Pages: [1]   Go Down

Author Topic: Argyll workflow (specifically profiling printers)  (Read 7341 times)

shewhorn

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 537
    • http://
Argyll workflow (specifically profiling printers)
« on: March 17, 2010, 12:10:47 pm »

Didn't find anything specifically on a workflow with Argyll so I was hoping a few people out there might be able to offer some general advice on getting started. I have an i1XTreme but I find that the profiles the Eye One Match software generates to be a bit disappointing when it comes to highlights on certain papers. It seems to do an okay job with mat based papers but when you throw something like Epson Exhibition Fiber or Crane Museo Silver Rag at it I find that it loses some detail in the highlights. One example... if you download the test images from Bill Atkinson:

http://homepage.mac.com/billatkinson/FileSharing2.html

... on the test image there is a yellow flower. Profiles generated with Eye One Match (regardless of how many patches there are... I've used the highest sample target from the i1Xtreme package as well as the 5202 patch target from Bill) lose a lot of detail in those yellow highlights compared to a profile generated with Monaco Profiler (this BTW is on a Canon IPF6100 (and in a few weeks an IPF8300). After seeing that I was prepared to pickup a copy of Monaco Profiler or X-Rite's Profile Maker however a little birdy has hinted that there will be a new profiling product available soon which could potentially be much better. In the mean time however I figured I'd try Argyll and I wanted to run the following by a few folks who might have some experience with the tool to see if I'm on the right track. I already have the 5202 patch target from Bill Atkinson printed out and it's had plenty of time to dry so since I already have it... I figured I'd use that. I need to generate a ti3 file so first...

txt2ti3 Convert Gretag/Logo/X-Rite or other format RGB or CMYK test chart results into Argyll .ti3 CGATS format.
If I wanted to generate my own target directly from Argyll I could use the printtarg command to generate a TIFF target chart.

chartread - This is the business section... use the Eye One Pro to scan in the cart and create a ti3 file

colprof - Create an ICC profile from the .ti3 test data.

I think that's it aside from the specific arguments I have to pass to each command. Does anyone have any hints or tips? I'd also be interested in any feedback anyone has if they're done any comparisons between Eye One Match and Argyll generated profiles.

Cheers, Joe
Logged

shewhorn

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 537
    • http://
Argyll workflow (specifically profiling printers)
« Reply #1 on: March 17, 2010, 12:43:47 pm »

Oh... and if anyone could shed any light on this error from txt2ti3 it would be greatly appreciated as well:

logo2cgats: Error - Input file 'RGB 5202 Eye-One.txt' doesn't contain field XYZ_X or spectral.

I have to try a file directly from X-Rite. This particular file is from Bill Atkinson. I compared it in a text editor to a file direct from X-Rite and the only obvious difference were some slightly different spellings of the variable names (I got some initial errors from that which was easily fixed to match what Argyll watned to see) but other than that the only major difference is that the X-Rite file lists the X and Y axis names where as Bill's file individually enumerates every patch. I suspect Argyll may not like this method but if anyone has done this... I'd be interested in your input. If this doesn't work I think the "scanin" tool might be able to read the actual TIFF chart.

Cheers, Joe
Logged

Czornyj

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1948
    • zarzadzaniebarwa.pl
Argyll workflow (specifically profiling printers)
« Reply #2 on: March 17, 2010, 12:54:59 pm »

First of all - did you try to change gamut mapping options in i1match?
http://people.csail.mit.edu/ericchan/dp/i1...ping/index.html

ArgyllCMS is also not a bad idea - you can get a GUI there:
http://colorhacks.blogspot.com/2008/08/gui...gyllcmsgui.html

http://colorhacks.blogspot.com/2008/08/gui...javascript.html
« Last Edit: March 17, 2010, 05:14:01 pm by Czornyj »
Logged
Marcin Kałuża | [URL=http://zarzadzaniebarwa

Czornyj

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1948
    • zarzadzaniebarwa.pl
Argyll workflow (specifically profiling printers)
« Reply #3 on: March 17, 2010, 01:15:46 pm »

Quote from: shewhorn
logo2cgats: Error - Input file 'RGB 5202 Eye-One.txt' doesn't contain field XYZ_X or spectral.

txt2ti3 is not for target conversion, but for measurement file conversion.
Logged
Marcin Kałuża | [URL=http://zarzadzaniebarwa

probep

  • Full Member
  • ***
  • Offline Offline
  • Posts: 149
Argyll workflow (specifically profiling printers)
« Reply #4 on: March 17, 2010, 03:08:07 pm »

shewhorn
Did you try free ColorXact profiler? It's good enough.
« Last Edit: March 17, 2010, 03:08:41 pm by probep »
Logged

shewhorn

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 537
    • http://
Argyll workflow (specifically profiling printers)
« Reply #5 on: March 17, 2010, 04:00:28 pm »

Quote from: Czornyj
txt2ti3 is not for target conversion, but for measurement file conversion.

Gotcha... I guess I would need to be able to convert a reference file to a ti2 file then?

Cheers, Joe
Logged

Czornyj

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1948
    • zarzadzaniebarwa.pl
Argyll workflow (specifically profiling printers)
« Reply #6 on: March 17, 2010, 05:10:24 pm »

Quote from: shewhorn
Gotcha... I guess I would need to be able to convert a reference file to a ti2 file then?

Cheers, Joe

No - just convert the .txt measurement file to .ti3 and use colprof to create .ICC profile

Use that GUI to make things simplier:
http://colorhacks.blogspot.com/2008/08/gui...gyllcmsgui.html
« Last Edit: March 17, 2010, 05:21:31 pm by Czornyj »
Logged
Marcin Kałuża | [URL=http://zarzadzaniebarwa

shewhorn

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 537
    • http://
Argyll workflow (specifically profiling printers)
« Reply #7 on: March 17, 2010, 08:43:19 pm »

Quote from: Czornyj
No - just convert the .txt measurement file to .ti3 and use colprof to create .ICC profile

Duh... that makes sense. I'm going to blame it on all this water (I'm in Massachusetts... we had 10 inches of rain in a few days, I figured my brain is water logged (it was 71 out today though.... woohoo)).

Alright, I managed to generate an ICC profile. Argyll does a better job with holding that yellow detail but totally fails with reds and oranges (the fall trees lose all the highlights).

Quote
First of all - did you try to change gamut mapping options in i1match?
http://people.csail.mit.edu/ericchan/dp/i1...ping/index.html

Ahhh no, hadn't seen that. LOGO Classic did a good job of retaining that detail but the saturation (as one might expect) was noticeably reduced. The Chroma Plus seemed to be better all around but "better" is still subjective at this point. What I found interesting was that if you look at the out of gamut warnings the Monaco Profiler generated profile has the narrowest gamut of all the profiles yet it still appears to have a decent overall saturation. I haven't played with any of Argyll's options yet so I'll have to see what's available to me there as well. I'll try that Java based GUI you posted (thanks).

Cheers, Joe
Logged
Pages: [1]   Go Up