[Leica] On camera gamuts and the M8

[reid at mejac.palo-alto.ca.us (Brian Reid)]

Mark,

Thanks for taking the time to write that. It's illuminating.

I have a question.  In the sentences following this:

> You could determine the camera gamut via the following experiment:
>  take a tunable source of monochromatic light....

aren't you making the assumption that the camera sensor hardware is 
distributive? For example, aren't you assuming that (within clipping limits) 
the Red sensor output to monochromatic light colors a and b satisfies
        R(a+b) = R(a)+R(b) ?

I have no idea if sensors actually behave this way or not, but given that 
they are made out of physical materials, I have a sense that R(a+b) is 
probably not equal to R(a)+R(b).

I would also fear that
        (R(a)+R(b))/(G(a)+G(b))
is not equal to
        R(a+b)/G(a+b)
in other words, that there is a color shift in the response to one color 
when you add another color to the mix. I guess I would call this property 
linearity.

If sensors really do get both of those right, then the world is a kinder 
gentler place than I am used to. And if they do not, then I'm not even sure 
I know what "device gamut" means, other than being a theoretical bound on 
the output of the sensor.

Brian Reid