Hi,
I am using LTE version with DaVinci Resolve as PG calibrating my Asus PA32UCDM with I1D3 and probe matching (extra delay 0,5 s, stabilisation 0,5 s, integration time 1,25 s). So far I have achieved excellent results on Rec709 gamma 2.4 with deltaE<0,5 and EOTF spot on (validated on Grey Ramp RGBCMY +). However, after creating a profile based on cube 15^3 (from native state) and creating a LUT converting it to Rec709 I saw the result: 99% within target gamut, but was actually expecting 100%...

Now, calibrating for DCI P3 D65 gamma 2.6 after making the same steps during LUT conversion I get 88% (only)... (but deltaE & EOTF after validation are still great! ). So, sorry maybe for a noob question, but, shall I be worried about it? Is this number indicating gamut coverage, which acc. to Asus should be ca. 98%? Should I get better results when profiling with f.i. 21^3 cube? Does the fact, through DaVinci Resolve patches are 8bit matters here (except for the EOTF accuracy)? Just trying to understand...

Thanks for any help or explanation.

First, see 'Volumetric Gamut Coverage' in the user guides, as well as the info on the 'Gamut Coverage' option in the Interactive User Manual.

And the bit depth used is irrelevant, even for EOFT accuracy, when doing LUT calibration.

Steve

Thank you for the reply Steve. So, is my conclusion correct, that:
- producers like Asus for marketing use 2d gamut coverage data, which in reality is not accurate and overestimated data?
- then, taking into account the above, the volumetric 88% might be quite a good result?
- finally, incresing the accuracy of a cube from 15 to let's say 21 may/should even gently increase the result? (will test it)

Best to post you profile so it can be assessed.
But so far all looks as expected.
And yes, more measurement points will make the value more accurate, as the user guides say.
And yes, that usually means a slightly higher value.

Steve

Hi again,
so I made another profile based on a cube 21^3 this time, the gamut conversion was actually even slightly worse - 85%, but the validation gave very good results (deltatE=0,6417 and EOTF=2.5937). How can I post the accesible profile to check? I exported it to .bcs file, but it cannot be attached here...
Regards, Marek

Zip the profile, and post it.

Steve

Ok, here it is.
Marek

Attached file: P3 D65 2.6 17^3 12.02.2026.zip

The EOTF (gamma) of the display as shown in the profile is approx. 2.2.
P3 D65 is 2.6.
If you target a colour space that has P3 D65 with a 2.2 EOTF, AND make the target have the same Min/Max Luma you get a coverage of 91%.
(Use the ^ arrow in Setting to set the Target Min/Max the same as the profile, or set to Auto.)

But, your black is rather high?
And you have strange/error reading here and there...
Oh - and the missing 9% is mainly due to the red/green edge being low gamut, which is unusual - as you sure the probe match is accurate?

Steve

Hi Steve,
thanks for your reply. So now I am a little confused...
So my workflow is following; after uploading a Unity Bypass on a native state I target DCI P3 D65 (when I check this color space by Modify - the EOTF is 2.6). Then I make gray scale and measure black and white point. By the way; the black point is 0.0129, is this high? I refered to Netflix ercomendation to get lower than 0,05 nits...
Then I made the profile incl. probe matching and saved this profile. Next based on this profile I made a LUT (source DCI P3 D65) and this profile as destination, id addition limiting luma to 100 nits, Finally after uploading this LUT and validation with Grey Ramp RGBCMY+ I got (deltatE=0,6417 and EOTF=2.5937).
Should I get this 95% or more on this monitor? I will test maybe again without probe matching..

The native EOTF of just about any monitor is 2.2, as that is the EOTF manufactured into the glass.
So when you do a Gamut Coverage verification any difference in the EOTF will result in a poorer values if the display cannot totally cover the target gamut.

If you have deliberately set the black to be 0.02ish, that is fine.
(It is actually a goo thing to do with OLEDs.)
I was just checking you were aware of the value, as OLEDS basically have a zero black, but do then tend to clip as they all suffer near black hysteresis.

The profile you provided was the native profile, not the calibrated verification.

And how are you probe matching, as the green/red edge is not normally so low gamut vs. the target on those monitors.
(Test with the Correlation File instead of probe matching.)

Edit: I just realised you are quoting the gamut coverage after a LUT is generated.
I used the option within Manual Measure, hence the difference with altering the target colour space values!
When actually generating a LUT the value will vary with LUT Generation Size (if you have a higher license level), exactly as per the number of points in the profile when you use the option in Manual Measure.
With a 128^3 LUT the result is 89%.
Anything around 90% is exactly what I would expect for that profile.

Steve

Hi Steve,
Today I tested with Correlation File and after profiling native state during LUT conversion I got 91% of DCI P3 D65. Would you please find a moment to look shortly on the profile? From what I see the red colors are slightly cut...
Regards, Marek

Attached file: DCI P3 D65 2.6 15^3 13.02.20.zip

Looks basically similar...
Guess that monitor is just low gamut on the green/red edge.
You do still have the occasional bad reading though.

Steve

Thank you Steve.