Mixing colors

Suppose that we have three pure sources of light, as in that produced by an LED, in red, blue, and green colors. We have already discussed how to produce yellow by blending red and green. In general, most perceptible colors can be matched by a mixture of three. This is called trichromatic theory (or Young-Helmholtz theory). A set of colors that achieves this is called primary colors. Mixing all three evenly produces perceived white light, which on a display is achieved as RGB $ =(255,255,255)$. Black is the opposite: RGB$ =(0,0,0)$. Such light mixtures follow a linearity property. Suppose primary colors are used to perceptually match power distributions of two different light sources. If the light sources are combined, then their intensities of the primary colors need only to be added to obtain the perceptual match for the combination. Furthermore, the overall intensity can be scaled by multiplying the red, green, and blue components without affecting the perceived color. Only the perceived brightness may be changed.

The discussion so far has focused on additive mixtures. When mixing paints or printing books, colors mix subtractively because the spectral reflectance function is being altered. When starting with a white canvass or sheet of paper, virtually all wavelengths are reflected. Painting a green line on the page prevents all wavelengths other than green from being reflected at that spot. Removing all wavelengths results in black. Rather than using RGB components, printing presses are based on CMYK, which correspond to cyan, magenta, yellow, and black. The first three are pairwise mixes of the primary colors. A black component is included to reduce the amount of ink wasted by using the other three colors to subtractively produce black. Note that the targeted colors are observed only if the incoming light contains the targeted wavelengths. The green line would appear green under pure, matching green light, but might appear black under pure blue light.

Steven M LaValle 2020-01-06