madVR - high quality video renderer (GPU assisted) - Page 3235

huhn · 15th April 2024, 01:31

https://www.itu.int/dms_pubrec/itu-r...3-I!!PDF-E.pdf
have a nice read.

the important part is there focus on CRT. so it'S supposed to be used event through we clearly know they didn't look at bt 1886 well we also know they did because bt 1886 is gamma 2.4 on perfect black displays. and on none perfect black display like CR 1000 it is even brighter in black parts then sRGB is.

i can tell you they didn't use the bt 709 curve (they are not supposed to they where using what ever they want) the spec even warns about using it for an EOTF:
https://www.itu.int/dms_pubrec/itu-r...6-I!!PDF-E.pdf
In typical production practice the encoding function of image sources is adjusted so that the final picture has the
desired look, as viewed on a reference monitor having the reference decoding function of Recommendation
ITU-R BT.1886, in the reference viewing environment defined in Recommendation ITU-R BT.2035.

it's an encoding function anyway not a decoding function like gamma 2.2.

tp4tissue · Yesterday, 04:14

Let me give you a solid example huhn.

https://ibb.co/NV3DcNG
https://ibb.co/dLfQLdh

This frame from the outlaw star bluray. My display is calibrated for 2.4 pure power. I've tried 2.2 pure, 2.1, but the entire bluray series has tons of crush.

UNTIL you either use srgb, or the bt709/601 gamma option. Then the details on the pipes just pop right out. Viewing the image in madvr, press play, or else the gamma function won't shift during pause.

I've tried this on all my LCDs calibrated to pur 2.4, none of them look right. ALOT of older blurays exhibit this problem. I think this is what someone posted about a long time ago on this forum, green crush, where pure 2.4 is NOT the intended target curve of older grades?

huhn · Yesterday, 09:44

i can only check with an ycbcr file.

this looks like wrong levels old anime doesn't have a perfect black in the first place. i'm not blaming you for the wrong levels. levels in master is not that rare.
try the newest beta and press control+shift+alt+print
there should be a video screenshot on the desktop RawScreenshots.
you also need ffmpeg in the madVR folder.

just to be clear relative 2.4 or absolutely?

here is sRGB:
https://drive.google.com/file/d/1360...ew?usp=sharing
there is a chance it is used.

green should be the issue the difference between gamma and sRGB is linear in terms of color.

tp4tissue · Yesterday, 15:03

Huhn, I'm not disputing that bt1886 2.4 "is" the standard, and is "most probably" the correct default.

My point of discussion here is that "where-ever" the mismatch occurred, mastering-hardware, colorist, compressionist, etc, such "mis-matches" are so incredibly common among the published material, that, it may be likely for these odd-curves to line up serendipitously as the correct display offset on our end^?

The screen shot is accurate as far as demonstrating the effect I'm talking about, it's from an untouched file. I'll look into this new beta, still on 113.

But can you take a look at the dark values, the difference between the shadow detail is rgb 0.1.1 and 3.5.4. That is barely anything. On purepower 2.4, you can not see that difference as detail, especially if the APL of the scene is enough to triggers your eyes to shutter.

Is it not crushed on your calibrated setup?

huhn · Yesterday, 22:38

i don't know i don't have the ycbcr image.

the RGB are very close to be dead on my oled on my LCD screen it is fine.
i have relative not absolute gamma on the LCD.

on at 1000 CR bt 1886 should elevate the black even more then sRGB.

and if they are from a BD then they are remastered to the standard used at the time the BD is released.

i would just old something like the sRGB 3d LUT i would clearly not use bt 709. sRGB is an actually used spec.

with a ycbcr file i may have a better idea what is actually happening. maybe this is far from a guarantee.

