- This topic has 45 replies, 2 voices, and was last updated 7 years, 7 months ago by
.
-
Posts
-
This looks a bit like an issue of TV vs PC levels (although the measured black is much, much too light even considering that possibility). E.g. if the TV expects TV levels, then either the graphics card needs to output TV levels as well (and madVR needs be set to PC levels), or the graphics card needs to output PC levels (and madVR needs to be set to TV levels). Note that the latter option has the drawback that you’ll get clipping while not using madVR.
Hi Florian,
these settings should be correct i have checked it with a levels pattern before. My system is setup as follows:
- Nvidia GTX 1080 – PC (0-255) RGB – also set with madvr Levels Tweaker
- LAV Filters output as input (unchanged)
- Madvr TV Levels 10bit
- TV Blacklevel set to low (for LG this should be TV Levels)
As i figured this out to be the way with less possible conversion made…
I would switch to a mode using TV set to blacklevel high (PC Levels) for gaming for example..
Something else just came to my mind… i am using my notebook and madvr over the network on my HTPC to calibrate and profile. As Dispcalgui is running on my notebook and not on the PC the TV is hooked to the 3DLut cant be applied automatically after the profuling is done right?
So how can Dispcalgui do any verification and proper reporting that way? Could this be the reason or at least one of the reasons?
So how can Dispcalgui do any verification and proper reporting that way?
Taking measurements over a network connection works transparently just like a local connection.
Could this be the reason or at least one of the reasons?
I don’t think so. I wouldn’t expect a raise in black level, for example.
Hm ok.. but the 3D Lut was actually saved on my notebook.
But i guess in this case thats not the issue so something else must have gone … no clue as i used the settings you recommended.
One more thing i recognized was that was only able to setup white level using RGB and White luminance while initial calibration. Black level was not included… guess thats normal.. but clicking on the Info button there was a point blacklevel wich was kind of ignored i guess as nothing was shown there.
No clue what went wrong then 🙁
Hm ok.. but the 3D Lut was actually saved on my notebook.
That’s normal. When you click “Install 3D LUT” after it has been created, it is transferred to the madVR instance you’re connected to (which doesn’t have to be on the same machine).
One more thing i recognized was that was only able to setup white level using RGB and White luminance while initial calibration. Black level was not included… guess thats normal.
Correct. There’s not much point in offering separate black point adjustment, because there’s really only two useful approaches: Leave black at the native level (for maximum contrast), or set black point to the same hue as the whitepoint (which raises the black level and decreases contrast, thus is usually not done for video calibration).
No clue what went wrong then
Me neither unfortunately. Only thing that comes to mind is that this could be a weird side-effect of A(S)BL.
Ok that makes some thing more clear to me ! Thank you 🙂 ASBL was turned off using the service menu before..
Guess i will start a second try tonight.
Just one more thing what peak luminance would you recommend i was choosing white level 120 nits as this seems appropriate for Rec.709. Can i use that setting as well for Rec.2020 3DLut calculation after i got the Rec709 Lut ?
Or does is it even matter which Luminance is chosen? (as i think colors change with it?!)
Just one more thing what peak luminance would you recommend i was choosing white level 120 nits as this seems appropriate for Rec.709
It is fine to alter the value if you feel it is too dim or too bright for your environment.
Can i use that setting as well for Rec.2020 3DLut calculation after i got the Rec709 Lut ?
If you’re aiming for HDR SMPTE 2084, then I would recommend to crank the TV brightness up as much as you can, because I feel tonemapped HDR looks best when the range where most of the significant detail is isn’t compressed too much.
I was thinking of a normal UHD SDR BT.2020 in this case for now as there is almost no HDR content available that can be played via an HTPC except of th all the demo stuff but i guess its not possible to get metadata passthrough using madvr yet.
I was thinking of a normal UHD SDR BT.2020 in this case
Alright. In that case same brightness as for Rec. 709 will work fine.
Perfect thanks:) Than i can have Dispcal calculate the BT.2020 Lut from the same base using the same picture mode:)
Will start a second try today… do you happen to know which color gammut should be used on TV side… I´ve set it to “extended” in order to support BT.2020 as well. Could this be the reason for my previous issues?
On the other hand i have read that one should use the widest gammut the TV supports so i thought this would be correct.
[…] i have read that one should use the widest gammut the TV supports so i thought this would be correct.
Yes, to give the 3D LUT the most possible “room” to work with, it is a good idea to use the widest available gamut.
Sounds good!
Strange thing i just recognized that i cant match White level in windowed mode 10 % and full 100%.
But i guess thats due to ABL… as i am getting round about 105nits in 100 % while using 10 % its about 250 nits…don´t know if i have to take care about that or just leave that aside for know. And when opening the TV menu for adjusting the OLED Light and contrast to match round about 120 nits in 100% i can easily do that… after closing the menu it drops to round about 105 nits again.
Strange thing i just recognized that i cant match White level in windowed mode 10 % and full 100%.
That would indicate some form of ABL is still active, and it’s not a good idea to use a pattern area that is larger than absolutely necessary to fit the instrument (which usually comes down to around 5%-10%).
