[argyllcms] Re: Generate a reference .ti3
- From: Gerhard Fuernkranz <nospam456@xxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Sun, 14 Jun 2009 13:50:24 +0200
Wallden Magnus wrote:
> -> and i have a printer profile , for some paper ok, and some a
> little bit less, I think the "jump" is to big.
> -------------------
> now I want to improve some of the printer profile. The workflow is
> described in the Refine manual text under "Discussion"
> I use refine and colprof , but the result wasn't good.
Basically, refine is rather intended for a different use case :
"Typically the charts would be printed on a target system (the one
being emulated, say a printing press), and the proofing system (the
one that is being profiled). The abstract profile that refine
produces will be a correction that makes the proofing system behave
more like the target."
I guess you don't want to setup a proofing system, do you?
Possibly the B2A tables of a single printer can nevertheless be refined
too, but if you run refine shortly after creating the profile (so that
no aging of the printer has happened in the meantime) I would not expect
much (if any) improvement (except possibly if the printer profile was
initially created from too few patches, or if you just refine a very
particular region (e.g. skin tones, etc.) of the color space). But
generally, I think any additional patches you would have to print and
measure for each refine pass are likely better invested into the target
you use to create the profile.
I can only speculate, but in your case I guess the problem is possibly
rather the (in)accuracy of the (scanner-based) color measurements. If
the measurements are wrong, the profile will be wrong too, and any
refinement will still give wrong results either. A soft proof on the
monitor will usually look fine in this case (since the same systematic
error is included in both, the A2B and B2A tables of the profile, and
thus cancels out), though the actual print will show color casts.
Regards,
Gerhard
Other related posts:
