Reproducing Reality

The challenge of color management.

We’ve been working with color management for many years now. In fact, it’s hard to remember a time when ICC profiles were not a key part of a production workflow, and words like "characterize" and "calibrate" had yet to become part of our everyday vocabulary.

At this point, then, you might believe that color management is a mature technology, that there are few significant advancements coming. We all have color management down pat, you might think.

But you would be wrong.

In fact, there is a sea change on the horizon when it comes to color management. Expectations for color are being broadened, requiring new and better solutions. Plus, changes are happening to some of color management’s core components, which are sure to have some impact on your production operation down the line.

Expanding the base of the pyramid

One trend everyone seems to point to is the greater use of color management, not only by professionals but also by consumers. The explosion in availability of high-quality color devices-from digital cameras to inexpensive- but-capable desktop inkjet printers-has raised users’ general expectations of an ability to get good color out of a production chain.

As Microsoft reports, in its white paper, "Windows Color System: The Next-Generation Color-Management System," a "swift adoption of color in the office and explosive growth of color in the home have brought us well past the demands and expectations associated with the first wave of color devices. Support statistics confirm that color-quality and color-management levels judged acceptable or good enough as recently as two years ago are considered inadequate by today’s standards." And that’s an assessment from 2005-imagine how much higher standards have risen since then. (Windows Color System is the color management system in Vista, replacing the Image Color Management system from earlier versions of the operating system.)
Doris Brown, vice president of marketing for Pantone, outlines the same scenario a little more succinctly when she points to the expectations of today’s color consumer: "‘I bought a new computer, I bought a new digital camera, I bought a new printer, I’m going to make great pictures.’ Wrong!"

But Brown sees help on the way: "The people who really get color management are making more and more tools for the uneducated to use. The companies with the knowledge are working on bringing this to more people. That is the trend."

One goal, she says, is to "package color management in a fun tool that would make someone want to use it, so that people get managed color without ‘color management.’" Perhaps not surprisingly, she points to the Pantone Huey pen-size monitor calibrator as an example. "This device comes with no manual, it’s so easy to use," she boasts.

Will Holland, vice president of corporate marketing for X-Rite, agrees that, "The technology is going to ‘go down-market’-it will get more and more inexpensive and more and more prevalent." And echoing the color-to-consumers theme, Holland predicts that color management will soon be embedded in home-theater displays: "We’re now seeing ‘colorphiles’ who are as picky about their screens as audiophiles are about their sound," says Holland. Not long ago, audiophile sound was the province of those few who were willing to do the work to put together a system. Now you can walk into Best Buy and walk out with a quality home theater sound system without knowing anything about how that quality was achieved. When consistent color is similarly easy to get on TVs, consumers will come to expect it elsewhere as well.

According to Brown, color management is also beginning to reach new professional markets. "One of the biggest trends is the creative community beginning to engage in color management for their own use." For a long time, she says, color management was implemented downstream from the design community, at the professional production stage-in the hands of print providers.

Among the early adopters of color management in the creative community were digital photographers. "They clearly understand the need for a color-managed monitor," says Brown, and from that starting point they acquired other color-managed devices. That same movement is now occurring among designers, she says: "The designers also are beginning to understand using color profiles and how they need to incorporate color-management practices into their workflow."

Peter Constable, color workflow expert for Adobe, sees the same transitions. "Photographers have traditionally been one of the leading adopters of color management," he says. At the other end of the spectrum, though, are designers: "Traditionally, they’ve been less comfortable with color-management implementations and technology. But we see that changing moving forward."

"The key to making ICC color management more effective and usable is a user-interface issue," Constable continues. Vendors need to get away from intimidating phrases like "rendering intent," he says, and make the process more transparent. Not surprisingly perhaps, he believes that Adobe is at the forefront of that process. In Creative Suite CS2, for instance, "many users say they don’t even realize color management is on," he says.

Doris Brown sees the intriguing possibility of an enlarged color-management user community beginning to create a feedback effect. "There’s an audience of millions of general users. Once we unveil some of the ‘dirty little secrets,’ they can put more pressure on the big companies."

For example, she cites the issue of inks for color printers: Printer manufacturers don’t want to standardize on inks because it would break their business models, she points out. "Are you going to get Epson or Canon to have a universal ink? Probably not." But standardized inks could help with color-management issues.

Built-in solutions

Not only will color-management tools become easier to use on their own, but we’ll also see more of them built into the devices themselves. Holland says, "There will still be things that you as a user will have to do, but much of it will be handled in the background. A lot of the technology will be built into the output devices." As an example of this trend, he points to HP’s Z2100, Z3100, and Z6100 Designjet printers, which have added built-in spectrophotometers.

Max Derhak, senior color scientist for Onyx Graphics (and Onyx’s voting member in the International Color Consortium, ICC), agrees: "One trend I see is adding color devices to the printer. But that also has some issues to it." HP touts the ability of those Designjets to create ICC profiles; but, according to Derhak, that can be done better in other ways. What the spectrophotometers can do is enable the printer to maintain a consistent state. "It makes it very easy for the device to be calibrated, which will provide more control."

Nevertheless, Onyx is working toward making its RIP products able to accept data from such built-in measurement devices and generate profiles from it.

Frank Hueske, product marketing manager for EFI Proofing Solutions, says that a RIP vendor will always have a lot of work to do, as long as new printers and new substrates keep coming on the market. "The wide-format and photo markets," he says, "use a lot of different substrates. That means we have to characterize lots of new combinations."

On the proofing side, Hueske says almost wistfully, "You only need to worry about a couple of papers, a couple of ink sets. In the proofing area, papers generally match SWOP or GRACol. But in the production market, you need profiles for textiles, fabrics, glass. Any time you change the ink, you have to start all over."

Not only that, but new printers use "six, seven, eight, even nine colors," says Hueske. Working with these new devices requires what he calls "n-channel" color management-the ability to characterize the interaction of any number of inks.

CMM changes

A Color Management Module is the part of a color-management system that translates the color data among different profiles-it’s the "number cruncher," in Constable’s words. And developments in this area, too, are opening up new possibilities.

For one thing, last spring, Adobe made the Adobe Color Management Module available as a free download. Previously, Constable says, the Adobe CMM was "baked into" the Adobe color management system, which meant that the Adobe transforms-the computer routines or equations that translate one set of color data into another; the profile for a scanner into the profile for a monitor, for instance-could only be used in Adobe products. But current color-management systems such as Apple’s ColorSync and the Windows Color System make it possible to plug in third-party CMMs, which can be used by other applications. QuarkXPress, for instance, has a selector in the user interface enabling the use of third-party CMMs, but it’s not automatic: ColorSync will offer application program interfaces (APIs) for hooking into the plug-ins, but Quark would have to build in support for them as well. If that were to happen, designers and production people would be able to use the same color transforms in QuarkXPress and in Photoshop.

Another advantage to breaking out the Adobe CMM, says Adobe’s Constable, is that it supports processing of version 4 ICC color profiles. So does Vista’s WCS color management, but the ICM system in previous versions of Windows does not. So incorporating the Adobe CMM would enable the use of v.4 profiles in a non-Vista Windows workflow.

Derhak points to another intriguing trend affecting CMMs: "A move toward using dynamic, ‘smart’ CMMs." With a smart CMM, a profile can program the CMM to perform the transforms based on the optimum gamut mapping for the output device. Referring to the ICC, Derhak says, "We’ve added the ability for the profile itself to define a sequence of transforms. This will give more flexibility."

Sound great? The problem so far, Derhak warns, is that dynamic-color processing makes it hard to achieve consistency. It’s fine, he says, in a workflow with, say, one digital camera and one printer, but it’s not so great for a more elaborate workflow. Unless everyone is using the same version of the same CMM, two different profiles can trigger different results. Even if the CMMs are the same, two different computers can also give different results.

"The whole idea is to come up with a programmable CMM," says Derhak, "but we’re not necessarily there yet."

Devising, and dealing with, new color models

The final trend, and the one that will require the most work but offer the greatest promise of transforming the way we handle color, is geared toward a new, expanded way to define color. Today, ICC profiles and color management are based on the LAB color space, a theoretical model based on the way the eye perceives color without regard for viewing conditions. That makes it subject to metamerism, the situation in which two different colors can look the same, or two samples of the same color can look different, depending on how they’re viewed. A frequently cited example is the way the same color would look on a matte versus a glossy surface.

This is because a particular color can reflect (or emit) light at different wavelengths from many different points on the visible spectrum, not just those corresponding to the "right color." The eye responds to all those wavelengths but can only process them through the three color receptors (red, green, and blue), so different spectral profiles can produce the same color sensation.

Because of that, says Holland, "It’s not about ink any more. It’s about what the color looks like. ‘Appearance modeling’ of colors will become more and more important as the technology advances. Wouldn’t it be cool if you could proof what a textured wool skirt in a particular color would really look like?"

The new Windows Color System, for example, incorporates a "color-appearance model" that’s based on the CIECAM02’s set of equations to transform the tristimulus color space (used in color management today) in ways that take into account viewing conditions. (In CIECAM02, the "CIE" part is from the International Commission on Illumination, which is the international expert body on light and perception; CAM is for "color-appearance model," pointing to the idea that it’s supposed to account for the context in which the color is viewed. And 02 is the year of the current CIECAM model-before that there was CIECAM97.)

There are good economic reasons driving this process. For one, color varies not just by surface or lighting but, even more, by medium or material. Says Brown, "In cross-media production, people have to produce direct mail, packaging, a Web site, even something for a cell phone screen, for instance, and maintain a consistent brand across all of them. As long as companies have a desire to lock down their color, there will have to be a system for doing so."

Holland broadens the definition of "medium" even further: "Almost every industry has a need for professional color control." Standards exist for specifying and controlling paint and other colorants on the manufacturing side, he points out, but the design and printing side operate with a different set of standards. Given that, Holland asks, how, for instance, does Procter & Gamble make sure the color of the plastic bottle for liquid Tide matches the color of Tide in a cardboard box? X-Rite has technology on both sides, he says. "We’re converging them right now. It’s our responsibility to drive that."

In Frank Hueske’s view, "The best way would be to have spectral-color management." Spectral data measures the color itself across the entire spectrum, not just the tristimulus values based on the limits of human perception. And that would make it possible for spectral-color management to take into account lighting and surface conditions in a way that current color-management systems cannot.

"You would need spectral inks-inks for the entire spectrum, more than six or eight inks," says Hueske, "and you’d need a spectral CMM. I’m pretty sure that an entire spectral color-management system will arrive in 10 to 20 years. However, we’ll see a spectral CMM in the RIP within 2 or 3 years."

Derhak agrees with the goal, but is not quite so optimistic about the time frame. "A spectral-reproduction system is inherently difficult because rather than just three channels, you deal with maybe 31. The problem is that it’s a lot of work, and the overall benefit may not justify it," he warns. It would be worth it in the world of fine-art reproduction, he says, where they’re working with paintings with different surfaces and textures and spectral work is being done in museums. But beyond that, he wonders, "Are people going to be willing to pay for the extra cost to get spectral reproduction?"

That’s not to say his company isn’t preparing for a spectral future. "We do actually characterize printers using spectral data, and then we convert the spectral measurements to a colorimeter profile," he says. "But we still have the spectral data if we need it.

Regardless of how the next few years play out, the goal is still the same. "All our color technology is around two things: capture the reality in the first place, then emulate it or reproduce it effectively," says Holland. That is, and will continue to be, the bottom line.

Jake Widman is a freelance writer based in San Francisco. His last article for The Big Picture, "Garnering Better Shop-Floor Workflow" (July) examined how five shops re-arranged their equipment as well as staff to expedite the flow of jobs on their production floor.

View more from this Big Picture issue