Digital Color Management: Encoding Solutions, 2nd Edition

13

Color-Encoding Data Metrics

To this point, the first principal attribute of a complete color encoding specification, the color-encoding method, has been discussed. The second principal attribute, the color-encoding data metric, defines the color space and numerical units in which the encoded color values are expressed. Although it seldom is necessary to restrict a particular encoding method to a single data metric, it often is convenient to do so. Most color-encoding methods do in fact have a particular data metric associated with them.

Many data metrics currently are in use, and attempts frequently are made to identify which of them is “best.” Almost every conceivable data metric seems to have its dedicated supporters and its equally dedicated detractors. Our opinion is that the particular requirements of a specific color-imaging system will dictate the most appropriate data metric for that system. In practice, systems vary widely in their principal uses, productivity requirements, quality requirements, and computational capabilities. Therefore, there is no single “best” data metric for all systems.

That does not mean, however, that the data metric for a given system can be chosen arbitrarily. Among other things, the data metric affects the efficiency of the color encoding and the requirements for signal processing. The data metric also will affect other factors, such as the visibility of quantization and compression artifacts. In virtually all applications, there are engineering ...