I guess this is a rather tangential question. I just found that after Gamma correction, in RGB mode, value (1,0.92,0.92) or something like that gives very, very weak perception of the corresponding color (red). My question is that is there any special nonlinearity at the ends of the internsity curve of the pixels? If so, then the aforementioned phenomenon should be due to the inaccuracy of the display to present the color… I am using sgi workstation…
Your red appearance will be due to the different exponents applied to the mid tones for each of the color components, what else did you expect with different correction for different components? Anything not 0.0 or 1.0 will display a distinct color shift if you apply different correction exponents.
Take a mid grey for example, .5 to the power of 1 is greater than .5 to the power of 1/.92, which is what you are seeing as far as I can tell since gamma on SGI is expressed as monitor gamma not correction gamma applied.
At the ends there is no change because 0 to the power of any correction you are allowed to apply is still 0 and 1 to the power of any correction is still 1. This is exactly what gamma does, it changes the distribution of intensity between two fixed end points to compensate for non linear monitor response.