cmb said:
My understanding is that it is mainly red and blue that we see as 'colours' and outside the fovea there are scarce few green cones and that job is done by the rods (black-white) with a similar spectral sensitivity in the 'green' wavelengths. So we don't have good acuity in green, unless it is 'just' green.
On the contrary. We are most sensitive to light in the green area of the spectrum and are least sensitive in the blue area, on account of the relatively few S-cone cells compared to M and L cones (S sees blue, M sees green, L sees red). Note that L cones actually have peak sensitivity in the yellow area of the spectrum, not red. But we often say that the L cone is the 'red' cone for reasons of simplification.
Rod cells are actually desensitized after exposure to bright light and do not play any role in color vision. My understanding is that they aren't even connected to the areas of the visual system and the brain that are responsible for processing color, but I could be wrong.
cmb said:
A broadband colour which has a strong red peak we see as red, a strong peak in blue we see as blue, but a strong peak in green we see as white.
That is true, but that's only because we have cone cells sensitive to wavelengths on both sides of green, not because we aren't sensitive to green or have few M cones. If we had another cone cell that was sensitive to wavelengths beyond blue, then a blueish-white object, like a very hot star, would probably look more violet and less white (or perhaps some other color that might result from having that extra cone cell).
The peak wavelength isn't actually what matters. It's more about how intensely each set of cone cells are stimulated by incoming light. This is why a small set of individual wavelengths can appear the same color as a broad spectrum source. Fluorescent lights are a perfect example of this. They commonly emit in only a handful of narrow wavelength ranges, yet they appear to be the same color as a broadband source.
For a thermally-based broadband light source the 'shape' of the spectrum (how intense each wavelength is) is set purely by its temperature, as is the peak. So while the peak may appear to be what sets the color of the source, it's not. It's just that both the peak and the shape of the spectrum are set by the temperature and cannot vary independently of each other.
cmb said:
(The cones also are particular on how much light is falling on them, needs to be enough but not too much. When the light is low the cones don't work well, the more sensitive rod cells take over and we see in black and white in the dark. You might not have noticed. Walk into a dark room that you can't see in straight away, let your eyes adapt, and notice that you can't see any colours.)
Yes, I am aware of how dark adaption works and the fact that low-light vision is purely black-and-white.
