Why Can Staring at Red Leave a Cyan Afterimage?

Stare at a bright red square for several seconds and then look at a white wall. A brief cyan-blue outline may appear. The wall does not contain that colour; the afterimage comes from a temporary change in the visual system's sensitivity.

The retina has three classes of cone photoreceptor, each with a different range of spectral sensitivity. Sustained red light adapts the cones and pathways that respond more strongly to longer wavelengths, temporarily reducing their response. When you shift to a white surface, all cone classes receive light, but the recently adapted channel responds less than the others. That altered balance is experienced as a cyan-blue signal. The shape remains because adaptation occurred in the retinal region that viewed the original figure.

This is often summarised as “red produces its complementary colour”, but the brain is not mechanically flipping a colour wheel. The precise hue and strength depend on the original colour, brightness, viewing time, background and individual variation. Recent experiments found that a cone-adaptation model predicted a broad range of colour afterimages better than a simple red–green and blue–yellow opponent rule.

An afterimage is not a photographic copy stored by the eye. It is a contrast created when short-term adaptation meets a new background. It usually fades quickly because the visual system continues adapting to the white field now in view.

https://www.ncbi.nlm.nih.gov/books/NBK11537/
https://pmc.ncbi.nlm.nih.gov/articles/PMC12579601/