[argyllcms] Re: Printed images too cool

  • From: Jos van Riswick <josvanr@xxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Thu, 14 Feb 2013 11:51:58 +0100

thnx for your advice!

I was printing a photo of someone, that's what I was referring to. In the
meantime Alan Goldhammer pointed me to a test image


Printing this one in the same way, with my own profiles, I now notice some
oddities in the saturated greens (the colored patches in the image seem to
merge together there, as opposed to the print from photoshop/windows). The
greys in the photoshop print are even slightly cooler than gimp/my own
profile!..  Having spent several days on this, I think I'll settle for
photoshop running in a virtual machine for now.

Maybe I will try cctiff at a later time, but how should I print the image
without Gimp? Or do you mean I should do the color conversion with cctiff
and the printing still with Gimp/gutenprint? When I do get around to doing
this, I will look into fwa compensation too.

One problem with using a command line utility would be that I do need to
edit in proof mode...


On Wed, Feb 13, 2013 at 1:00 AM, Graeme Gill <graeme@xxxxxxxxxxxxx> wrote:

> Jos van Riswick wrote:
> Hi,
> > The printed result now looks much much too blue. This holds for both
> normal
> > and softproof viewing.
> What part of the image ? ie, are neutrals too blue ?
> What intent were you using ?
> How fluorescent is your paper ? (CM is effectively UV filtered).
> > Am I missing some temperature option or something? My monitor is
> > calibrated/profiled to match my studio lamps (solux, 4700K). White on the
> > monitor matches a white test chart pretty well....
> To figure out what's going on, you need to remove extraneous elements
> such as Gimp. What happens if you us ArgyllCMS to manage the color ?
> ie. standard test image -> sRGB -> cctiff -> Epson profile -> printed on
> printer using
> exactly the same settings as your test chart.
> If that still shows a problem, you can then pin down a particular example
> (such as RGB .5 .5 .5) and run it though the color pipeline to examine and
> test what's happening at each step.
> If that's not the problem, then you need to investigate what Gimp is doing,
> or how your test chart printing differs from printing from other programs.
> Graeme Gill.

Other related posts: