the scanner hardware, by the scanner software, by the scanner operator, by the digital proofer, and by the paper on which the proof is made.
The digital proofers that use dye sublimation are causing dye to be absorbed into the surface of special paper. This is not ink-on-paper. Therefore it cannot match the ink-on-paper of a printing press. It can look similar to it, but not match it.
The digital proofers that use ink-jet technology put ink on paper. This may seem to be a step in the right direction, but the inks are far less viscous than offset litho inks, the dyes are not necessarily the same as those used by the ink manufacturer for your printer, and the patterns in which the inks are splattered onto the paper are not the same as the patterns of halftone dots or stochastic spots transferred from film to printing plates.
In most cases, the proofer cannot print on the same paper stock that you will use for the printed piece. Even if it can, the inks are different enough that they will perform differently.
The digital proof can give you a good representation of what a printer should be able to get on press, but it won't match. It can't. It can be very close, but it can't match.
The digital proof made from a calibrated system can allow you to make good, informed color decisions. That is how they should be used.
Now go to film, and get a film proof.
A printer will be able to very closely match a contract film proof on press at color OK. It will not be an exact match, but it can be very close, especially when dealing with highly reflective, white, cast-coated stocks, printing one image on one page at a time. But normally you print 4 or eight pages at a time, each with different images, different ink densities, and requiring different trade-offs for color. Then as the run progresses, various conditions on press cause color variation. So only a portion of the final print run will match the sheet from color OK.
Down to the Bottom of this Document
We are frequently asked why an Iris® digital proof doesn't reproduce Pantone Reflex Blue (or any number of other PMS® colors) faithfully. Well, it can't. We are also often asked to choose a color that more accurately represents a PMS color on our Iris proofer. We can't. Pantone has spent a great deal of time doing just that; we can't do any better than they already have.
No digital color printers can print spot colors; they all show CYMK builds of the spot color. If you have a Pantone Process Color Imaging Guide, it shows Pantone Match colors next to the best CMYK screen build they can come up with for it. Many CMYK screen builds are very different from the spot color they are to represent. The same results can be expected with any color printer.
What do you do to proof spot colors?
This is another question frequently asked of us. Well, how much money have you got? How close of a match do you really need?
Few film proofing systems can show accurate spot colors.
A Cromalin® (a specific brand of proof) can be made with special color mixes; however, their accuracy depends on that of the operator who mixes the color.
Some proofing systems, including Cromacheck, have PMS color films. These can be color-accurate, but few pre-press houses stock the entire Pantone library in proofing materials.
Some companies will do actual ink proofs. Such proofs tend to be more saturated than results on press; such proofs tend to have a thicker ink film than will the press-printed piece.
The most accurate proof for a spot color job is a press proof. Few projects warrant that kind of expense.
So what do you do about spot color proofing? We have two recommendations:
Down to the Bottom of this Document
Computer Screens
There are programs and devices that make a computer screen look more like a printed piece. Only a few programs will represent colors accurately in such calibrated environments, however. Photoshop® with a properly calibrated monitor can show you a close representation of how a photograph will print. Place that image in QuarkXPress® and the computer screen no longer shows the color correctly. Make a box in Illustrator® and fill it with a color. Make a box in FreeHand and fill it with the same color. Place both in a page layout program in which you have made a box with the same color. No two will look alike. Only recently are the various software and hardware manufacturers coming to grips with creating screen views of colors in the same way.
The only way to see how all color elements in a project will look is to get a decent proof of the entire layout.
Digital Proofs
You might wish to get a digital proof early in a design cycle in order to avoid the expense of film. Make sure that your proofing device, or that of your service bureau is calibrated to accurately represent ink-on-paper from a printing press. If a service bureau offers both digital and film proofs, ask to see one of each for the same images or file. In Tandem Design, Inc. has such examples readily available, because we had to make them in order to calibrate our digital proofer. So should any pre-press house dedicated to good color reproduction.
If your service bureau has an accurately calibrated digital proofer, you can make process color decisions based on it.
However, most digital proofs are not on the stock which will be used on press. They are often on unrealistically glossy paper, which looks nice but doesn't represent your project. If you get the digital proof to wow your boss, go for the glossy. If you get the proof to see what your project will look like, consider a less glossy stock. Many digital proofers allow for such a choice.
Also, a digital proof will normally not represent the dot gain you may experience on press with the stock you will use. Be aware that you should ask the service bureau to build dot gain into the film, if you expect dot gain on press. Ask your printer how much is expected with the stock you've chosen. Even if you choose a sheet which is very bright and has good ink hold-out, a press operator will have more latitude on press if the film has some dot gain built into it. We recommend 8% dot gain even with the best press and the highest quality sheet. You can always bring the ink up on press; problems arise when you can't bring the inks down.
Film Proofs
However, be aware that there is another very good reason for a film proof. The film proof and the printing plates are made from the same film. If there are any technical flaws in the film production, they will show up in the film proof. Also, digital proofers don't accurately represent overprints, traps, and various other file problems (the element left as PMS 185 instead of being separated to CMYK). The film proof will show these problems.
If you need to make color decisions, they can be made on calibrated monitors (for color scans), or on calibrated digital proofs, or on contract film proofs. If you need to make a final technical check on the film, you must do so with a film proof.
Why not just use the blue line? A blue line is normally made from only one, two or three of the CMYK films. Technical flaws in the other color(s) won't be seen. Also, blue lines are not in color; you cannot see that everything came out right on a blue line.
Contract film proofs are not all the same. In fact, proofs made with Cromalin®, Matchprint®, AgfaProof®, Waterproof®, and 3M Contract® proof might each be slightly different from the other.
Film proofs do not represent what you will get on press; they represent what you can get on press.
You can also get results on press that do not look at all like the film proof, nor the digital proof, nor the computer screen. There is a great deal that can be adjusted on a printing press to alter the color of a printed piece. There are many conditions associated with layout that can influence color on press. And there are conditions on press that vary over time, and, therefor, during the press run. The first sheet and the middle sheet and the last sheet won't match each other, let alone the film proof. Sorry, it's true.
Down to the Bottom of this Document
One last word on proofing, calibration and color control: you should get your digital proofs, your film, and your film proofs from the same source for any given project. There are aspects of calibration that permeate all functions of a given pre-press shop. They will be different from shop to shop. If you get a scan and digital proof from one shop, image film at another, and have the printer make a film proof, you may not get the results you want or expect. And you may have a difficult time getting someone to accept responsibility for correcting the problem.
What is the Objective
"Matching" color is a goal rarely reached. Paper used as the base for proofs is rarely the same as that used on press; even if the same paper is used, the colorants used for proofs are not the same as the inks used on press (and inks from different manufacturers do not match each other on press); lighting conditions can cause perceived color shifts; some people are more sensitive to color differences than others.
But is the objective to "match" color? Match what? ... the original photo shoot? ... the computer screen? ... the digital proof? ... the film proof? ... your idea of what it should be?
The objective might more accurately be stated as:
To get a printed piece that looks great as a result of all your efforts and expenditures.
There are procedures you and your vendors can follow to get as close to that objective as possible. There are checks along the way that you should make to achieve it. There are things you can avoid which make the objective harder to reach. There are compromises you will have to make; so that the overall project is a success, even if certain aspects of it are not exactly the way you wanted them.
Be aware of what your pre-press and printing vendors can and should do. Be aware of what they cannot do. And ask beforehand if you are not sure.
Forward comments or questions to harry@email.intandem.com
Back to the Top of this Document
Back to the Color Matching Start Page
Over to the short dissertation on color
HTML, graphics and textual content © In Tandem Design, Inc. 1996.
An Example in Process Color
Nothing Can Actually "Match"
Light strikes the white cloth. Some of the light is absorbed; the rest is reflected. No particular color is absorbed more than any other; thus it appears white. The reflected light is dispursed by the uneven texture of the white cloth. Thus the cloth does not appear to be a point source of light.
You are looking at a transparency on a light table. You are looking at transmitted light ... light that eminates from the bulb, goes through the transparency where it is selectively absorbed, and shines in your eye. Your eye perceives Red, Green, and Blue (RGB) light, the additive primaries.
At some point, you or the scanner operator is looking at the scanned image on a computer screen. The monitor is analogue (not limited to the 256 shades of an 8-bit word). The monitor projects light. It projects light in increasing or decreasing strength from little phosphors on the screen. This implies selective projection, not selective absorbsion as in all previous situations with the transparency.
The scanner operator and you might tweak the curves a bit before or after scanning. Then the digital image is translated to CMYK, the Subtractive Primaries. This step should always be done by the best piece of gear and/or software in the production cycle. Photoshop® will translate from RGB to CMYK. Photoshop costs under a thousand dollars and does a lot more than just translate from RGB to CMYK. High-end scanner software costs many thousands of dollars. A good deal of what you pay for is the color transformation algorythms and tables. Use them. Getting a repro-quality scan to disk in RGB and translating it to CMYK at home may not assure the best print-media results.
Back to the Top of this Document
Let's print a digital proof.
Some film proofs are put on the stock you will print on; some are not. Some use laminated layers of plastic, some use powdered pigments, some transfer films of ink ... there are many different "contract film proofs." The term "contract" means that the printer and the customer can treat it like a contract; the printed piece will look like the proof (... within industry standards, within variations on press, within a tolerance associated with paper stock, the eye of the pressman, the eye of the person doing the color OK, etc.).
Go to Press
Back to the Top of this Document
Various Proofing Problems
Problems with Proofs and Pantone®
Back to the Top of this Document
Computer Screens, Digital Proofs and Film Proofs
Is an image too light or too dark? Are the colors saturated enough or too much? Are the colors the right colors? Do colors chosen for non-photographic elements work together well? To answer such questions, you must be viewing a relatively accurate representation of what will happen on press.
Computer screens tend to be too bright and too blue. They also can present a much broader gamut of color than can a CMYK printing press. Finally, they are projected light whose color depends on additive primaries: red, green and blue. A computer screen will never look exactly like the printed piece. It can't.
Some printers or magazines will insist on a film proof for use at press. One reason for this is the possibility that a digital proof is not properly calibrated. If your printer works with your pre-press house, this might not be an issue.
Back to the Top of this Document
Wrap Up
Just to reiterate: proofs do not represent what you will get on press; they represent what you can get on press.
Thanks for visiting!