File talk:Cones SMJ2 E.svg
Shai mach[edit]
I believe this plot is misleading in two ways:
First, it claims to be adapted from Stockman et al. Vol. 10, No. 12/December 1993/J. Opt. Soc. Am. A 2491). However, it would appear the adaptation was not sufficiently careful, as it missed the UV behavior, see figure 12 in above reference.
Second, and more importantly, it does not accurately reflect reality, and fails to account for the following phenomena: We know the shortest wavelength on the rainbow appears purple, as does and equal combination of red + blue (RGB 128,0,128). This is inconsistent with the above plot.
Instead, see discussion on StackExchange. The correct plot should be http://i.stack.imgur.com/z3dtf.png, which is referenced to Bowmaker, J.K., & Dartnall, H.J.A. Visual pigments of rods and cones in a human retina. Journal of Physiology, 298, 1980, 501-511 figure 2, as it allows us to understand how an RGB-based screen can generate a color which appears as purple. — Preceding unsigned comment added by Shai mach (talk • contribs)
Shai mach: Nyet, graph is fine. It's a good match of http://www.cvrl.org/cones.htm, linear Energy. Human vision does not automatically match LMS to RGB, but has more layers of weighing and opponent color process after that. In addition, Stockman et al.'s later review notes:
As Fig. 11a makes clear, MSP is of little use in defining cone spectral sensitivities except close to the photopigment λmax. The large discrepancies between MSP and other estimates of cone spectral sensitivities arise because of the small signal to noise ratio of the MSP measurements.
