[argyllcms] Testing dispcal

  • From: Richard Hughes <hughsient@xxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Sat, 17 Nov 2012 15:17:15 +0000

I'm trying to build some dogtail scripts to test color management GUIs
on Linux. Part of this is testing profile generation too.

Testing targen and colprof are pretty easy, and dispwin can even be
tested using pre-canned values from a program that injects -C and -M
and returns a set number of values.

I'm having problems testing dispcal as it seems to use some kind of
iterative algorithm to set the gamma curves, and also needs access to
the device for each sample and bases the next request on this
simulated value. This makes testing very hard.

Is there a way to feed the ti3 file into dispcal and for a data file
containing the VCGT data to pop out? I.e. a non-iterative version. It
doesn't matter if we have to use 500 samples, it's just for a
dedicated test VM.

Thanks for any ideas,

Richard

Other related posts: