|
|
|
|
|
|
A White Paper on Color Management: Framework for Imation Color Solutions September, 1996
Presented By:
1.0 Introduction The term "color management" has appeared in graphic arts and computer trade literature with exponentially increasing frequency over the past few years. This phenomenon is reminiscent of how the term "desktop publishing" exploded onto the scene in the 80's. Back then, the jury was still out about whether desktop publishing could really work, and whether it could really pay for itself. Debates raged about whether the software applications and digital hardware of the time could ever close the gap on the expensive high-end systems, or whether they were merely sophisticated and resource-consuming toys. Nonetheless, the conviction among those on the ground floor of desktop publishing was that it would eventually supersede the reigning high-end systems. Indeed, the high-end systems slowly became regarded as expensive and inefficient. This occurred as desktop software applications, RIPs (raster image processors), and digital hardware (input and output) reached an acceptable level of quality. The software applications and RIPs found ways to utilize an existing communication framework, PostScript™, in a way that was an adequate substitute for traditional high-end workstation proprietary technologies. The PostScript standard continued to evolve in response to various issues raised by the graphic arts industry, especially with regard to color quality. Today, we are seeing the graphic arts industry respond to digital color quality needs with the evolution of color management systems and a standardized color management framework for these systems. The International Color Consortium (ICC) was formed in 1993 to develop one communication framework for all color management systems. The ICC is currently developing a communication mechanism for digital color. It is our belief that the ICC standard, while not completely finished yet, is close to becoming a communication mechanism for color that is adequate for high-end color applications as well as for the consumer market. We believe the benefits of color management will cause it to become the norm of the late 90's, much like desktop publishing became the norm of the early 90's. Imation is committed to helping the ICC and the graphic arts industry make color management a fundamental and trusted part of the graphic arts workflow.
2.0 Background on Color Management and the For those not familiar with color management or the ICC, a brief review might be helpful. 2.1 Evolution of Color Management and the ICC Color management is a systematic approach to performing color conversions from image creation or capture to final output. It involves the creation of digital "profiles" in which color values are measured; the profiles are then converted for device color fidelity by Color Matching Module (CMM) software. Color management is based on color science technologies that have been around since the early 1920s. However, it's only in recent years that color science has been allowed to closely intertwine with real life graphic arts workflow. A number of factors contributed to the emergence of color management systems in the graphic arts workflow: 1) The advent of low-cost spectrophotometers 2) The availability of relatively inexpensive high-speed computer processing 3) The development of color communication standards 4) The support of these color communication standards in commonly used software applications 5) The heightened awareness and education throughout the industry of the existence of these technologies and their potential impact on digital color workflow As color management systems evolved and multiplied, the need for standardized communication and processing of color values became apparent and members of the graphic arts community formed the ICC. 2.2 International Color Consortium (ICC) The creation of the ICC marks the beginning of the serious color management movement outside of the proprietary high-end systems. It was created to establish a color management framework for communicating and processing color calibration information in a standardized manner. The ultimate goal of the ICC is to provide a way for color management to occur seamlessly across multiple applications and platforms. The advantages of such a framework are obvious for both venders and users. For venders, it means they do not need to develop formats unique to software applications such as Adobe PhotoShop™, QuarkXPress™, etc. or create operating system drivers for systems such as Apple™, Sun™, etc. Rather, the color management framework will enable the creation of standard profiles which can be read and understood by multiple applications and operating systems. For users, it means they can at least begin to consider "plug and play" color that simply works and works simply. The color management framework allows a user the option of configuring off-the-shelf hardware and software capable of achieving critical color results. NOTE: The ICC is currently moving toward a goal of "plug and play" color, building upon the foundation in Color Sync 2.0. However, users should be aware that the ICC specification is still open to certain vender-specific interpretation in some key areas such as gamut mapping. This means that mixing and matching profiles from different venders is risky at this point in time. Imation, as a committed member of the ICC, strongly supports the concept of the ICC standard because we see it as a winning situation for venders and users alike. That's why Imation has taken a leadership role in the ICC and is focused on achieving consensus and agreement among ICC venders to further define the ICC specification, while still enabling vender-specific enhancements. 2.3 Essentials of Color Management There are two crucial elements of a color management system: the creation of a digital profile and the conversion of that profile for color fidelity from device to device. 2.3.1 What is a Profile? A profile is a description of a digital color device or system, usually saved as a file. A digital color device is a stand-alone piece of equipment for capturing, displaying, or outputting color images. A digital color device is fundamentally digital in nature, such as an RGB scanner. A digital color system, on the other hand, contains components that may not be fundamentally digital in nature. For example, a printing press may use plates exposed by films that were created using non-digital screening methods. However, if the plates were exposed using films generated on an imagesetter, one can regard the imagesetter/film/plate/press system as a digital system. As such, a single profile can be created which describes the color of the final image printed in relation to the original digital image file. How is a profile made? A profile is created by performing physical measurements with a color measuring device on a select group of color patches. These patches are either physically created and captured as a file (in the case of an input device) or output (in the case of an output or display device). A mathematical description is then calculated which relates the color values of the device (such as RGB or CMYK) with device-independent color values such as CIELAB. If a profile is correctly made, it should be able to predict accurately the CIELAB value you will measure for any set of values of RGB or CMYK.
2.3.2 What are Color Matching Modules (CMMs)? Once the values are measured, the data needs to be processed. This is done by a Color Matching Module (CMM). A CMM is the software that is used to connect together and convert the color data from device to device in order to maintain color fidelity using the device profiles described above. ICC compatible applications read ICC profiles for source and destination in association with images and then direct the operating system to process the image data using the indicated profiles. The actual processing of the image data via the profile information is done by the installed CMM in the operating system.
NOTE: One can have "sophisticated profiles" and a "dumb" CMM or very basic profiles and a "smart" CMM. The most common implementation of ICC at this time is to have a "dumb" CMM and "smart" device profiles. That is, the technologies such as gamut mapping are built into the profiles rather than into the CMM. Since the core differentiating color technologies reside in the profiles rather than in the CMM, there can be significant issues in mixing profiles created by different venders. For this reason, Imation strongly recommends that for high-quality applications customers use profiles created by the same profile-generating application. 2.4 Critical Components of Successful Color Management In the previous sections we discussed a basic definition and description of color management and the ICC. Now let's consider the requirements for color management to work simply and effectively. 2.4.1 Device-Independent Color Space In order for a CMM to work, an appropriate device-independent color space has to be used. Currently, the most commonly used color space is CIELAB. CIELAB uses the "RGB" responses of the eye, which are known as XYZ, and converts them into values which are more practical for the actual quantifying of color. CIELAB does this in two ways: 1) it relates all XYZ colors to the XYZ values of a neutral white, and 2) it scales the values according to equal visual steps, similar to the steps used in color systems such as Munsell. The values calculated using CIELAB are as follows: L* indicates on a scale of 0 to 100 the appearance of a color ranging from black -> white. a* indicates on a scale of -120 to +120 the appearance of a color ranging from green -> red b* indicates on a scale of -120 to +120 the appearance of a color ranging from blue -> yellow By definition, a "perfect" white would have the values L*=100,a*=0,b*=0 and a "perfect" black would be L*=0, a*=0, b*=0. The ICC also offers a modified CIELAB which uses the actual paper or monitor white as the neutral reference rather than a "perfect" white. Imation color technology utilizes a proprietary L*a*b* color space that is very effective in translating between colors of different systems with differing white points. Our philosophy is that color spaces are like software--new revisions and modifications can be developed to meet the needs of new applications while maintaining forward compatibility. This ability to maintain color fidelity between systems of differing white points means that it is now possible to obtain an optimal match between printing presses with white, grey, or yellow paper bases and the pre-press proof, such as a 3M™ Matchprint™proof, which is signed off on by the customer. 2.4.2 Accurate Profiling As discussed in section 2.1, it is critical to have profiles that give an accurate description of the color behavior of digital color devices. Most profile- generating software will create profiles that accurately predict the color of the patches used to create the profile. For example, if the profile-generating software requires the user to measure a patch with tint values CMY = 50%,50%,50%, there is a good chance that the profile will accurately predict the measured values of L*a*b* for that particular patch. However, this does not guarantee that the profile will accurately predict the color of a patch with values CMY = 50%,65%, 45%. Therefore, a good test of a profile is whether it predicts the L*a*b* values for a random set of patches which are not the same as the patches used to generate the profile. A good test of the accuracy of a color management system as a whole is whether the device profiles and CMM together allow you to scan a given printable color such as a mid-tone grey, and automatically reproduce that same color using digital output without human intervention in the conversion process. 2.4.3 Gamut Mapping If one has an effective device-independent color space and accurate profiles, it is easy to convert colors from a source device to a destination device if the colors of the destination device all lie within the colors of the source device. For example, if the original source is a SWOP proof and the destination is a commercial press with very saturated colors, one can achieve a one-for-one conversion to make the image printed on the commercial press resemble the less saturated colors of the SWOP proof. But suppose the source device is a transparency on a scanner containing highly saturated colors? The printed output can never achieve the exact same appearance. Gamut mapping is the technology that allows the automatic scaling and optimization of the destination image's appearance in order to best match the appearance of the source image. 2.4.4 Graphic Arts Images and Black Channel Reproduction In the previous sections we considered the importance of having a valid device-independent color space, accurate profiles, and good gamut mapping technology. In the graphic arts however, the values of YMCK in an image are used for more than just color reproduction purposes. The black channel (K) in particular is used for accentuation and detail, especially in areas of the image where one may wish to minimize the softening effects of misregistration between CMY separations. In the Imation™ Rainbow™color proofing system, we not only optimize color reproduction, but also maintain the integrity of the black channel information. Thus, if the original image data has a value of K that is high or low, the reproduced Rainbow™proof will have a value of K that is also high or low, without loss of color fidelity. Although the ICC framework can be used in such a way so as to preserve the K channel information, currently few systems, if any, actually maintain this aspect of graphic arts reproduction.
Copyright 1996 Imation. All rights reserved. 
|
