[argyllcms] Re: Camera input profiles and gamma correction.

On Sat, Mar 28, 2009 at 3:44 AM, Paul Hunt <prhunt@xxxxxxxxxxxxx> wrote:
> I have built camera profiles using ufraw and argyll.
> I developed the raw image of my Wolf Faust C1 target twice, once using
> gamma/linearity settings of 0.45/0.1 (gamma 2.2), the other using
> gamma/linearity settings of 1.0/0.0. The developed images were saved as 16bit
> TIF.
> I take it, therefore, that the first image was gamma corrected, and the second
> was not gamma corrected.
> After measuring the target images with scanin, I created LUT profiles with
> colprof, using these parameters - -qh -al -bn
> I tested the new profiles in ufraw.
> Results:
> Using the profile made from gamma-corrected target, images appear linear
> (histrogram skewed left) unless I use ufraw gamma correction (ie gamma and
> linearity sliders at 0.45 and 0.1).
> Using the profile made from the non gamma-corrected image, images appear gamma
> corrected without ufraw gamma correction (ie gamma and linearity sliders at 1
> and 0).
> Can someone help me better understand what is happening?

I've already explained this on the UFRaw forums:

"So if UFRaw doesn't do gamma correction (gamma 1 = no correction), then the
profile will do the full correction"

"But please note... A profile will always do gamma correction! Since
it always tries to
match the values in the reference file. However, if your image is
already gamma corrected,
the corrections done by the profile will the negligible... "

Pascal de Bruijn

Other related posts: