I don't disagree that UV and IR exposure may be important. Variations in UV exposure are thought to account for changing rates of myopia, for example. There may

I don't disagree that UV and IR exposure may be important. Variations in UV exposure are thought to account for changing rates of myopia, for example. There may be non-image-forming receptors with different spectral sensitivities for circadian rhythms. However, the structure and function of circuits involved in color representations has been studied to death, and it overwhelmingly points to a tristimulus model where the activation of the S/M/L cones matters, rather than the complete power spectrum of the illuminant. The sensitivity of rods and cones has measured measured with exquisite sensitivity, both behaviorally and by directly recording their electrical activity. Many downstream neurons in visual cortex get their input from individual L/M cones (parvocellular pathway). The others (magno, konio) have a fairly simple mix of inputs from a simple, spatially organized combination of the cones. In some cases (especially within the retina), the individual fibers have been traced and mapped. There is a hell of a lot we don't know about the brain, but the very early representation of color isn't one of them. If you've got sources saying otherwise, I'd love to see them.