[argyllcms] Re: printtarg with custom ti1

  • From: Elena [service address] <1007140@xxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Sat, 12 Feb 2011 20:31:59 +0100

Hello Marco

Allow my comment here, because I recently stumbled across the very same
problem. The quick answer is that there's no "clean" way to do it yet,
also I was asking kindly to support some option in targen to add custom
CMYK (or RGB) patches taken maybe from an external list file - you may verify
that in some message of mine one month or so ago.

To achieve that - but note that I'm very unsure on the effectiveness of
adding custom patches - I had to write some very dirty custom code to batch
estimate (by 4D interpolation) the XYZ value of a CMYK patch from those
limit values already present in a fresh .ti1 file [sorry my stuff is not
portable and won't run on your machine]

Those XYZ values are a first aid to check during chart readings against possible
reading errors or line mismatch. They're estimations the first time, of course.

I understood that Graeme is preferring this standard approach:
-print a testchart with randomly spreaded points (they should better cover,
 statistically, the device gamut and avoiding resonances)
-measure, generate a first profile, then generate a second testchart passing
 the previous profile to targen so it will be able to better place the test 
 in device space (see targen options)
-iterate if needed

Expecially when working with CMYK, I suspect that it's not so trivial and
intuitive to select problematic patches to add at will in order to improve
results, there're too many combinations involved. It should be (imho) colprof
itself to possibly suggest ranges of CMYK combinations needing a better 
But it's not the case yet.


> Hi all
> I'm new to color management and color stuff in general.
> I'm trying to generate a custom CMYK tif target test chart with values that
> I consider problematic for my printing system.
> I think that I could edit the ti1 file because automatic choosen values
> doesn't fit my needs
>  and give them to printtarg but I faced the issue of XYZ values in the ti1
> file.
> I think I could remove them but I got a
> printtarg: Error - Input 'pro2.ti1' file doesn't contain field XYZ_X in
> first table
> what is the purpose of XYZ values in a target file ?  can I fake them
> probably not and if no why?
> thanks in advice.

Other related posts: