Possibility to convert csv-file with only L*A*B* values into a ArgyllCMS ti3 file

[pharmacist]

Here it is:

 

[pharmacist]

This is how the profile should look like:

 

[pharmacist]

I think I messed up all the files, so this is the correct set of data: ti1, ti2 and the ti3 that is scanned with a known spectro: i1Pro1 (to be used as a reference file to build a new ti3 using the date from the CR30 obtained scandata in csv-format) + the csv-file based on the LAB + spectral data from 400-700 nm obtained by scanning with the Aliexpress CR30 colorimeter.

 

[x64]

Yep that works now, before you rescan, the last remaining step is that the csv needs unique per-patch names that we can match to SAMPLE_LOC of the .ti2 (e.g. A1, B2 etc.). Right now I'm just matching the closest values but that is not accurate or reliable. I noticed from the screenshots that there are quite some measurement parameters available, see if you can include a reproduceable index to the csv, so we can match the measurement to the correct SAMPLE_LOC

 

[pharmacist]

Unfortunately it is not possible but what I can do is make a new target and choose not to randomize the patch layout (standard setting in printtarg command) so the data layout in ti1 and ti2 are in line with another making comparison easier. I keep your informed many thanks for all your work and help already Thanks

 

[x64]

Yes that works too, if you can get all sets across ti1 ti2 and csv uniform (set 1 = 1 = 1, etc) that would make things a lot easier.

 

[pharmacist]

So here you are, it took me more than an hour to produce the set:

the patches in the ti2 are no longer scrambled but are aligned with the ti1 color patch sequence and there is a ti3 file generated from scanning with the i1Pro1 to be used as a dummy ti3 for the data from the csv file to be converted and to be pasted into it.

 

[pharmacist]

Warning for those who already downloaded the previous zip file in previous post: PLEASE download it again, because the ti3 file was corrupted. I had to rescan it again to regenerate a functional ti3 file again.

This is the profile 3D model when using the right ti3 file:

 

[x64]

First results looking good! I'm noticing some difference between the 1iPro1 and CR30 results. Not sure if this is an undesirable offset caused by our process, or just instrument measurement difference.

I'm going to clean up the script and prepare it for upload today. Any simple convenience functionality you would like? I already included direct colorprof execution with config file -> when running the script to convert the CSV you can toggle to directly proceed to run colorprof with the selected arguments.

 

[pharmacist]

Hi @x64 : that is some fantastic job you did. I think the difference is totally explainable. The i1Pro1 ref ti3 is reprinted with a smaller patch size and scanned (but with the same patch layout) and scanned on a yellow table and the CR30 target was printed on a table with a mousepad with greenish pattern underneath. The difference should be acceptable provided the two profiles have more or less the same 3D-body with the some curvature in each XYZ-axis, but a bit smaller or larger is ok.

I know I can cramp extra 19 extra patches on a A4 paper to make up 247 patches, so maybe I will redo the job with 247 instead of 228 patches increasing sampling rate in the RGB color space. The Spyderprint High Quality target has only 225 patches, so 228 should be enough: What I will after you have uploaded your script, is to reprint on a different paper (gloss) and redo the trick and compare it to a profile made with a larger patch set (maybe 648) and using a i1Pro2 and ArgyllCMS and compare those two profiles and see if the magic is still there.

At this moment it looks already very good: that is a great opportunity for a poor man's printer profile. I am pretty sure the device can be used to calibrite a screen as well with the necessary scripts in ArgyllCMS