Thank you steve! I have ordered the Calibrite SL and it should be delivered in the next few days. I will repeat the whole test systematically and can then also measure the 21^3 Cube, which will be quicker with the Tristimulus. I hope I can get the problem under control.

Daniel

I got the Calibrite SL and did further tests. My realisation is that it is not due to the LUT or the measurement.
I have measured and verified a Cube 17^3 with the internal patch generator. The result is good.
However, when I load the LUT, the grey balance is distorted and banding can be seen in grey gradients.

I then borrowed a CG279X and repeated the test. The banding does not occur here even with a simple LUT from a GreyRampRGBLarge Profile.

The error must be related to the HDMI port of my CG247X. When I go from the Ultrastudio HDMI output to the DVI input of the monitor and use the same Cal Slot with the same LUT, the banding does not occur and the Grey balance is good.

Very strange, has anyone already observed this?

Daniel

Please post Verifications profiles for the LUT when enabled as an Active LUT within ColourSpace, and when the LUT is loaded into the monitor.

Steve

I have created two LUTs from the original 17^3 profile using PMActive with the following settings:
1. Hybrid, disable Gamut Mapping - Scaling: With PassBlack
2. Hybrid, disable Gamut Mapping -Scaling: Without PassBlack
(I have not uploaded the LUTs, they are too large)

I loaded the LUTs with PassBlack in the EIZO and verified them that way.
I verified Active LUT and EIZO with Unity Bypass, with and without PassBlack.

Folder Profiles:
EIZOCG247X_CUBE_17^3_PMActive_1.bcs
EIZOCG247X_CUBE_17^3.bcs
Folder Verifications:
EIZOCG247X_17^3_PMActive_VERIFICATION-ActiveLUT_noPassBlack.bcs
EIZOCG247X_17^3_PMActive_VERIFICATION-ActiveLUT_PassBlack.bcs
EIZOCG247X_17^3_PMActive_VERIFICATION-LoadedLUT_PassBlack.bcs

Daniel

Attached file: Original 17^3 Profile

Attached file: Verifications LUT from 17^3 Profile

The Active LUT and Loaded LUT are the profiles of interest.
Looking at them, the Loaded LUT is the better result, as it should be.
The Active LUT is obviously applied before all monitor electronics, so shows what 'signal path changes' are occurring within the monitor.
The original profile, and the LUT generation, takes those into account - hence the Loaded LUT result being better than the Active LUT.

As you say, the LUT isn't the issue.
Seems your monitor has a internal signal processing issue that is below the granularity of the measurements you have performed.
I doubt the i1D3 would be good enough to measure such small variation in grey scale, as you would need to use something like a 256 grey scale step patch set.

Steve

If the issue is related to the internal processing of the monitor, it may be an idea to adopt a Matrix Profiling approach, reducing the number of patches used to an absolute minimum, and generating what is basically a Technical LUT from a User Colour Space.

This is generating a LUT that is a colour space to colour space LUT, without the use of a Profile for LUT Generation.

I will add this info to the Advanced Operation page of the website.
But the idea is to profile with just 7 colour patches - Black, White, mid Grey, and Red, Green, & Blue.
From the generated profile, Use Modify/Extract to generate a new User Colour Space of the native response of the monitor.
Then generate a LUT using Rec709 as the Source, and the Extracted colour space as the Destination.

You may need to use colour patches that are not 100% peak - for example 235, 0,0 for Red, rather than 255, 0, 0.
And maybe an additional grey scale patch or two to help extract an accurate EOTF value.

Steve

I tested the matrix method today, it gives a slightly better result. You can still see steps (banding) in the grey gradient but no more colour shifts. However, this method does not produce a better representation of dark grey tones than Colornavigator, the result looks pretty much the same. The problem is only present at the HDMI input of the monitor, I now use Ultrastudio HDMI out to Eizo DVI In with an appropriate cable. I now have very good verification and clean grey gradients.

Thank you Steve for your support.
Daniel

That really does suggest an issue with the HDMI input on that monitor.
Or the cable used, if you have not tried a different one.

Steve