Printer Profile Validation: M1 vs. M2 and ΔE2000 Evaluation

[itsab1989]

Issue1: The command boxes begin to appear before the next working screen is displayed

After the ICC profile is built:
Click Create Chart.
Change the Run Type to Verification.
Observe the attached video.


Issue 2: Fixed Seed Is Not Saved in Preset

Preset files are stored in the following directory:
\Users\UserName\AppData\Roaming\ChromIQ\presets

Example file: 1Pro-4x6in-192p-2pages.json

Currently, the preset file contains the following values:

"randomize": true,
"seed": null,
"seed_fixed": null

Expected behavior:

If the user defines a seed value (e.g., 1234), the preset file should save that value in the seed field:

"randomize": true,
"seed": 1234,
"seed_fixed": null

The user-defined seed value should be saved correctly in the preset file and retained when the preset is loaded again.



.../2 zips are attached

That's really strange. I will see what I can do about it

Edit:

Edit 2:

Fixed:

https://github.com/itsab1989/ChromIQ/releases/tag/v4.3.7

 

[know]

After the ICC profile is built:
Click Create Chart.
Change the Run Type to Verification.
Observe the attached video.

Nice catch We have not seen anything like that before, so it is good that some more people, like you, help out with reporting bugs

Thanks.

 

[jctay]

ChromIQ Profile Verification – Part 1: General Checks

Sharing a few screenshots showing the initial measurements and builds, without any profile refinement.

Profile Quality Check

Patch Distribution - Verify Target

Patch Distribution (RGB) - Verify target vs Profile under test

Soft Proof Using the Built-in ChromIQ Test Image
For illustrative purposes only. Actual colors may vary from those shown in the screen capture.

.../

 

[jctay]

ChromIQ Profile Verification – Part 2: Measurement Validation

Full colour check Judged against Quick check

Grey and tone check Judged against Quick check

Colour summary (one page) Judged against ISO 12647-8 2021 values

Validation print check (ISO 12647-8) Judged against Custom ISO 12647-7

Others: Please refer to the attached document.

 

[know]

Hi @jctay ,
Thanks for your feedback and showing us your results....

I have a few comments from what I can see...

The "Contract proof check (ISO 12647-7)" and "Validation print check (ISO 12647-8)" report you posted have a red text warning:
the profile has been rebuilt since this sheet was printed, so these figures describean older profile than the one now in the run

This means you have created a new profile at some point, but did not re-create the colors of the verification chart, and then print that and then re-measure. Thus, it seems (if I interpret correctly), that the measurements you did measure are describing the old profile, not the new one. You should check what you did exactly, and preferibly re-create the verification chart in "From Profile's Gamut" mode, then print and measure again.

Then, the reports contain both "Within gamut" and "Beyond it" columns, which means that you did not use colors on your verification chart that were only within gamut, using "From Profile's Gamut" mode. (you have to remember to click on Generate Chart after clicking on "From Profile's Gamut" mode in Create Chart). Thus, you printed and measured a chart that was the original patch set you had for your 192 patch chart. The

This image you posted with the distribution of the colors for the 192 patch verification chart proves also that the patch distribution is not evenly distributed, which is how it looks like when the normal patch set is viewed through the profile. If you had used "From Profile's Gamut" mode it would be much more even. Though, for some reason, this view does not show the out-of-gamut patches.

Also, the distribution comparison image you shared, shown again below, shows a patch set of the 572 patch profiling chart that is not the best. There are no near-neutral gray rings for better neutral. I suggest you use one of the 600 patch presets supplied in ChromIQ instead, shown below:

Distribution of 600 patch preset for 10x15cm / 4x6" photo card. This is a better chart for profiling than the 572 patch chart you used.

Note also that the Evenness check matrices
- Maximum ΔE00, between two of the nine sheet areas (within gamut)
- Maximum ΔE00, one sheet area against the whole sheet
They require a minimum of 9 x 9 grid of patches on the verification chart, thus it could not be judged for the 192 patch chart you used.
If you wonder which chart can be used, among all saved presets (custom made or built-in) you can click on button "Which presets can be used for verification" under the Create Chart tab, while Run type = Verification.

Questions about the reporting feature:

What is your impression about the reports? Does it supply you with the information you need for quality assurance?

Comments? Impressions?


Cheers,
Knut

 

[jctay]

Hi @jctay ,
Thanks for your feedback and showing us your results....

I have a few comments from what I can see...

The "Contract proof check (ISO 12647-7)" and "Validation print check (ISO 12647-8)" report you posted have a red text warning:
the profile has been rebuilt since this sheet was printed, so these figures describean older profile than the one now in the run

This means you have created a new profile at some point, but did not re-create the colors of the verification chart, and then print that and then re-measure. Thus, it seems (if I interpret correctly), that the measurements you did measure are describing the old profile, not the new one. You should check what you did exactly, and preferibly re-create the verification chart in "From Profile's Gamut" mode, then print and measure again.

