Re: [Gegl-developer] [Gimp-developer] Don't make an architectural mistake based on a groundless premise

[Elle Stone <ellestone ninedegreesbelow com>]

On 10/07/2014 07:32 PM, Simone Karin Lehmann wrote:
> In babl/unbounded sRGB world your input space and output space differ
> in a very important aspect to the PCS space unbounded sRGB:
> unbounded sRGB can have values < 0 or > 1, input and output don't.
> Whereas in ICC / XYZ world input, output and PCS have well defined boundaries,
> namely 0 and 1 and well defined clipping, so everything is 0 <= x <= 1.

Display-referred editing does require bounded RGB channel values and 
will require an incredible amount of hacks to work within the proposed 
"sRGB as PCS" architecture.

This picture illustrates why gamma slider adjustments only make sense 
when done on display-referred channel values:
http://ninedegreesbelow.com/gimpgit/gimp-srgb/data-container/red-car-after-levels.jpg

The image on the left is a ProPhotoRGB image. The image in the middle 
was given a Levels gamma slider adjustment in the ProPhotoRGB color 
space. Results are normal and expected. The image on the right was 
converted to unbounded sRGB and given the same Levels gamma slider 
adjustment. Results are quite abnormal and unexpected.

This article explains why gamma slider adjustments requires values that 
are bounded by the range 0.0 to 1.0:
http://ninedegreesbelow.com/photography/unbounded-srgb-levels-gamma-slider.html#gamma-slider-mathematics

Gamma slider adjustments should be added to the official list of editing 
operations that require a hack in order to work in "sRGB as PCS".

With respect,
Elle Stone