Improving Madvr HDR to SDR mapping for projector - Page 196 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 5576Likes
Reply
 
Thread Tools
post #5851 of 6789 Old 05-11-2019, 12:20 PM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,739
Mentioned: 526 Post(s)
Tagged: 0 Thread(s)
Quoted: 2605 Post(s)
Liked: 3476
Quote:
Originally Posted by Soulnight View Post
I have programmed something for the madmeasurehdr optimizer tool to recognize such area (flat bright area) based on the histogram information only wth the help of the examples provided by @Neo-XP and @Fer15 (my thanks to them ). It works pretty well. I hope to release it this WE.
Can you make this available for both FALL and BT2390 algos? I'm asking because it seems some users prefer FALL and some BT2390 algos.
Manni01 likes this.
madshi is offline  
Sponsored Links
Advertisement
 
post #5852 of 6789 Old 05-11-2019, 12:42 PM - Thread Starter
AVS Forum Special Member
 
Soulnight's Avatar
 
Join Date: Dec 2011
Location: Germany
Posts: 1,402
Mentioned: 215 Post(s)
Tagged: 0 Thread(s)
Quoted: 1035 Post(s)
Liked: 1611
Quote:
Originally Posted by madshi View Post
Can you make this available for both FALL and BT2390 algos? I'm asking because it seems some users prefer FALL and some BT2390 algos.
It's already build like this and available for both.

I'll make it all available to you ("sky detection" and bt2390algo ) once people have tested it a bit.

I hope to release it "officially" by tomorrow.
Soulnight is offline  
post #5853 of 6789 Old 05-11-2019, 12:48 PM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,739
Mentioned: 526 Post(s)
Tagged: 0 Thread(s)
Quoted: 2605 Post(s)
Liked: 3476
Great, thanks!
madshi is offline  
Sponsored Links
Advertisement
 
post #5854 of 6789 Old 05-11-2019, 01:47 PM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 993
Mentioned: 151 Post(s)
Tagged: 0 Thread(s)
Quoted: 700 Post(s)
Liked: 875
Quote:
Originally Posted by madshi View Post
So you're saying these brightness reaction speeds should not increased further for anything above 200nits? However, when going below 100nits, maybe they should be reduced?
I didn't find the necessity to do it for a dynamic tuning value of 50.

I don't know if it makes sense to set a dynamic tuning value higher than 50 with only 50 real display peak nits, but if it does, the brightness speeds have to be decreased like this to be safe:

Dynamic tuning 50 / 75 / 100



Quote:
Originally Posted by madshi View Post
What is your latest opinion of Metric1 and Metric2? I assume the combination of both works best, since that's what your settings are currently using? Did you notice (m)any scene detection problems when watching HDR movies during my absence?
I still don't trust Metric2 too much, but with a low value it helps a bit and doesn't hurt too much.
For the last movies I watched, the scene detection didn't cause me any issue (once enabling the sub).


Quote:
Originally Posted by madshi View Post
And what about the "disable sub of previous frame" option? You seem to recommend to users to actually do the sub, but you seem to have it disabled in your settings. Can you clarify if you think this option is useful or not? (For tests it should be disabled, of course).
This option is the most useful to avoid a whole lot of false positives (enabling the sub). I have it disabled in my settings on purpose, so I can check if the false positives can be avoided by enabling the sub when I notice a false positive

Most of the time it does, unless there is a sudden high Metric1 value for no real reason (but that can maybe be avoided with some histogram smoothing).

Quote:
Originally Posted by madshi View Post
2) Brightness adjustment speed is "infinite" during fade-ins now.
It doesn't work for me with the fade in at the beginning of HP and the Deathly Hallows part 2 (http://www.mediafire.com/file/qtb3v8..._jump.mkv/file ).

It would also be nice to have it for fades out I think.
For instance, for the BvS scene at 02:11:13 to 02:11:19. When the green spear goes in/out of frame, a fade in/out is detected, but the target is too slow to adapt.

