ArgyllCMS and X-Rite i1Pro2 target sets

[Ink stained Fingers]

I'm looking to the 648 target sheet, and it's probably the same with the other files - they are stored as 48 bit RGB files, have you seen any benefit to start off printing the patch sheets with these *.tiff files , and not just with the common 24 bit (8 bit per color) files. ? I remember already some discussion about the (in)ability of printer drivers to print such 48bit files .

 

[pharmacist]

I have no problem at all printing those files at all. Note: the printer driver should print those files without any color management at all.

 

[thebestcpu]

In the past years I have made several target sets for the X-Rite i1Pro2 device using ArgyllCMS to create printer profiles. I want to share them with all those who have bought an i1Pro2 without an working RGB printer profiling license.

All target sets are in ZIP-format containing *.ti1, *.ti2, target in TIFF-format (360 dpi resolution), AdobeRGB1998.icc and the makeprofile.bat file so it is easier for you to create the profile without typing in the cryptic colprof command with alot of parameters.

The first set uses 875 patches: there are 2 zip-files. one for A4 (210x297 mm) European paper size and the other for Letter (American paper size).

This 875 patch target will be more than enough for most purposes, based on 729 colour patches (RGB: 9x9x9) and 128 (-2 > 126) grey steps + 10 white patches and black patches to make up 875 patches.

Thanks!! You desire an appropriate award
(and yes I know you are on Windows rather than Unix Mac yet still a wizard)

 

[thebestcpu]

I'm looking to the 648 target sheet, and it's probably the same with the other files - they are stored as 48 bit RGB files, have you seen any benefit to start off printing the patch sheets with these *.tiff files , and not just with the common 24 bit (8 bit per color) files. ? I remember already some discussion about the (in)ability of printer drivers to print such 48bit files .

That's a good question that might not yield to a good answer without an experiment with target sheets done both ways.
I did find out that I could see a visual difference when printing an image in 16-bit presented to the printer vs the same image in 8-bit presented to the printer:
https://www.printerknowledge.com/th...s-and-8-vs-16-bit-workflows.16794/post-143833

A bit different with color targets as most of the color chips are spaced from each other. In the @pharmacist's profile of 9x9x9 color cube, the color numbers could be ~28 separated from each other. With 8 bits, a separation might be off by 1 when rounded.

However, with the greyscale of 128 shades, the steps between them would be ~2, so an occasional 8-bit rounding error could separate one step by 3 instead of 2.

I'm unsure what difference that would make once the RBG is translated into printer CMYK and other measurement inaccuracies and finally turned into a profile.

I was planning on doing some grayscale target experiments on my Espon P900 to tease out what is going on with Black Enhanced Overcoat, and Carbon Black so I just may add targeted experiments doing a target in 8 bit vs 16 bit to see if it makes any difference.

Even if you could measure a difference in the experiments, I bet you would never see the difference in a printed image.

For me, since the 16-bit print option is there on my Mac, it did show a difference, as shown in my linked post, I am, by default, going to print in 16-bit for profiling targets.

 

[Ink stained Fingers]

You would have to measure color differences of prints printed either with a 24 bit color file or a 48 bit color file, and your measurements would need to be more accurate than reading variations of the instrument and color variations of the prints. You could take such measurements and calculate a DeltaE difference to get a first glance how small these variances are .

Delta E =

http://zschuessler.github.io/DeltaE/learn/

These readings just would be input data into the profile calculation engine of which we don't know if they process 8 or 16 bit per color from input to the end, the profile as the intended result. That's all beyond my capabilities and the typical process steps to create an icc-profile.

 

[thebestcpu]

You would have to measure color differences of prints printed either with a 24 bit color file or a 48 bit color file, and your measurements would need to be more accurate than reading variations of the instrument and color variations of the prints. You could take such measurements and calculate a DeltaE difference to get a first glance how small these variances are .

Delta E =

http://zschuessler.github.io/DeltaE/learn/

These readings just would be input data into the profile calculation engine of which we don't know if they process 8 or 16 bit per color from input to the end, the profile as the intended result. That's all beyond my capabilities and the typical process steps to create an icc-profile.

Great inputs thanks.
John Wheeler

 

[skyer]

In the past years I have made several target sets for the X-Rite i1Pro2 device using ArgyllCMS to create printer profiles...

Exuse me, do you have similar targets for Colormunki Photo?

 

[pharmacist]

Exuse me, do you have similar targets for Colormunki Photo?

Here you are. Please note: a custom scanning ruler is necessary to read this patches.

 

[skyer]

Here you are. Please note: a custom scanning ruler is necessary to read this patches.

Thank yo so much!
As for the scanning ruler, I saw your DIY device. I noticed that in order not to damage a paper surface there is a layer between it and a spectrophotometer. Thus the spectrophotometer is a few mm above the paper. Doesn't it influence the chart readings?

 

[pharmacist]

Not really, but you can use a paper version as well: My previous ruler is paper based and it works perfectly. Have a look at my other test with both i1Pro2 and Colormunki and the resulting profile is the same.