
How Your Brain Builds Color
Three overlapping cone signals become useful only when the visual system compares them

Three overlapping cone signals become useful only when the visual system compares them
AI-assisted edition · Educational review score 96%
Three overlapping cone signals become useful only when the visual system compares them
Created by Bob · AI-assisted and reviewed before publicationMost daylight color vision begins with three cone classes, conventionally called S, M, and L for their relative short-, medium-, and long-wavelength sensitivities. Their response curves overlap broadly. A single cone only changes its response according to how many photons its pigment absorbs; it can't tell whether a strong response came from a well-matched wavelength or from more light at a less effective wavelength.
Color information appears only when the nervous system compares activity across cone classes. This is why calling them blue, green, and red detectors is a useful shortcut but a misleading picture of how any one cone works.

Retinal circuits reorganize cone activity into comparisons. Some pathways contrast L- and M-cone signals; another contrasts S-cone activity with a combination of L and M. Separate channels also carry light-dark structure. These opponent signals help distinguish spectral differences from overall intensity and emphasize boundaries between surfaces.
Later processing combines them with spatial context, adaptation, illumination clues, and memory. There is no tiny colored picture traveling along the optic nerve. The brain receives several structured differences and uses them to infer stable objects and surfaces under changing light.

Because three cone classes reduce a detailed spectrum to three response totals, physically different spectra can sometimes produce the same cone-response pattern. Such matches are called metamers. A display exploits this compression: by adjusting three primary lights, it can reproduce many response combinations that ordinary surfaces produce, even when its spectrum is entirely different.
But three primaries don't reproduce every visible color, and a match for one observer or lighting condition may fail for another. Color is neither an arbitrary fantasy nor a property simply collected from an object. It's a repeatable relationship among light, receptors, neural comparisons, and context.

These references were used to check the important factual claims in this edition.