Then, the reports contain both "Within gamut" and "Beyond it" columns, which means that you did not use colors on your verification chart that were only within gamut, using "From Profile's Gamut" mode. (you have to remember to click on Generate Chart after clicking on "From Profile's Gamut" mode in Create Chart). Thus, you printed and measured a chart that was the original patch set you had for your 192 patch chart. The
View attachment 19456
This image you posted with the distribution of the colors for the 192 patch verification chart proves also that the patch distribution is not evenly distributed, which is how it looks like when the normal patch set is viewed through the profile. If you had used "From Profile's Gamut" mode it would be much more even. Though, for some reason, this view does not show the out-of-gamut patches.
View attachment 19457

Also, the distribution comparison image you shared, shown again below, shows a patch set of the 572 patch profiling chart that is not the best. There are no near-neutral gray rings for better neutral. I suggest you use one of the 600 patch presets supplied in ChromIQ instead, shown below:
View attachment 19458

Distribution of 600 patch preset for 10x15cm / 4x6" photo card. This is a better chart for profiling than the 572 patch chart you used.
View attachment 19459

Note also that the Evenness check matrices
- Maximum ΔE00, between two of the nine sheet areas (within gamut)
- Maximum ΔE00, one sheet area against the whole sheet
They require a minimum of 9 x 9 grid of patches on the verification chart, thus it could not be judged for the 192 patch chart you used.
If you wonder which chart can be used, among all saved presets (custom made or built-in) you can click on button "Which presets can be used for verification" under the Create Chart tab, while Run type = Verification.
View attachment 19460


Questions about the reporting feature:

What is your impression about the reports? Does it supply you with the information you need for quality assurance?

Comments? Impressions?


Cheers,
Knut

Thanks for your comments.

I recreated a profile using ChromIQ’s built-in 600-patch set, and the profile creation process went smoothly.

However, during verification, I spent considerably more time than expected taking measurements because the i1Pro repeatedly failed to read certain strips.


**My personal perspective**

Users who need verification features to meet certification requirements as part of their organizational processes may find these functions valuable. While I find them impressive, I believe new users will need some time to become familiar with the workflow.

For my part, I’m simply looking to improve color reproduction on an affordable photo printer for home use while enjoying the process of experimenting with printer profile creation. Certification requirements, on the other hand, are entirely new territory for me.

The most discouraging aspect has been the measurement process, which has taken up far more time than I anticipated. I’ve followed most of the precautions discussed in this forum to minimize measurement-related issues. In fact, I’m quite comfortable with the measurement workflow in i1Profiler.

I’m not sure whether other users have encountered similar difficulties when measuring with ChromIQ. I’m running it on a PC that’s more than 10 years old, so I wonder whether the CPU’s processing speed could be a contributing factor. Alternatively, could the timing or responsiveness of certain program scripts affect data transfer from the instrument and cause some strips to fail to read?

I’d be interested to hear about other users’ experiences, particularly those using an i1Pro on older hardware.

 

[itsab1989]

Thanks for your comments.

I recreated a profile using ChromIQ’s built-in 600-patch set, and the profile creation process went smoothly.

However, during verification, I spent considerably more time than expected taking measurements because the i1Pro repeatedly failed to read certain strips.


**My personal perspective**

Users who need verification features to meet certification requirements as part of their organizational processes may find these functions valuable. While I find them impressive, I believe new users will need some time to become familiar with the workflow.

For my part, I’m simply looking to improve color reproduction on an affordable photo printer for home use while enjoying the process of experimenting with printer profile creation. Certification requirements, on the other hand, are entirely new territory for me.

The most discouraging aspect has been the measurement process, which has taken up far more time than I anticipated. I’ve followed most of the precautions discussed in this forum to minimize measurement-related issues. In fact, I’m quite comfortable with the measurement workflow in i1Profiler.

I’m not sure whether other users have encountered similar difficulties when measuring with ChromIQ. I’m running it on a PC that’s more than 10 years old, so I wonder whether the CPU’s processing speed could be a contributing factor. Alternatively, could the timing or responsiveness of certain program scripts affect data transfer from the instrument and cause some strips to fail to read?

I’d be interested to hear about other users’ experiences, particularly those using an i1Pro on older hardware.

Thanks for doing all this testing!
Although I doubt your assumption about this being related to the measurement part of ChromIQ I will let AI check it once it is idle again.
The measurement engine is based on the code from Argylls chartread. We copied („forked“ it as it is called) to build some features into it that the regular version does not support. For example the ability to save the measurements after every strip instead of only after you have read the whole chart. This then enables the ability to show the measured vs expected overlay on top of the patches you have measured.
So in theory this should not affect the transfer from the device to the computer. But since I have noticed similar issues myself on some occasions and I have learned to never trust your instincts in this coding process I‘d rather let this get checked. Might take a while though as I am not that available this week.