I see many times that people use only ColorChecker Classic to evaluate the display's performance after a 3D LUT generation.
ColorChecker Classic is the default measurement method to measure the performance of a display to some calibration software.
ColorChecker Classic is enough to evaluate a performance of a display since it will not measure any edge color; it's just a quick and very limited patchset for a quick evaluation.
ColorChecker SG contains more colors but still is not covering a lot of areas of the gamut.
When users want to evaluate the difference between different patchsets or cube sizes; profiling from 100 measured colors or from 729 (9-Point Cube), or larger cube size (1000/3000/5000/10000 colors), it's recommended to measure with 10-Point Cube patchset (1000 patches) to evaluate for differences in performance.
10-Point Cube patchset can be used also to evaluate with a lot more detail the results of a manual calibration using regular calibration controls (10/20-Point Grayscale and 6 Color CMS).
As we know, internal TV CMS controls are limited to 6 adjustable control points (primary+secondary colors) to modern TVs / projectors.
These six control points will try to reduce the errors for the whole color gamut.
Usually, the default patchlists to perform CMS calibration based on RGB-CMY colors with:
100% Saturation with 100% Luminance
100% Saturation with 75% Luminance
75% Saturation with 100% Luminance
75% Saturation with 75% Luminance
Some displays can be calibrated better with a specific selection of Saturation and Luminance level.
When you are using 75% Saturation patterns, it may reduce the color errors in ColorChecker (since ColorChecker measures colors away from the edge of the gamut) while it will increase the errors to borders of the gamut.
The ideal method to evaluate for all these is to perform a calibration using a specific CMS saturation/luminance level and then verify the performance with 1000 patches.
If time is available, perform a new calibration using a different saturation/luminance level and verify with 1000 patches also.
Ideally, it will require performing 4 CMS calibrations and always verify with 1000 patches.
After comparing the CIE chart of ColorSpace ZRO with 1000 points, the user will know which CMS patterns provide better performance with lower dE, so the user will know which particular patterns to use in future calibrations of that specific display/projector.
This method is a good practice for users to understand better the capabilities of the display.
The animated PNG below compares the color points measured between ColorChecker Classic / ColorChecker SG / 10-Point Cube:
For better understanding since the 2D CIE chart doesn't show the luminance levels of the patches, below you see normalized RGB Cube pictures of REC.709 colorspace.
These colors are measured using ColorChecker Classic:
Even ColorChecker SG is a limited patchset, not measuring edge colors:
10-Point Cube provides a deeper coverage with different saturation/hue/luminance levels and a lot of color points to the border of the gamut:
* These 3D Charts is not available to ColourSpaceZRO; I took that picture from ColorSpace HTL.1000 Point volumetric verification run using ColourSpaceZRO
About the 10-Point Volumetric Verification patchset, you can download the patchsets from here
The 10-Point Cube (09 April 2021) ZIP will contain three folders with 1000 point patchsets for different patchscale settings:FULL
(0-255)...PC users / DaVinci ResolveLEGAL
(16-235)... Consumer TV/projector evaluationEXTENDED
(16-255)... LG OLED 3D LUT evaluation
You can see this video with 0.5x Speed and enabled subtitles about how to use them: