What is the connection between cone sensitivities and the perception of violet?

[hl_world]

Is violet the colour we get when only the blue cones are stimulated or is there some sort of "octave" effect with light where 400nm light shares properties with 800nm (near IR) light and the red cones also pick up photons giving a purple colour?

 

[Pythagorean]

There are no octaves in seeing. Our brain doesn't store the information about tones the same way it does color.

Dogs were actually found to store smell similar to the way we store tones.

 

[Proton Soup]

http://en.wikipedia.org/wiki/Color_vision#Physiology_of_color_perception

 

[hl_world]

Thanks. So makes ρ cones sensitive to shorter wavelengths and why doesn't this show in spectral sensitivity charts as a 2nd peak?

 

[Proton Soup]

Normalized response spectra of human cones, S, M, and L types, to monochromatic spectral stimuli, with wavelength given in nanometers.

it would take more digging than i want to do, but normalized means that you shouldn't assume that each of those response curves has the same amplitude. also, given that those curves appear roughly gaussian ("bell" curve), the sum should also be gaussian. and indeed, the spectral response is roughly gaussian, [STRIKE]with slight weighting towards blue.[/STRIKE]

http://en.wikipedia.org/wiki/File:Eyesensitivity.png

edit: hmm, i dunno. looking at area under the curve, it appears that there is more area under the curve to the right of the peak. which i guess makes more sense with two receptors on that side.

 

[hl_world]

it would take more digging than i want to do, but normalized means that you shouldn't assume that each of those response curves has the same amplitude.

Well I prefer a graph like this:
http://homepages.inf.ed.ac.uk/rbf/CVonline/LOCAL_COPIES/OWENS/LECT14/cones.gif
Where the sensitivities of the cones are relative. But no apparent 2nd red peak near the 400nm mark.

 

[Proton Soup]

Well I prefer a graph like this:
http://homepages.inf.ed.ac.uk/rbf/CVonline/LOCAL_COPIES/OWENS/LECT14/cones.gif
Where the sensitivities of the cones are relative. But no apparent 2nd red peak near the 400nm mark.

well, there doesn't have to be a peak, it just depends on how the curves add up. and i think i was wrong before about them adding up to form another gaussian curve. i was thinking of http://en.wikipedia.org/wiki/Sum_of_normally_distributed_random_variables" seems to address the situation here, and the difference in means may actually be close to 2 s.d., but then the amplitude of the blue isn't comparable, either. so, yeah, the blue cones would cause another peak in the spectral sensitivity at the blue end, if only their sensitivity were higher.

or maybe I've completely missed your point. it certainly does appear that stimulation of red and green cones would be almost zero for color vision at the shortest wavelengths (violet).