i1Profiler miniature + small RGB printer targets: 10x15 cm, 13x18 cm, A5

[pharmacist]

Thanks to @Ink stained Fingers idea using a very small target (96 patches) on a small 10x15 cm (4x6") paper to obtain a rudimentary but still a satisfactory RGB printer proifle, I took the chance to optimize the amount size (increase to the maximum amount that can be printed on small photo cards ranging van 10x15 cm to A5 (210x148.5 mm) with patch sizes from 192 to 450 patches.

The idea was based on the rather small patch size the Spyderprint uses with its High Quality target containing only 225 patches, that was based on a RGB-grid of 5x5x5 and with additional patches (mostly around the visually most important area around the neutral axis) up to 225 patches) and can still produce very good printer profiles:

By using the same idea: based on the RGB cube of 5x5x5 for 10x15 cm, 6x6x6 for 13x18 cm and 7x7x7 for A5 and enhancing it with extra grey steps (important for good B&W printing with a smooth gradient) and a regularly distributed patch set around the neutral axis to optimize sampling rate (limited by the chosen amount of total patches) of the visually important area of the RGB colour cube, I could generate 3 sets of miniature/small RGB printer targets that, despite the small patch count, can surprisingly create very good printer profles, confirmed by the expertise of @Ink stained Fingers .

The 10x15 cm (4x6 inch) target with 192 patches (can be used for bot manual scanning as well as the i1iO2 robot):

The 13x18 cm (5x7 inch) 285 patch target:

The A5 (210x148,5 mm or half size A4) 450 patch target:

The A4 (210x148,5 mm or half sizeA4) 450 target with i1iO layout (can be used too for manual scanning):

The target sets consists of 3 files of which the .txf and (and eventually the target file in tiff-format .tif, but not necessary because this is the file you want to print) is to be placed into:

\ProgramData\X-Rite\i1Profiler\ColorSpaceRGB\TestCharts

And the workfile file *.pwxf file into:

\ProgramData\X-Rite\i1Profiler\ColorSpaceRGB\PrinterProfileWorkflows

In i1profiler check the "advanced" user mode and on the left of the windows: printer > RGB printer: profiling and go to "saved workflows"and choose the corresponding *.pwxf workflow to start your profiling job after you have printed your target on the corresponding paper size and have it dried properly (at least 1 hour for pigment ink, 24 hours for dye ink).

 

[pharmacist]

Here is another target I have made based on the TC2.83 target that is originally included in the Gretag Macbeth iMatch software alongside with the famous TC9.18 target (and which can still be found in the i1Profiler software package). It is a small target that has initially alot of white patches in the target up to 294 patches. By removing those redundant white patches and adding 2 extra near neutral patches up to 285 patches I could make a target that fits on a small 13x18 cm photo card and also on a half size A4 paper aka A5 (210x148.5mm). Not sure what will give better profiles: this slightly modified TC2.83 target or my 285 patch target with extra grey and near neutral patches. Maybe @Ink stained Fingers can give it a try and report back:

There are 2 sets: one for manual scanning and the other for the i1iO2 robot. The manual scanning set also has a modified target for a smaller 13x18 cm photo card, with extra separation bars between the patches.

 

[Ink stained Fingers]

I checked already the TC2.83 patch sheet

https://www.printerknowledge.com/th...e-needed-for-a-good-profile.16935/post-144751

The Delta to profiles based on higher patch counts is very small and adding a few control points does not make a difference as long as essential control points like the corners of the color space are not removed which is not the case here.

 

[pharmacist]

So actually the TC 2.83 target has no extra benefit compared to the other targets if I have understood it.

 

[Ink stained Fingers]

I'm reaching the limits of data accuracy and validity with the current setup; let me just show you what is happening;
I printed the 0192 patchsheet, created a profile from it , reprinted the 0192 patchsheet with this profile and scanned it twice - directly one after the other and let i1Profiler extract the differences between these 2 scans, the average of all DeltaE's is 0.17 with a max of 0.71 . When you now add just a few more patches a possible delta of improvement will just be covered up by the scan variations. A .pdf file of all the deltas is attached.
Better data accuracy would need a tighter look to the scanning data, are the variations caused by the scanning process, would multiple scans improve the results, are the patch prints uniform enough. The i1io scan table can run in patch by patch mode - very slowly - and would even allow multiple scans per patch, but a scan may not take some minutes but hours.
Or an instrument like the Myiro9 would deliver more accurate results - I don't know.

https://www.myiro.com/myiro-9






There is something else visible from the data of the gray columns on the right side that i1Profiler does not create neutral grays along the gray axis but gives the spots along the gray axis - a=b=0 - a cool touch with some negative b-value. This was already discussed and reported longer time ago - ArgyllCMS values along the gray axis compared with a i1Profiler created profile - here in this table

https://www.printerknowledge.com/th...arison-argyllcms-i1profiler.15717/post-137159

When I browse around for 'Gray axis' and similar I find comments that neutral grays - technically neutral with a=b=0 tend to have a slightly brownish look, and the preference of people would rather be a slightly cooler tone interpreted as neutral which explains the negative b-values and that some profiling software supports that trend.
I cannot really comment on that , I haven't looked in detail into B/W printing and its complications, and this goes beyond the subject of this thread.