[argyllcms] Re: A beginner's questions

  • From: "Preben Soeberg" <prsodk@xxxxxxxxx>
  • To: <argyllcms@xxxxxxxxxxxxx>
  • Date: Sat, 2 Feb 2008 12:33:39 +0700

> -----Original Message-----
> From: argyllcms-bounce@xxxxxxxxxxxxx [mailto:argyllcms-
> bounce@xxxxxxxxxxxxx] On Behalf Of Graeme Gill
> Sent: 2008-02-02 08:04
> To: argyllcms@xxxxxxxxxxxxx
> Subject: [argyllcms] Re: A beginner's questions
> William Winfield wrote:
> > Anyway, this seems to be a bug in the dispcal -o option.
> Sorry, what bug is that ? Can you provide an example ?

I have checked the data in vcgt and RGB tables and can not see any problems.
Smooth transitions, all as expected. The coordinates of the primaries are as
expected from EDID data.

But is there a bug in Photoshop?

I have recently started to use the profiles created with the -o option in
dispcal as well as profiles created with -as option in profile. I did this
because the LUT profiles showed some annoying banding and artifacts.

As a result of these matrix/shaper profiles, I got a much worse problem,
i.e. that when I made pictures in PS7, they appeared much darker, than when
I later look at them in IE7 or the image viewer. The tone also appeared
If I applied an inverse gamma 1.20 in PS, the images looked almost like in
the MS programs.

I have returned to use LUT profiles again and these brightness/contrast
differences have disappeared.
I have read somewhere that Photoshop can give banding when using LUT
profiles, presumably because Adobe has made some compromise in favor of
speed. Making sensible parameters to targen has reduced the artifacts.


Other related posts: