[Tool] ChromIQ – a macOS and Windows GUI for ArgyllCMS printer profiling (v3.0.0 beta 10)

[itsab1989]

I just wanted to post you the questions that AI gave me
So is it working as intended now?

Edit: I just released 3.5.10 with improvements in the measure tab. I have a feeling though that the patch consistency settings default is too strict. I mean the user can change it but it might cause confusion.


I'll post the AI text anyway in case it might be useful:

Thanks for the detailed report — I tried to reproduce on my end (Windows, 240-patch RGB chart with a preconditioning
profile, sRGB.icm as -S source) and the build completed cleanly, so I need a bit more info from you to pin this down.

A few things I noticed:

1. The colprof warnings on their own are harmless. Warning - gammap internal: input bounds has changed! is ArgyllCMS's
gamut-mapping optimizer adjusting its working bounds mid-iteration. It's noisy but doesn't normally break the
profile. However, many repetitions of it combined with a profile that iccview.de rejects suggests the underlying .ti3
measurements are off — gammap is reacting to outlier data, not causing it.

2. Your ArgyllCMS lives at CArgyllCMS\bin — that's a manual install location. ChromIQ auto-detects CProgram
Files\ArgyllCMS\bin and %LOCALAPPDATA%\ArgyllCMS\bin, so you must have pointed it there in Preferences. Which version
is it? Run CArgyllCMS\bin\colprof.exe -? and the first line will show the version — I'd like to make sure it matches
ChromIQ's tested 3.5.0.

3. The hang during chart generation with a preconditioning profile is the part I most want to understand. targen -c
<profile> can take a while, but it shouldn't lock up. If the preconditioning .icc is malformed, targen can spin for a
long time trying to invert the CLUT.

To dig in, could you share:

1. The output of CArgyllCMS\bin\colprof.exe -? (just the version line)
2. Where the preconditioning profile came from — a previous ChromIQ run, or imported from somewhere else?
3. The .ti3 file and the preconditioning .icc, attached or zipped
4. The ChromIQ log file (%LOCALAPPDATA%\ChromIQ\logs\chromiq.log or wherever it lands on your system)
5. Your targen settings for that run — patch count, grey steps, good-mode on/off

Once I can see the .ti3 I should be able to tell whether this is a measurement-data issue, a preconditioning-profile
issue, or something in ChromIQ's pipeline staging the file wrong.


Edit 2: Light mode added in 3.6.0

 

[pharmacist]

Thanks for the update. I find the light version almost unreadable, because the letters are so light in my case.

But I still cannot get a good profile, despite 600 patches:

I get a very strange flatten 3D profile:

The profile is unusable when I try soft proofing checked:

 

[itsab1989]

Wow, this looks horrible...
1. I added a light mode because somewhere else someone was complaining that the dark theme would hurt his eyes. Might still need some work probably.

2. I would need the exact steps you took to try and reproduce this to figure out what went wrong here. Best would be to have the data from the log file and maybe even the files that were created.

 

[pharmacist]

The latest version also removes the scanning arrow (showing which row is being read by the spectro), that was present in the previous version: it was very convenient to show which row is being read, especially in the new version the contrast is so low I can hardly read the letters in the command window in the lower left side.

 

[itsab1989]

The latest version also removes the scanning arrow (showing which row is being read by the spectro), that was present in the previous version: it was very convenient to show which row is being read, especially in the new version the contrast is so low I can hardly read the letters in the command window in the lower left side.

It should not be removed but apparently it had issues to get the right position. Working on a fix already

 

[pharmacist]

OK, I just keep on posting feedback. For comparison here is my profile generated with old school commands in dos box for a 648-patch target:

Above profile seems to be legit and here is the profile in action in softproofing in Qimage:

The profile is ok with the old fashioned command typing method in ArgyllCMS.......

 

[itsab1989]

I released an update that invludes some missing option for targen that was requested and tried to fix the highlighter in measure tab. It is working better on my machine but I‘d be interested in your feedback.

Would you mind telling me the commands you are using for each tool so I can try to figure out what is breaking?

And if the light mode is not what you like you can click the little settings button on the right side of the header. In the menu you can switch back to dark mode.

 

[pharmacist]

My commands for this particular profile:

targen -v -d2 -G -f648 -g128 -e4 -B4 -N0.75 -648-target

For scanning with my i1Pro2 I use this:

chartread -v -H -T0.4 648-target

For generating the profile this is my command (batch file called: makeprofile.bat):

----------------------

@echo off

set /p par1=Please enter the internal ICC file name (use brackets like "x x x" for names with spaces):
echo.
set /p par2=Please enter the (optional) copyright description (use brackets like "x x x" for names with spaces):
echo.
set /p par3=Enter the name of the .ti3 file created by the chartread command:
echo.
set /p par4=Please enter the desired external name of your printer profile:
echo.

colprof -v -qh -i D50 -o 1931_2 -S AdobeRGB1998.icc -cmt -dpp -D%par1% -C%par2% %par3%

rename %par3%.icm %par4%.icm

-----------------------------

Als you can see: I use AdobeRGB1988.icc as the mapping target.

 

[itsab1989]

Thanks a lot for this information. I will have a look at it later but it seems you are using some options that I have never touched myself so far because some of them won’t work with my i1Studio. I have a feeling that the colprof settings are the culprit and something is wrong with how my app hands them over to Argyll. For another tab there was a windows specific issue where the OS only accepts lowercase letters in the code or something. Very strange behaviour.

Sorry for all of these invonveniences and thanks for your feedback. I really hope to get this right.