Overview

Classical color science, explained in the previous chapter, establishes the basis for numerical description of color. But color science is intended for the specification of color, not for image coding. Although an understanding of color science is necessary to achieve good color performance in video, its strict application is impractical. This chapter explains the engineering compromises necessary to make practical cameras and practical coding systems.

Video processing is generally concerned with color represented in three components derived from the scene, usually red, green, and blue, or components computed from these. Accurate color reproduction depends on knowing exactly how the physical spectra of the original scene are transformed into these components, and exactly how the components are transformed to physical spectra at the display. These issues are the subject of this chapter.

Once red, green, and blue components of a scene are obtained, these components are transformed into other forms optimized for processing, recording, and transmission. This will be discussed in Component video color coding for SDTV, on page 301, and Component video color coding for HDTV, on page 313. (Unfortunately, color coding differs between SDTV and HDTV.)

The previous chapter explained how to analyze SPDs of scene elements into XYZ tristimulus values representing color. The obvious way to reproduce those colors is to arrange for the reproduction system to reproduce those XYZ values. That approach works in many applications of color reproduction, and it's the basis for color in video. However, in image reproduction,...