[argyllcms] Re: Visualizing calibration charts

  • From: Mikhail Matrosov <mikhail.matrosov@xxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Wed, 3 Oct 2012 15:07:15 +0400

>
> Not at present. It's an interesting idea though. For an x,y reader
> (like the spectroscan) it would be useful to use .cht like information
> to locate the patches, but the print and scan sides of Argyll aren't that
> well integrated at the moment.


What about printtarg <http://www.argyllcms.com/doc/printtarg.html>?
Description says it will produce a TIFF image given a .ti1 file. Is it a
good idea to manually convert layout and reference data from .cht and .cie
to a .ti1?

-----
Best regards, Mikhail Matrosov


2012/10/3 Graeme Gill <graeme@xxxxxxxxxxxxx>

> Mikhail Matrosov wrote:
>
> Hi,
>
> > I wonder, whether it is a tool in argyll, which allows to generate a
> > synthetic image of a chart, given its description in .cht, .cie, or other
> > formats? Images like these:
> >
> http://www.babelcolor.com/main_level/ColorChecker.htm#ColorChecker_images
>
> Not at present. It's an interesting idea though. For an x,y reader
> (like the spectroscan) it would be useful to use .cht like information
> to locate the patches, but the print and scan sides of Argyll aren't that
> well integrated at the moment.
>
> > For now I manually parse .cht file for layout of color patches and
> > reference XYZ colors, and drow it myself in Matlab. Is there a simpler
> way?
>
> A lot of the nessesary logic is in scanin. I guess it could be
> modified to do this.
>
> > And, by the way, why colors in .cht file are just approximates and one
> need
> > .cie file to run colprof? Wouldn't it be simpler just to include
> reference
> > colors in the same .cht file?
>
> Because a .cht defines chart layout for recognition purposes, while the
> reference values depend on the chart instance. So the same .cht
> can be used with any number of different manufacturers or instances
> of (say) an IT8/7 chart, while (ideally) each one needs a different
> set of reference values.
>
> Graeme Gill.
>
>

Other related posts: