So, if you remember, I had experienced some very bad results with Calman which I described in this post
in the main Calman 5 thread.
I was trying to calibrate the TV as a PC monitor, that is: Calman configured for full range patterns, RGB graphics card with full range output, using the TV with full range input (Black level set to High), on a HDMI input marked as PC.
The end result was that when viewing a image like this:
it would look like this on the TV:
I left Calman aside for a month, and now I noticed in this thread that WiFiSpy recommends to do the full range calibration using the TV with the input marked for video levels (Black level set to Low).
Indeed, this way the calibration of the grayscale no longer causes the issues with the patches near to 235. More precisely: Calman configured in its Workflow Basic Options setting tab with Luminance Levels Video = 16-235, doing a Autocal Grayscale for 16-255 (0-109%), graphic card configured for full range 0-255 output, and TV input configured for video levels (Black Level Low). Switching the TV input afterwards back to PC levels (Black level set to High), does not give the same response as when the TV had TV levels, but it's good enough.
Measuring something, and then using something else, is not the right way to do calibration though, I am surprised that none of its customers is asking Portrait Displays to implement 0-255 8 bit RGB calibration properly.
Unfortunately, I am not able to do a calibration of colors using anything else but the Matrix LUT. When I try using Lightning LUT for instance, Calman fails after some time with this message:
And if I do what the dialog say, that is to enable Sync for the i1D3, it still fails, only that it happens much quicker:
There are no such issues when doing the Grayscale AutoCal, when doing pre and post calibration measurements, or when creating the i1D3 correction matrix. Only the saturation AutoCal fails like this.
This is a Rev.A i1D3, btw.
Any tips on how to get rid of this issue?