Printer Profile Validation: M1 vs. M2 and ΔE2000 Evaluation

[itsab1989]

@pharmacist

It is added to the list, hope it is ready later today

 

[pharmacist]

@itsab1989 : many thanks for the update

Are there any plans to export the ti2 working files towards i1Profiler pwxf working files so one can use the ChromIQ generated targets on both ChromIQ and i1Profiler as well. I believe @jctay is hinting in this directon to have universal target compatible to both ChromIQ (json/ti2-file) and i1Profiler (pwxf-file) ?

In the past before ChromIQ I exported my ti1 into pxf-files and use i1Profiler to generate the corresponding pwxf-file (similar to the ti2-file), but all the patches ar turned around (the row sequence turned into the strip sequence and vice-versa when exported from ti2 into pwxf format). So I think it must be not so difficult to do this export.

 

[itsab1989]

I have no plans in this direction. The AI did a quick analysis of this request when jctay made it but if I remeber correctly that wasn‘t as simple as one would think. Maybe I will look into this again at some point.

 

[know]

It would be nice if the row markers can be rotated for a better visual experience, because scanning on the scanning jig of the target. It is only possible to rotate the strip markers, not the row markers.

Hi @pharmacist ,
For configurability this is logical. The letters are thinner standing upright, so they are taking more space when they are turned, but for multiple pages the two-letter strip labels would benefit from the labels being turned, as the heigth of the letters are smaller than two letters are wide, like "AM". So from a practical view this also is good. However, I think also that the strip letters should be allowed to be placed along side or above the helper markers (only on the top-side of the page) without getting lots of margin-text overlap warnings. Only if the letters them selves overlap with an actual line from the helper markers, then it is good with a warning, and of course if the strip letters are pushed beyond the limits of the T-parameter for the text from the edge distance. This would change the thoroughly tested margin and overlap warnings feature for the top part of the page and would need new rounds of testing before it would work satisfactory, because this was something claude was never able to build properly without my very specific instruction. So if this it done, someone else must take charge in testing that this feature is working to satisfaction. I will unfortunately not be able to do it now.

 

[know]

- Create the Chart: Create a ColorChecker Classic project using the built-in preset.
- Measure the Chart: Print and measure the chart.
- Validate the Results: Compare the measured values against the reference data to assess color accuracy.

Future Consideration: Support for the ColorChecker Digital SG.

Hi @jctay (and @itsab1989),
ChromIQ already supports the "ColorChecker Digital SG" target and many more targets, for scanning them with a flatbed scanner. Your suggested feature is thus something that could be made using many of the already existing features. This would either need to be a separate tool under the tools menu, or a new category of a chart/preset where the user is required to input a reference file (this latter option might be most integrated with the measurement feature). These targets come with a reference file, which must be used to know the correct reference values for the patches on the target. A separate window could be designed where the user selects which target type he has (which is associated with the loaded chart/preset), then selects the reference file for the target, then, same as for normal measurement of a chart, the user can go through measuring the target using the connected instrument (no printing of chart required or needed, and must be registered by the tool, so that it is correctly handled during verification and in the measurement report). When a chart is made to be resemble the layout from a ready made target, then the colors in the target would not originate from a ti1 file, but the target's reference file would define what the ti1 file should contain, thus becomes another source of origin for colors of a chart's ti1 file, in addition to argyllCMS "targen" or the internal ChromIQ Patch Editor.

Also, supported targets must only be those that have large enough patches that an instrument like i1Pro or i1Studion/ColorMunki can be used. Targets with smaller patches (below approx. 8mm wide patches), only for scanners, like LaserSoft Advanced target IT8, or Wolf Faust IT8 etc, should not be included in the selectable list. Targets that are not on paper (passport, and other targets in a plastic frame or box), which are only for photography, should also be excluded.

I would though suggest that this would be a feature relevant for verification, not profiling, since these ready made targets usually do not hold a sufficient amount of patches for good printer profiling. Thus, it is the chart related to run type = verification that is relevant for this.

After measurement, the current features for Check & Refine (Analyse Profile Quality), and the tools "Inspect a measurement", "Measurement Report", "Verify a profile", and "Verify against reference", should be usable as they are (possibly with very small adaptations because the target was not a printed chart).

 

[itsab1989]

@jctay @pharmacist

New update is out. Row labels can rotate now and the instrument buttons can be configured for double and triple presses. Hopefully it works without mayor issues…

https://github.com/itsab1989/ChromIQ/releases/tag/v4.3.14

 

[itsab1989]

@pharmacist

Does this look like a chart from i1Profiler or one from ChromIQs patch generator? Keith emailed me that he was trying ChromIQ now. Just wondering whether this might be a first sign of his testing.

 

[know]

Does this look like a chart from i1Profiler or one from ChromIQs patch generator? Keith emailed me that he was trying ChromIQ now. Just wondering whether this might be a first sign of his testing.

Looks definitively like the Patch Editor we made in ChromIQ
i1Profiler barely creates any greys, and the gamut-corner emphasis or color extremes would look like what he has on that chart...

 

[jctay]

@itsab1989 : many thanks for the update

Are there any plans to export the ti2 working files towards i1Profiler pwxf working files so one can use the ChromIQ generated targets on both ChromIQ and i1Profiler as well. I believe @jctay is hinting in this directon to have universal target compatible to both ChromIQ (json/ti2-file) and i1Profiler (pwxf-file) ?

In the past before ChromIQ I exported my ti1 into pxf-files and use i1Profiler to generate the corresponding pwxf-file (similar to the ti2-file), but all the patches ar turned around (the row sequence turned into the strip sequence and vice-versa when exported from ti2 into pwxf format). So I think it must be not so difficult to do this export.

Blue-Sky Thinking

 

[itsab1989]

I don’t even understand why this would he helpful in the first place. You create a chart in either application and read it with the same one. Why go through the struggle to make one chart compatible with both? The exports of the patchsets is mostly there because ChromIQ does not support some instruments i1Profiler supports and so those can benefit from Knuts patchset generator as well.
But for other instruments you could as well just read the chart in ChromIQ. For i1Profiler compatibility the arrangement is not the only problem. The patches on a chart are not always perfectly square but a little longer in reading direction. If this direction changes this might cause issues with the measurement.