Question about my custom profiles done with Argyll & Colormunki

[mristuccia]

Hello Everyone,

I'm creating some custom profiles with Argyll and my old Colormunki Photo spectrophotometer.
Using around 700 patches right now, and the process finished with encouraging statistics which say:

Profile check complete, peak err = 1.622554, avg err = 0.272795, RMS = 0.340088

Specifically, I did profile the Canson Infinity Baryta II for my Epson Stylus Photo R2880. Working on macOS Monterey 12.6.

Now, this is more of an exercise, as Canson already provides a profile for that paper on that printer.
My question arises when I graphically compare my custom profile against the official Canson one.

Here are the two:

Canson official profile


My custom profile


Both profiles print perfectly well, and when I compare normal prints I see no difference.

Now, I'm not an expert at all in such color profiling science. My question is, can anyone please explain to me the reason why my custom profile has this gap at the bottom end towards the black point while the Canson one does not have it?

Thank you in advance for any light you can shade on this.

 

[anwarp]

One possible reason is that the color munki is not sensitive enough with the darker colours.
To add to this, I think the second image looks rolled upwards a little more, if I count the squares.

Anwar

 

[mristuccia]

Hi Anwar,

thank you for your feedback.
Here below an image where I overlapped the two profiles them (Apple ColorSync "compare"). This shows better the differences in the bottom part.



The coloured volume represents my custom profile, while the greyish one is the official Canson profile.
It seems indeed that either Colormunki in not sensitive enough on the dark areas, or that Argyll 3.0.0 has some problem in interfacing the Colormunki.
On the other side, it also seems that my custom profile is very slightly better on the upper part of the gamut.

I'm going to try the official X-Rite/Calibrite software (ccStudio) to see if the generated profile sill shows this lack in the darker areas.

I did not want to use the official software as it only has 50+50 color patches and I wanted to try a way to generate a more precise profile. With Argyll I can use as many color patches as I want.

Will report my findings here, for you and for anyone else who in the future can find himself/herself in the same situation.

 

[mristuccia]

Hello again,

I've generated a custom profile by using the official X-Rite/Calibrite ccStudio software.
This process implies the generation of a first set of 50 color patches which is always the same, and then the generation of a second set of 50 color patches generated as a refinement after reading the first one.

Here is the result (custom profile coloured, official custom profile greyish):



So, here the bottom black point is reached, but the overall volume is way worse than the one I've generated with Argyll and 700 color patches.

So, I'm even more perplexed

 

[anwarp]

I use Argyll for my profiling, but I have an old rebadged i1pro. It does not have the license for xrite software.

Did you use the canson profile as the starting point for your custom profile?

And of course, how did you print the targets?

I follow these steps that torger published:

Printer profiling with Argyll and Colormunki

 

[anwarp]

Have you tested the color munki with displaycal and monitor calibration?

 

[mristuccia]

I use Argyll for my profiling, but I have an old rebadged i1pro. It does not have the license for xrite software.

Did you use the canson profile as the starting point for your custom profile?

And of course, how did you print the targets?

I follow these steps that torger published:

Printer profiling with Argyll and Colormunki

Hi anwarp,

Yes, I followed the very same link as the one you posted.

When I've made the profile with Argyll I did not specify any starting point/profile. Just started from scratch, because I wanted to simulate a situation where I have paper non coming from a well know vendor and without a profile already avaiable.

Of course, I print the target profiles without color management, by using Adobe ACPU utility.

However, in order to have as many patches as possible in one A4 sheet, I followed the indication on other forums where they suggest a series of targen parameters to obtain small but still readable patches, like the i1 pro ones. Maybe such small patches are creating the issue, I should try with the standard patch size for Colormunki, which is bigger than the one of i1 Pro.

 

[mristuccia]

Have you tested the color munki with displaycal and monitor calibration?

I did not, as I have an i1 Display Pro for calibrating my monitor.
Maybe I should try, in order to rule out any issues on Colormunki device.

 

[anwarp]

In my experience, using the manufacturer’s profile as the starting reference did give me a much smoother profile.
I asked about the printing because I did mess that up at one point trying to print directly from C1.
Somehow ACPU did not work well for me and I finally resorted to using qimage one.
I use windows.

Anwar