I spent last few days testing ArgyllCMS further but only with BT.1886 2.4 tone curve. Flow used:
dispcal.exe -v -dmadvr -c1 -yn -Yp -G2.4 -f0 -k0 -A4.0 "LUT"
targen.exe -v -d3 -e4 -B4 -s40 -g128 -m9 -b9 -f0 -N1.0 -V1.6 -p1.0
dispread.exe -v -dmadvr -c1 -yn -Yp -k LUT.cal
colprof.exe -v -qh -bl -aX -V1.6
collink.exe -v -qh -G -iaw -3m -et -Et -IB:2.4 -a LUT.cal Rec709.icm profile.icm "3DLUT_name.icm"
I realize some of the commands had no effect (like targen.exe -V1.6), but they did not harm anything either. Using targen.exe commands I listed above resulted in a more accurate 3DLUT and lower black levels than using:
targen.exe -v -G -e4 -B4 -s40 -g128 -m9 -b9 -f0 -N1.0 -V1.6 -p1.0 -c Rec709.icm -d3
To load and test a .cal file for MadVR, I used:
dispwin.exe -dmadvr -s LUT.cal
To generate a .cal file for desktop environment, I used:
dispcal.exe -v -d1 -c1 -yn -m -Yp -qh "-w0.3127,0.3290" -G2.4 -f0 -k0 -A4.0 "LUT"
To load and test a .cal file for desktop environment, I used:
Regardless of whether I loaded and tested a madVR 3DLUT or just a .cal file for desktop, each and every single case raised black levels. Checking generated .cal file or madVR 3DLUT black level with a colorimeter was not as effective as simply running a black clipping test pattern in MPC-HC. In each case, the original brightness setting required to achieve the lowest visible black was either 48 or 49. The end-result .cal file and madVR 3DLUT required a setting of either 46 or 47 to achieve the lowest visible black level. This doesn't bother me as much as I have a TV that let's me do that. Additionally, I managed to increase my contrast to improve CR, which allowed me to decrease brightness further, while retaining calibration accuracy. These post-calibration adjustments resulted in a better contrast ratio than the one I had before calibration. The results are still better than with ArgyllCMS 1.6.1, but still far from "slightly raised black levels". Another thing that did change was slighly increased banding when using a .cal file for desktop, visible in this test - http://www.lagom.nl/lcd-test/gradient.php
. I am not sure about madVR 3DLUT - I don't have a gradient pattern.