Re: [ft-devel] [GSoC] ftinspect

[Philipp Kerling]

Am Donnerstag, den 30.03.2017, 12:23 +0200 schrieb Nikolaus Waxweiler:
> > As far as gamma is concerned, it seems that Qt does not support the
> > necessary blending calculations [1] that were implemented in the
> > old
> > utilities. So: Should the gamma blending work exactly the same as
> > before (meaning that the blending code from gblender.c needs to be
> > ported)?
> 
> I'm not familiar with Qt's painting system, does
> QPainter::CompositionMode_SourceOver do any gamma correction? Or is
> this what the bug report you referenced is about? 
It does not. SourceOver is just plain alpha blending. The bug report is
about providing gamma-corrected blending in some way or another, not
necessarily integrated into SourceOver.

> I think that since ftinspect is to be a reference application, it
> should support both sRGB and pow()-based gamma curves. The sRGB curve
> would be more correct as the default (almost everything you see on a
> screen is supposed to be in sRGB color and gamma space), but reality
> testing at Adobe suggests 1.8.
> 
> If Qt can't handle this out of the box, I guess you'd have to
> implement
> custom pixel blending code anyway (custom composition mode?), might
> as
> well take existing code then.
Yep, that's the plan. Also, the existing code does support the sRGB
gamma curve (you can check it out in ftgrid or similar by setting gamma
to the special value zero), so that would naturally be ported, too.