Last edited by Neo-XP; 05-11-2019 at 01:54 PM.
Neo-XP is offline  
post #5855 of 6789 Old 05-11-2019, 02:14 PM
Advanced Member
 
chros73's Avatar
 
Join Date: Jan 2015
Posts: 524
Mentioned: 12 Post(s)
Tagged: 0 Thread(s)
Quoted: 313 Post(s)
Liked: 140
Quote:
Originally Posted by tommarra View Post
Personally I like it turned off because I don’t like the image sharpness to be impacted due to tone mapping
Which value did you use? Try 25%.
tommarra likes this.

Ryzen 5 2600,Asus Prime b450-Plus,16GB,MSI GTX 1060 Gaming X 6GB(v385.28),Win10 LTSB 1607,MPC-BEx64+LAV+MadVR,Yamaha RX-A870,LG OLED65B8(04.10.25+PC4:4:[email protected]/24/25/29/30/50/59/60Hz)
chros73 is offline  
post #5856 of 6789 Old 05-11-2019, 04:07 PM
AVS Forum Special Member
 
Javs's Avatar
 
Join Date: Dec 2012
Location: Sydney
Posts: 8,029
Mentioned: 490 Post(s)
Tagged: 0 Thread(s)
Quoted: 6917 Post(s)
Liked: 6620
Quote:
Originally Posted by chros73 View Post
Which value did you use? Try 25%.
Yeah I had to turn Dynamic Clipping down to 15%.

100% just clips way too much.

I am not sure if its a bug or not, or the settings just differ from how they do in the measurements tool from Soulnight and Anna.
tommarra likes this.

JVC X9500 (RS620) | 120" 16:9 | Marantz AV7702 MkII | Emotiva XPA-7 | DIY Modular Towers | DIY TPL-150 Surrounds | DIY Atmos | DIY 18" Subs
-
MadVR Settings | UHD Waveform Analysis | Arve Tool Instructions + V3 Javs Curves
Javs is online now  
post #5857 of 6789 Old 05-11-2019, 05:52 PM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 993
Mentioned: 151 Post(s)
Tagged: 0 Thread(s)
Quoted: 700 Post(s)
Liked: 875
I have a clear winner:

stevenjw, madshi and Manni01 like this.
Neo-XP is offline  
post #5858 of 6789 Old 05-11-2019, 05:57 PM
AVS Forum Special Member
 
Javs's Avatar
 
Join Date: Dec 2012
Location: Sydney
Posts: 8,029
Mentioned: 490 Post(s)
Tagged: 0 Thread(s)
Quoted: 6917 Post(s)
Liked: 6620
Quote:
Originally Posted by Neo-XP View Post
I have a clear winner:

Which settings are these?

JVC X9500 (RS620) | 120" 16:9 | Marantz AV7702 MkII | Emotiva XPA-7 | DIY Modular Towers | DIY TPL-150 Surrounds | DIY Atmos | DIY 18" Subs
-
MadVR Settings | UHD Waveform Analysis | Arve Tool Instructions + V3 Javs Curves
Javs is online now  
post #5859 of 6789 Old 05-11-2019, 06:39 PM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 993
Mentioned: 151 Post(s)
Tagged: 0 Thread(s)
Quoted: 700 Post(s)
Liked: 875
@Javs

From the last build:

Quote:
Originally Posted by madshi View Post
Neo-XP is offline  
post #5860 of 6789 Old 05-11-2019, 06:45 PM
AVS Forum Special Member
 
Javs's Avatar
 
Join Date: Dec 2012
Location: Sydney
Posts: 8,029
Mentioned: 490 Post(s)
Tagged: 0 Thread(s)
Quoted: 6917 Post(s)
Liked: 6620
Quote:
Originally Posted by Neo-XP View Post
@Javs

From the last build:
I meant your settings?

Do I just follow the settings from here:

https://www.avsforum.com/forum/24-di...l#post58033168

JVC X9500 (RS620) | 120" 16:9 | Marantz AV7702 MkII | Emotiva XPA-7 | DIY Modular Towers | DIY TPL-150 Surrounds | DIY Atmos | DIY 18" Subs
-
MadVR Settings | UHD Waveform Analysis | Arve Tool Instructions + V3 Javs Curves
Javs is online now  
post #5861 of 6789 Old 05-11-2019, 09:45 PM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,739
Mentioned: 526 Post(s)
Tagged: 0 Thread(s)
Quoted: 2605 Post(s)
Liked: 3476
Quote:
Originally Posted by Neo-XP View Post
I have a clear winner:

Oh wow, that looks pretty good - almost no false positives, anymore!! And it seems *all* of the missed true scene changes are clearly detected by Metric1, so combining both should produce great results?
stevenjw, Manni01, NoTechi and 1 others like this.
madshi is offline  
post #5862 of 6789 Old 05-11-2019, 11:17 PM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 993
Mentioned: 151 Post(s)
Tagged: 0 Thread(s)
Quoted: 700 Post(s)
Liked: 875
@Javs @madshi

Once the sub of prev frame option is enabled, I get an extremely good detection ratio with these:

stevenjw, CoryW and Manni01 like this.
Neo-XP is offline  
post #5863 of 6789 Old 05-12-2019, 05:39 AM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 993
Mentioned: 151 Post(s)
Tagged: 0 Thread(s)
Quoted: 700 Post(s)
Liked: 875
@madshi Now that Metric2 is more reliable, I was thinking that we could set a minimum threshold which has to be reached (for Metric1&2), for a scene to be detected.

For instance:

Metric1

Threshold: 12
Minimal threshold: 4

Metric2

Threshold: 6
Minimal threshold: 2

For an immediate change to happen, the combination of both Metrics would have to be >= 10 (like it is now), but in addition, Metric1 would have to be >= 4 and Metric2 >= 2.

By doing this, even if Metric1 is extremely high due to a bar shift in the histogram, Metric2 minimum threshold would prevent the false positive (Metric2 is very close to 0 on these cases + no need for smooth histogram).

Just an idea passing by...

Edit: This sample contains the brightness speeds issue (not adapting immediately) during fades in, followed by multiple false positives caused by Metric1 and "ignored" by Metric2: http://www.mediafire.com/file/i11tmq...tives.mkv/file
madshi, Manni01 and tommarra like this.

Last edited by Neo-XP; 05-12-2019 at 06:18 AM.
Neo-XP is offline  
post #5864 of 6789 Old 05-12-2019, 06:43 AM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,739
Mentioned: 526 Post(s)
Tagged: 0 Thread(s)
Quoted: 2605 Post(s)
Liked: 3476
Quote:
Originally Posted by Neo-XP View Post
@madshi Now that Metric2 is more reliable, I was thinking that we could set a minimum threshold which has to be reached (for Metric1&2), for a scene to be detected.

For instance:

Metric1

Threshold: 12
Minimal threshold: 4

Metric2

Threshold: 6
Minimal threshold: 2

For an immediate change to happen, the combination of both Metrics would have to be >= 10 (like it is now), but in addition, Metric1 would have to be >= 4 and Metric2 >= 2.

By doing this, even if Metric1 is extremely high due to a bar shift in the histogram, Metric2 minimum threshold would prevent the false positive (Metric2 is very close to 0 on these cases + no need for smooth histogram).

Just an idea passing by...
I have something like this (not exactly the same but similar in concept) planned for test build 82. There will be 2 more metric2 specific test builds before that, though: Build 80 to fine tune Metric2 quality. And build 81 to check if I can simplify the Metric2 math without losing quality, in the hope of maybe speeding processing up a bit, for those of you guys with slower GPUs.

I'm now waiting for Fer15 to complete his tests before I move on to build 80. I'm pretty confident that Fer15's results will match yours, but I'll wait, just to be safe.
madshi is offline  
post #5865 of 6789 Old 05-12-2019, 08:13 AM - Thread Starter
AVS Forum Special Member
 
Soulnight's Avatar
 
Join Date: Dec 2011
Location: Germany
Posts: 1,402
Mentioned: 215 Post(s)
Tagged: 0 Thread(s)
Quoted: 1035 Post(s)
Liked: 1611
Quote:
Originally Posted by madshi View Post
Can you make this available for both FALL and BT2390 algos? I'm asking because it seems some users prefer FALL and some BT2390 algos.
So here it is. With the "Sky/flat bright area" detection algo.

New version of the madmeasuredynamicclipping / optimizer Tool V3.8.7

Download:
http://projectiondream.com/download/...namicclipping/



And the new Option(s) under "tweaked FALL".


For further discussion and feedback, as always, please report in this thread:
https://www.avsforum.com/forum/26-ho...l#post58035720

@Fer15 and @Neo-XP please provide feedback when you get time so that it can get confirmed or improved to get included in the LIVE algo.


Thank you and enjoy,
Flo
AmigoHD, Manni01, neo_2009 and 4 others like this.
Soulnight is offline  
post #5866 of 6789 Old 05-12-2019, 09:15 AM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,739
Mentioned: 526 Post(s)
Tagged: 0 Thread(s)
Quoted: 2605 Post(s)
Liked: 3476
Oh man, got the results from Fer15 and they're completely different to Neo-XP's results. Basically for Neo-XP the "temporal motion smooth + 0.1 * prvIgnPix" algo produces the best results by a noticeable margin. But for Fer15, this algo shares the last place. Now I'm screwed...

One thing I don't understand, though: E.g. for the "temporal motion smooth + 0.1 * prvIgnPix" algo, Neo-XP's sheet suggests an optimal threshold of 11.551. However, if I apply this threshold to Fer15's sheet, I get *significantly* worse results (20 overall misses compared to 14). @Neo-XP , how would your sheet work with a threshold for 9.5 for "temporal motion smooth + 0.1 * prvIgnPix"?

Could both of you share your sheets here? Then we could all look at both sheets and maybe try to draw some conclusions. Maybe you could even double check a few random measurements from each other's sheets, to check if you get similar results? I'm really at a loss to explain this wide gap in which algo works best for you guys. If all else fails, I guess we could merge both sheets and then check which algo does best that way?

Or any other ideas how to get to the bottom of this?

Sorry guys, I had hoped this would be more straightforward...
stevenjw and Manni01 like this.
madshi is offline  
post #5867 of 6789 Old 05-12-2019, 09:26 AM
Senior Member
 
Join Date: Jan 2018
Posts: 363
Mentioned: 70 Post(s)
Tagged: 0 Thread(s)
Quoted: 200 Post(s)
Liked: 240
Here are my results:


https://mega.nz/#!QyIgiCLI!BllQ2PJNGxCIDGJAkf_4FLYv7rdqZQKCNQdmH-6ea10




(PS I have "metric 1 would have detected where m2 missed" entered (green shade), but haven't looked at this yet, so that's the reason there's an absence of green atm, not that there was no difference)
madshi and Manni01 like this.
Fer15 is offline  
post #5868 of 6789 Old 05-12-2019, 10:38 AM - Thread Starter
AVS Forum Special Member
 
Soulnight's Avatar
 
Join Date: Dec 2011
Location: Germany
Posts: 1,402
Mentioned: 215 Post(s)
Tagged: 0 Thread(s)
Quoted: 1035 Post(s)
Liked: 1611
Quote:
Originally Posted by Soulnight View Post
Maybe this can help us differentiate what a scene is compared to what a shot is:

https://www.research.ibm.com/haifa/projects/imt/video/

https://www.research.ibm.com/haifa/p..._DataSet.shtml

https://www.ibm.com/blogs/research/2...ene-detection/

So madVR live algo current "scene detection" is actually a "shot detection " algo instead.

And what I called "chapter" as a collection of scene which belongs together should actually be called "scenes" instead.

And there is quite a lot of litterature online for both.

But shot detection seems much easier than real scene detection. And I believe real scene detection can only be performed before movie start when analyzing the full movie to regroup shots with the same semantics together.
In the 2nd link, there are a few movie already pre-analysed with a text files giving all the correct time code for the "shot" change.