Yesterday, 04:14	#64682 \| Link
tp4tissue Registered User Join Date: May 2013 Posts: 710	Let me give you a solid example huhn. https://ibb.co/NV3DcNG https://ibb.co/dLfQLdh This frame from the outlaw star bluray. My display is calibrated for 2.4 pure power. I've tried 2.2 pure, 2.1, but the entire bluray series has tons of crush. UNTIL you either use srgb, or the bt709/601 gamma option. Then the details on the pipes just pop right out. Viewing the image in madvr, press play, or else the gamma function won't shift during pause. I've tried this on all my LCDs calibrated to pur 2.4, none of them look right. ALOT of older blurays exhibit this problem. I think this is what someone posted about a long time ago on this forum, green crush, where pure 2.4 is NOT the intended target curve of older grades? __________________ Ghetto \| 2500k 5Ghz Last edited by tp4tissue; Yesterday at 05:04.

Yesterday, 15:03	#64684 \| Link
tp4tissue Registered User Join Date: May 2013 Posts: 710	Huhn, I'm not disputing that bt1886 2.4 "is" the standard, and is "most probably" the correct default. My point of discussion here is that "where-ever" the mismatch occurred, mastering-hardware, colorist, compressionist, etc, such "mis-matches" are so incredibly common among the published material, that, it may be likely for these odd-curves to line up serendipitously as the correct display offset on our end^? The screen shot is accurate as far as demonstrating the effect I'm talking about, it's from an untouched file. I'll look into this new beta, still on 113. But can you take a look at the dark values, the difference between the shadow detail is rgb 0.1.1 and 3.5.4. That is barely anything. On purepower 2.4, you can not see that difference as detail, especially if the APL of the scene is enough to triggers your eyes to shutter. Is it not crushed on your calibrated setup? __________________ Ghetto \| 2500k 5Ghz Last edited by tp4tissue; Yesterday at 15:08.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

15th April 2024, 01:31	#64681 \| Link
huhn Registered User Join Date: Oct 2012 Posts: 7,923	https://www.itu.int/dms_pubrec/itu-r...3-I!!PDF-E.pdf have a nice read. the important part is there focus on CRT. so it'S supposed to be used event through we clearly know they didn't look at bt 1886 well we also know they did because bt 1886 is gamma 2.4 on perfect black displays. and on none perfect black display like CR 1000 it is even brighter in black parts then sRGB is. i can tell you they didn't use the bt 709 curve (they are not supposed to they where using what ever they want) the spec even warns about using it for an EOTF: https://www.itu.int/dms_pubrec/itu-r...6-I!!PDF-E.pdf In typical production practice the encoding function of image sources is adjusted so that the final picture has the desired look, as viewed on a reference monitor having the reference decoding function of Recommendation ITU-R BT.1886, in the reference viewing environment defined in Recommendation ITU-R BT.2035. it's an encoding function anyway not a decoding function like gamma 2.2.

Yesterday, 09:44	#64683 \| Link
huhn Registered User Join Date: Oct 2012 Posts: 7,923	i can only check with an ycbcr file. this looks like wrong levels old anime doesn't have a perfect black in the first place. i'm not blaming you for the wrong levels. levels in master is not that rare. try the newest beta and press control+shift+alt+print there should be a video screenshot on the desktop RawScreenshots. you also need ffmpeg in the madVR folder. just to be clear relative 2.4 or absolutely? here is sRGB: https://drive.google.com/file/d/1360...ew?usp=sharing there is a chance it is used. green should be the issue the difference between gamma and sRGB is linear in terms of color.

Yesterday, 22:38	#64685 \| Link
huhn Registered User Join Date: Oct 2012 Posts: 7,923	i don't know i don't have the ycbcr image. the RGB are very close to be dead on my oled on my LCD screen it is fine. i have relative not absolute gamma on the LCD. on at 1000 CR bt 1886 should elevate the black even more then sRGB. and if they are from a BD then they are remastered to the standard used at the time the BD is released. i would just old something like the sRGB 3d LUT i would clearly not use bt 709. sRGB is an actually used spec. with a ycbcr file i may have a better idea what is actually happening. maybe this is far from a guarantee.