Actually the article is confused on this very point. I think it's actually implied that it's sampled in HSV and not weighted for saturation or value.
Edit: his blog entry implies RGB, and it looks like the average color is actually brownish grey, which would be highly unsurprising to anyone who has mixed paint.
I just tried my own suggestion. For a random sampling of wavelengths from 390 to 700, I converted each to RGB using this function http://stackoverflow.com/a/3407960 Then I averaged the R, G, and B values. The result is 144, 122, 90 or #907A5A which is... a dim brown color, with a "hue" of 36 which is very orange.
Edit: undid sRGB conversion since I think i did it wrong.
Comments
Actually the article is confused on this very point. I think it's actually implied that it's sampled in HSV and not weighted for saturation or value.
Edit: his blog entry implies RGB, and it looks like the average color is actually brownish grey, which would be highly unsurprising to anyone who has mixed paint.
I just tried my own suggestion. For a random sampling of wavelengths from 390 to 700, I converted each to RGB using this function http://stackoverflow.com/a/3407960 Then I averaged the R, G, and B values. The result is 144, 122, 90 or #907A5A which is... a dim brown color, with a "hue" of 36 which is very orange.
Edit: undid sRGB conversion since I think i did it wrong.
Good thought, and I think that the warm bias in digital cameras (and human tastes) probably contributes as well.