[argyllcms] Re: Absolute colorimetric - dark saturated colours excessively light

  • From: Gerhard Fuernkranz <nospam456@xxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Fri, 03 Feb 2006 22:16:40 +0100

Alastair M. Robinson wrote:

Hi,

I've been experimenting with various paper types on my R300 and hit a
snag with the Absolute Colorimetric intent and my profiles.

The problem concerns very dark, saturated colours becoming excessively
light when transformed to a profile with poor black density.

Here are some images to illustrate the problem.  Note particularly
columns 2, 3 and 4.

Firstly, the IT8 target, as scanned on my Perfection 640 and transformed
to sRGB with Argyll.

http://i14.photobucket.com/albums/a346/robinsonb5/IT8_sRGB.jpg

For the following files, I did an absolute colorimetric transform from
the scanner's profile to the printer profile, saved as a temp file, then
transformed that from the printer file to sRGB, again as absolute
colorimetric. Thus a side-by-side comparison of the final file on-screen and a print should match pretty well, which it does.


Firstly, 7DayShop glossy, which has excellent dense black:
http://i14.photobucket.com/albums/a346/robinsonb5/IT8_Glossy_Proof.jpg

Then Tesco matte inkjet paper, not far behind the glossy.  Note columns
2, 3 and 4 becoming lighter:
http://i14.photobucket.com/albums/a346/robinsonb5/IT8_Tesco_Proof.jpg

Finally, bog-standard plain paper:
http://i14.photobucket.com/albums/a346/robinsonb5/IT8_Plain_Proof.jpg

Any ideas what's going on here?  Should I be able to solve this by
fiddling with the profile program's parameters?
Alastair,

I do not understand what you want to "solve"? What you explain, is the expected behaviour. You simply encountered the gamut limitations of your printer on different paper types, and a colorimetric transformation to the printer color space clips colors which are out of gamut on the printer to the gamut boundary. And with the 2nd transformation from printer color space to sRGB, you get a soft proof, i.e. if you would actually print the image after conversion to the printer color space, then the print should look similar to the soft proofs you had posted above.

Regards,
Gerhard
(I can let you have TI3s for all four profiles if you want to see them...)

All the best,
--
Alastair M. Robinson


Other related posts: