American Printer's mission is to be the most reliable and authoritative source of information on integrating tomorrow's technology with today's management.


Jan 1, 2000 12:00 AM

         Subscribe in NewsGator Online   Subscribe in Bloglines

The management of color in print historically has been under the control of a select group of experts, usually living in a prepress department or hovering over a light box in a publishing company or ad agency. Their talents, while fairly well developed, entailed a mixture of color science, production limitations, physiology and psychology, since certain laws of physics prevent the reproduction of color for print to be an exact science.

Their expertise in analyzing and predicting color resulted from years of training and experience. They usually relied on an excellent understanding of the subtleties of color, and the creation of those colors as halftone dots on film. The loupe and a color proof were usually the tools used to affirm their expertise and support their judgement. Even the film proof required some interpretation, however.

As electronic prepress entered the mainstream, the monitor became a tool that could be used in place of the loupe, or even the densitometer, to read the dots on film. While it still required interpretation, if you understood halftone values, you could usually be confident of the final outcome, even if the monitor did not accurately represent the finished product.

As the cost of color printing started to decrease with increased demand, ways to supplement color expertise with computer technology were investigated. One of the solutions introduced enabled users to bring a device back to a standard setup to compensate for machine drift. This allowed users who were less experienced with halftone dot readings to achieve repeatable, acceptable color results.

In the early 1990s we saw the introduction of tools to calibrate color printers. While each of these tools was a step forward, none addressed the core problems associated with inherent differences in devices or human perception.

The real breakthrough came with the introduction of new tools that could bridge the differences between devices by characterizing each. This characterization allowed users to identify and map the entire color space of a device and, as a result, achieve a better cross device match within the limitations of the device. This all happened through the introduction of "color management" software.

Initially this software solution was introduced by a number of different manufacturers using different systems, and we wound up with a "tower of Babel" of color management. Color management varied based on which layout or image application was being used. Since this scenario wasn't going anywhere fast, a committee of concerned vendors and users was formed. The ICC (International Color Consortium) was formed to negotiate the marketing and technological problems associated with developing an application standard that would address this "color management."

When this committee was formed, the promise of ICC color management was that it would be a simple, transparent solution. When properly used, it would decrease the need for highly experienced color professionals and develop consistent cross application and device color output. This was critical, since the user base was increasingly less skilled.

Today, while we have made significant progress toward that end, the complexity of color management implementation is still viewed as akin to rocket science. When implemented properly, it is a valuable and "semi" transparent process support utility.

The basic working structure of ICC color management is based on the interaction of a number of components: profiles, Color Management Modules (CMMs), and the CMS or Color Management Solution, such as Apple's ColorSync or Microsoft's ICM2 (Integrated Color Management). Each of these components has a distinct function in conjunction with the files that will be processed using the system. The most widely discussed components currently are profiles.

Profiles are unique files that contain information describing the specific characteristics of a device. Those characteristics can include color space, white and black points, dynamic range, etc. Usually these are determined by measuring the device's output against a standard target. In the case of a monitor, a colorimeter is used to measure the values presented to the screen by a software-generated specific target. In the case of a scanner, the values from the output of a scanned target are then measured against the known values of that target. And finally, in the case of print output, a software-generated target is output and measured using a spectrophotometer and processed against the known target values to create the profile.

The structure of the profiles should comply with the structure defined by the ICC, allowing for easy distribution and use. Software to create profiles is available from many vendors, including Agfa, Itec, Southwest Software, Fuji, Heidelberg, Scitex, Kodak and Monaco Systems.

CMMs are the color transformation engines that actually do the image color conversions based on the profile information. They act as translators, or look- up tables, that convert color data from the specific characteristics of one device to another. CMMs also are available from many vendors. While the default CMM included with Apple ColorSync and Microsoft ICM originally came from Heidelberg (formerly LinoColor), there are CMMs available from Kodak, Agfa, Scitex, Imation and others. Some are available as options to the ColorSync installation process and some, such as the one from Imation, are available as a separate purchase option.

The selection of CMMs for image processing can happen manually or automatically. Users can choose to process files with a specific CMM. In addition, Apple ColorSync has an automatic option that looks inside the profile to see what CMM was used to create the profile in question. If it is available on the system, ColorSync will use the CMM to do the transformation. If it isn't on the system, ColorSync uses a default CMM.

Color Management Solutions such as Apple ColorSync and Microsoft ICM2, are the OS level system structures that facilitate the process of color management. These system level solutions replaced the early solutions provided by individual manufacturers such as those from EFI, Agfa and Kodak.

In order to fully understand color management you should appreciate the differences between calibration and characterization. Calibration can bring a device back to a standard or known configuration set. Characterization and color management can be thought of as methods to calibrate not just a device but an entire process.

Normally device calibration is done prior to characterization (profiling) to ensure that any device drift can be brought back to a center point prior to use. Characterization can be done to an uncalibrated device and still work, however. The device calibration cannot move from the state it was when the profile was made, though. This isn't the ideal method, since you may be able to achieve the widest range available from an uncalibrated device.

Apple ships ColorSync with its OS software, and along with the necessary supporting graphic arts applications, you have most of the core tools necessary for implementing a color managed workflow. The only missing piece is the necessary device characterization profiles.

As part of ColorSync, Apple has added a software tool to allow users to visually create a monitor profile. This isn't the most reliable way to manage color across devices because it is based on a visual interpretation of targets rather than one that was specifically measured. In addition, many input and output devices are now equipped with their own individual profiles. While these may be better than nothing, not all the devices that a manufacturer ships are exactly the same.

In order to create accurate profiles, which in turn will allow users to achieve the optimum results, you need good measurement tools. You will need good color profiling software, a colorimeter to create monitor profiles and a spectrophotometer for determining the output profiles. In the past few years there have been many of these measurement devices introduced to the market. They come from firms such as X-Rite, GretagMacbeth, Color Savvy and Color Vision.

Colorimeters are fairly manual in their operation. You usually mount them to a monitor using a suction cup. The Color Savvy device uses an adjustable strap to profile LCD monitors.

Spectrophotometers are available with many more options and configurations. There are manually operated and require users to measure each color patch on the target individually, which can be a long and exhausting process.

There also are a number of automated devices. They range from those that scan strips of patches that are fed into the device to others that are robotic. The robotic devices allow users to place an entire target on the bed of the device. Each patch is read automatically.

Obviously the decision to select manual or automatic options needs to be made based on cost vs. available time to read the patches. If you have more than one or two output devices, it is probably worthwhile to look at an automated solution.

All profiling software is not created equal. Most support widely used measurement devices, but check to make sure which devices are supported by which profiling program prior to purchasing either the software or the devices.

Some software packages don't come with targets. This approach can get rather expensive if targets need to be purchased separately. In addition, some software may have support for only certain targets. This can be an issue if the targets you are using don't match the type of work you are doing. While this is usually not the case, increasingly there are specially designed targets that address some of the perceived weaknesses of the historically standard IT8.7 targets. Some of the differences include the number of target patches and specially created targets that are supplied.

Some of the software, such as that from Scitex and Imation, have very specific software tools to handle individual issues, including CMYK to CMYK profile generation, while maintaining the black. Others have a complement of profile editing and evaluation tools. This is one of the biggest areas of difference.

Itec's ColorBlind was one of the first color management software packages that included extensive profile editing and diagnostic tools. As the market and demand continue to expand, however, others, such as GretagMacbeth with its ProfileMaker, have expanded their offerings to include an increasing number of editing and process support tools.

In addition to including an editing tool, each of the editing packages take a different approach to editing. Some of the tools are complex and require a good understanding of color reproduction, while others simplify the process. A new color management package from Fuji, for example, takes a multi-level approach.

There are intuitive tools for those with a minimal experience level, and tools for those who are experienced with the more sophisticated controls of drum scanner operation. You should familiarize yourself with those tools prior to purchasing anything, since most output profiles will usually need to be edited.

The whole process is based on a profile connection space (PCS). This is an intermediary color space that, in theory, encompasses the widest viewable color space available. When a device profile is created, it is created to describe the transformation from the device's color space to the reference color space (PCS). The CMM actually processes the image data from one color space to the PCS and then from the PCS to the new color space.

One of the real benefits of a good color managed workflow is that you can store and edit all of your images in RGB, the originally captured color space. This allows you to keep all of the captured information for cross media repurposing at a later time, and have the transformation include conversion to CMYK.

There are those, however, who believe that images should be stored and edited in CMYK space. ICC color management workflows can support both of these scenarios. It is important to understand that there are not just differences in the workflows, but also in the profiles used and potentially in the required application or device support.

Where to have the files transformed is a commonly found argument. Conceptually it is best to have the actual color transformation occur at the last possible minute prior to imaging, especially if you are going out to a number of different devices. However, this isn't always possible or practical. Transformation is a computational intensive process. It can be done as early in the process as creation or editing in Photoshop, as part of the PostScript print output file creation through the OS, on an image management server or as late as the RIPing stage.

Many of the Digital Asset Management (DAM) systems on the market today have added color management support to their output server options. In addition, individual color management server support can be added through the use of products such as ColorBlind Parachute. This server software allows users to individually configure output device queues. These queues will process PostScript files with embedded profile information destined for output. In addition, users can override any imbedded profile information.

Increasingly RIP manufacturers have added that same level of support into their output solutions. Basic Color Management support was included in the Level 2 PostScript specification and has been enhanced in the PostScript 3 specification. The specific levels of support vary based on the individual RIP implementation, so check to see what level of support is offered in your current or future RIP solutions.

In order to fully take advantage of color management, individual graphic arts applications need to support it. If color management implementation fails, this is usually the weak link. Part of the problem can be attributed to non-standard setups and terminology; inherent software limitations also are a factor. An example of inherent limitations would be that of Quark XPress (prior to 4.1), which didn't output color managed RGB files. While this wasn't a problem for those who place CMYK images, it did prevent those who didn't from getting color managed output without installing a specially designed XTension.

Most of the major professional publishing applications such as QuarkXPress, PageMaker, Photoshop and Illustrator support color management in their latest versions, providing the basis for a color managed workflow.

At a very basic level, color managing an individual work site is a fairly straightforward process. Once set up, it usually requires only device calibration (which should be done anyway) and occasional profile checking. While the profile of a device shouldn't change, any changes in media, ink batch colorants, and even device aging can affect the reliability of the profiles.

Managing the number of profiles in a workplace can be a challenge, depending on the number of devices you have. If there are five monitors, two output devices and one input device, there already are a minimum of eight profiles. If you add multiple workflow options such as both RGB and CMYK, there are twice as many profiles.

If you have any input from outside sources such as designers, the number of profiles can increase exponentially. In a recent study, it was estimated that the average workstation has 19.46 device profiles. However, a medium-size printer could conceivably have hundreds or thousands, depending on the customer base.

You don't really need a profile file in hand to process an image that already has an imbedded profile. The problem arises if you need to edit that image. At that point, having the original profile is the only way to ensure that you are seeing what the image originator saw.

Just in case your customers' images need work, profiles should be treated as fonts. In addition to keeping track of the profiles, if customers use a specific CMM in their systems, you will need to make sure that you have a copy of that too. CMMs and profiles should be kept with the job and stored or discarded when the job is complete. As you can see, this is where the implementation can get overwhelming.

Finally, you will have to decide where your transformations will occur based on your available configurations.

Depending on your available skill sets, there are a number of options that would enable you to move into color management. There is a growing list of qualified consultants and service organizations that can get you started. These consultants can be found through the Apple ColorSync website, as well as the websites of color management hardware and software developers.

In addition, the Graphic Arts Technical Foundation (GATF) recently started to train, test and certify color management consultants. GATF has also started a program that will allow plants to be tested and certified based on their ability to maintain certain levels of color control.

While color management can be implemented using currently available products and technology, there are still issues that need to be addressed. Two recent developments have begun to make that possible. With the release of Apple ColorSync 3.0, there is a way to set up color management settings once, at the system level, and with special API applications use those settings again.

In addition, the committee recently approved a plan that would allow "device link profiles" (a profile that was created as a combination of two output profiles) to be included in the image file. This opens the door to a number of issues including "blind transfers," a workflow in which the originator has no idea where the image will be output, and/ or the output service has no conversation with the originator. There is still a lot of work to be done to fine tune the designated PCS, align CMMs and their relationship with profiles (smart profiles with dumb CMMs, dumb profiles with smart CMMs), and a host of other issues.

The ICC meets regularly to address the many issues that are raised in the defense of the industry, the process and the individual needs of each of the member companies. Eventually we may see a day when an image can be captured without regard for the media that was used or the device on which it is output.

Ultimately color management, when implemented properly, can give additional control of the color publishing process. It can reduce the time and cost of the process. In fact, a 1999 ROI study of color management performed by Gistics Research, found that the fully burdened annual cost of implementing color management was approximately $146 per device. The study also found that the one-year return on investment ranged from 4.53 times to 23.41 times per year, depending on the type of company and work performed. This provides a substantial case for the financial benefits of color management.

There are other benefits to color management implementation. These include competitive differentiation, developing closer client relationships and facilitating cross media.

If you haven't looked seriously at color management for your individual production process, it is something that should be addressed in the short term. Postponing it will not make the process easier. The sooner you implement color management, the faster you can begin to take advantage of its many benefits.