[argyllcms] Re: Use of calibrated display to profile digital camera

Graeme Gill wrote:
> ...and therefore may not represent the spectral composition that you
> want to capture in the real world
>
> (but how do you know what that will be anyway ?).

That's actually the reason why I got doubts in the mean time, whether a
CLUT profile is really too useful for camera characterization, or
whether a matrix model isn't possibly still the better choice for a
general purpose camera profile, which does not make strict assumptions
about the spectral properties of the objects being captured [assuming of
course, that we're dealing with linear raw RGB from the camera sensors,
and not with a weird RGB space, already processed by the camera
arbitrarily].

Regards,
Gerhard

Other related posts: