Gamut issues with i1Profiler & i1Pro2

[MilleMiglia]

Hi everyone,

I have been quietly reading this forum for a bit, and I already learned a lot here, so thank you all for this ! Today I have an issue with creating profiles with i1Profiler.

I am using an Epson SC-P600, mostly with Luster paper. I was using Epson's Metallic Luster Profile v2, but I wanted a profile that would be more accurate ( I always thought that Epson's profile was too magenta in the skin tones and a little crazy in the Reds). I recently bought an i1Photo Pro 2 kit and started playing with it.

After creating my first profile in i1Profiler with a 3000 patches target, it was clear very early on that there was a serious issue with the colors of the generated profile, both while soft proofing and also on the final print. My first observations were a lack of contrast, lack of saturation and a reduced gamut, especially in the green tones. This has been confirmed upon comparing i1Profiler's profile with the Epson one in ColorSync (I attached the screenshots of these, along with a print comparison).

While scanning the targets, I noticed that all green patches that were more saturated seemed to already have an important visual delta in i1Profiler.

Here is what I tried so far to resolve the problem :

- I generated profiles with i1 Profiler both in single and dual scans (for OBC paper and without);
- I tried with i1Profiler v3.4.2 and v3.5;
- I ran i1Diagnostics on the i1Pro2 (everything looked fine);
- I noticed that the profile optimization feature produced a larger gamut in the blue tones, but reduced contrast and lifted the black point even more (this was with a 3000 patches optimization target);
- I tried using the profile both in Photoshop 2022 and Capture One 21, with the same results;
- I tried using the profile to produce prints both on a M1 MacBook Pro and a Intel MacBook Pro;
- Soft proofing was made on a calibrated monitor.

I am far from being an expert on this, so there is probably something I missed - However I am a bit stuck at the moment, hence this (lengthy) post. Any help would be greatly appreciated !

Cheers

Link to pictures : https://www.dropbox.com/sh/98eeh3e07a5wcil/AAAPvBo1SkMb3z2NlN6yEZICa?dl=0

 

[Doug Gray]

Took a look at your profile. It's really distorted and min L* is around 16. That's awful for a non-matte paper. It is far different from the KL_LUSTER profile. Not even in the same ballpark.

I notice the profile is labeled: SC-P600 Series Metallic Photo Paper Luster PK v2.icc

Metallic papers have small particles in them that reflect a portion of the light. This may account for the really odd gamut shape and high DMin. Spectrophotometers measure the light by shining an annular beam at 45 degrees onto the paper then reading the spectrum directly facing the paper.

You should try using a non-metalic paper to verify your workflow. An M3 (polarizing) spectro might or might not work better but that has it's own problems in addition to costing a lot more.

Added:
I made a profile from your 3k patch set (it's embedded in the profile) with i1Profiler. Same awful results and with weird indentations on the gamut surface. Oddest is that near L* min, the profile inverts inwards in a sort of cup. Very strange. But could well be cause by specular reflection from the metallic particles.  M3 spectros use polarized light to remove these but they also have their own distortions. Mostly with Matte papers where they produce false DMins. BTW, M3 profiles are not ICC compliant though some paper suppliers offer them.

Until you verify workflow, it's best, and a lot faster, to use smaller patch sets. 600-1200 or so. There is rarely a visual difference above that.

[smilem]

In the old days I bypassed the problem of metalics sort of, by using good quality scanner profile (that I made) and scanning the printed target, then extracting the color data.