Yes, treating gama corrected stuff as linear is a very common misdemeanor, but
obviously it is not such a big problem, otherwise people would already be
running with guns after the authors of (most of the) imaging software!

You can see from the link, that the author had some hard work finding/creating
examples that make the error visible.

I do find it very interesting, that commiting such an crude mathematical sin,
once can still get results, that even top visual artists seem to swallow
without indigestion!

But again, since we almost never know the exact mapping curve used in the
camera, we can't exactly convert to linear anyway.

Also, if you do no interpolation, but use just "nearest neighbors" this
problem does not occur!

So I think, for the time being, correcting major flaws, as the ones you found
in balanc0r and levels, should be the more important goal.

Marko Cebokli

P.S. the "original sin" is usually already done in the camera, where the
Y'CrCb values are matrixed from gama corrected RGB, so there is no salvation
for us!




On Friday 21 September 2012 16:32:54 Steinar H. Gunderson wrote:
> On Fri, Sep 21, 2012 at 04:20:31PM +0200, Marko Cebokli wrote:
> > Many plugins, like geometry transforms, blurs, sharpens, etc. do not need
> > to worry about colorspaces, and are in this respect "universal".
>
> Geometry transforms should certainly worry about colorspaces (in the sense
> of the used gamma, at least). Blurs, too.
> See http://www.4p8.com/eric.brasseur/gamma.html
>
> /* Steinar */