- This topic has 30 replies, 2 voices, and was last updated 6 years, 1 month ago by
.
-
Posts
-
I don’t even see a setting for PC or TV levels in my graphics card panel.
What’s your graphics card? You can set video levels in DisplayCAL (advanced option) in case the grahipcs driver software doesn’t let you choose (or chooses the wrong levels).
Do you not think the TV expects full range levels?
Unlikely. Unless the TV explicitly lets you choose 0-255 input (I don’t think Philips ever offered that option for its LCD TVs, mine doesn’t have it as well), it will be expecting video levels.
My graphics card is Radeon HD 7770. Why does it look dark and contrasty after profiling with display output levels set to TV range in Displaycal? Why does it look more similar to my PC monitor after profiling with display output levels set to full range in Displaycal?
Several problems:
- The TV is still not setup correctly. This is the first thing that you need to fix. Use a “pluge” pattern in a non-color managed image viewer (NOT video player) to setup the correct brightness (backlight) setting for the TV, also saturation sweeps (make sure none of them clip), or use a standalone (NOT software) BluRay or DVD player with (e.g.) the AVS HD 709 disc.
- TV levels. Ideally you should set that in the graphics card settings (for AMD, the setting is called “Pixel format” under “My digital flat panel displays”), and let DisplayCAL output full range (the graphics card will convert full range to TV levels). That way, the desktop will also look correct even without calibration.
Thanks for the advice. I have adjusted the backlight level. Neither my AMD or iGPU allow me to set the level range. I suspect the TV was manufactured with incomplete EDID data or something. In the meantime I should set DisplayCal and MadVR to 16-235?
I noticed now that DisplayCAL can do 0-255 to 16-235 scaling via calibration curves. So I shouldn’t matter that my graphics card panel doesn’t allow to chose levels, right?
I get an error after finishing a verification run with “use simulation profile as display profile” unchecked:
Calibrite Display Pro HL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.Sorry, I cannot reproduce this. I would need detailed step-by-step on what your settings are, and what you did (and in which order).
I create a profile. For example I select “Default (Gamma 2.2)” under settings and press “Calibrate & profile”. After the profiling run finishes i install the new profile. Next I select the verification tab. I select sRGB and check “simulation profile” and hit “Measurement report”. After it reads the last patch I get the “key does not exist” error.
If I check “use simulation profile as display profile” I the error does not show up. If Iunchecked “simulation profile” it finishes without error also.
Sorry, still can’t reproduce. What settings do you use on the “Display & instrument” tab? Please attach the profile and related files (compressed archive) as well.
Sorry, still can’t reproduce. What settings do you use on the “Display & instrument” tab? Please attach the profile and related files (compressed archive) as well.
Sure. Please see below.
I get an error after starting “report on calibrated device”. “report on uncalibrated device” ends with no error. Maybe its a related issue?
Not getting the error anymore. I did two things, not sure which solved it: #1 Moving the ArgyllCMS folder to the drive root. #2 Installed the display profiles as system-wide instead of per-user.
I did two things, not sure which solved it: #1 Moving the ArgyllCMS folder to the drive root. #2 Installed the display profiles as system-wide instead of per-user.
None of these are related 🙂
Here’s the run-down of the actual cause of the problem: When measuring with video levels set in DisplayCAL (doesn’t matter if autodetected or explicitly selected), ArgyllCMS quantizes the test patch device RGB values in the resulting measurement file to 8 bits by default. Normally this doesn’t cause an issue, but when doing a verification with a simulation profile (not using it as display profile), the input CIE values are looked up through the display profile to obtain the target device RGB values, but these device RGB values are not quantized to 8 bits. So, if by chance these values end up not lining up with the quantized test values for white (which depends on the display profile), there’s an unexpected mismatch between the measurement file and the looked up device RGB values. The solution is to quantize the target device RGB values in the same way as the device values in the resulting measurement data. This will be fixed with the next update.