"Open Video Scene Detection Dataset"

Maybe we can/should use that to see how madVR scene detection performs against this known benchmark?

Maybe we could even train an algo based on those?
Manni01 and omarank like this.
Soulnight is offline  
post #5869 of 6789 Old 05-12-2019, 05:41 PM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 993
Mentioned: 151 Post(s)
Tagged: 0 Thread(s)
Quoted: 700 Post(s)
Liked: 875
@madshi

The doc: http://www.mediafire.com/file/nv4tyw..._V10.xlsx/file

I will double check Fer15's doc measurements and add them to mine (for the titles I own), to see if it changes something.
madshi and Manni01 like this.

Last edited by Neo-XP; 05-12-2019 at 05:45 PM.
Neo-XP is offline  
post #5870 of 6789 Old 05-12-2019, 11:54 PM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,739
Mentioned: 526 Post(s)
Tagged: 0 Thread(s)
Quoted: 2605 Post(s)
Liked: 3476
Quote:
Originally Posted by Soulnight View Post
In the 2nd link, there are a few movie already pre-analysed with a text files giving all the correct time code for the "shot" change.

"Open Video Scene Detection Dataset"

Maybe we can/should use that to see how madVR scene detection performs against this known benchmark?

Maybe we could even train an algo based on those?
They don't seem to have a list of how other algos fare with these videos, or did I miss that? What would it help to check these videos for how madVR's algo handles these, if we have no comparison to other algos? We won't be able to judge whether our algo is better or worse compared to others.

Quote:
Originally Posted by Neo-XP View Post
@madshi

The doc: http://www.mediafire.com/file/nv4tyw..._V10.xlsx/file

I will double check Fer15's doc measurements and add them to mine (for the titles I own), to see if it changes something.
Awesome, many thanks!
madshi is offline  
post #5871 of 6789 Old 05-13-2019, 12:12 AM - Thread Starter
AVS Forum Special Member
 
Soulnight's Avatar
 
Join Date: Dec 2011
Location: Germany
Posts: 1,402
Mentioned: 215 Post(s)
Tagged: 0 Thread(s)
Quoted: 1035 Post(s)
Liked: 1611
Quote:
Originally Posted by madshi View Post
They don't seem to have a list of how other algos fare with these videos, or did I miss that? What would it help to check these videos for how madVR's algo handles these, if we have no comparison to other algos? We won't be able to judge whether our algo is better or worse compared to others.
Well, the goal is to compare to yourself as you improve the algo.

And we could use an automatic optimization process to see which settings or combination of shot detection algo wields the better results/match to the set of data.
And this with a much larger database than what neo-xp and fer15 are working on manually.
Soulnight is offline  
post #5872 of 6789 Old 05-13-2019, 12:18 AM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,739
Mentioned: 526 Post(s)
Tagged: 0 Thread(s)
Quoted: 2605 Post(s)
Liked: 3476
Automatic testing has been suggested before, in this thread. It would be too much work to make it worth it at this stage. I've also already said that I might look into neural network logic in the future to improve scene detection. But it's not the right time to do that now, either. I just want Metric2 to be optimized reasonably and then use Metric1+2 as a reasonably well working solution for the time being (to be improved in the future). I can't justify spending all my development resources on just shot/scene detection for many more weeks to come, there are just too many other areas that need work. So work on shot detection will soon be concluded (for now).
madshi is offline  
post #5873 of 6789 Old 05-13-2019, 12:40 AM - Thread Starter
AVS Forum Special Member
 
Soulnight's Avatar
 
Join Date: Dec 2011
Location: Germany
Posts: 1,402
Mentioned: 215 Post(s)
Tagged: 0 Thread(s)
Quoted: 1035 Post(s)
Liked: 1611
Quote:
Originally Posted by madshi View Post
Automatic testing has been suggested before, in this thread.
I know. But at the time we were missing an available large set of data to benchmark against.

Quote:
Originally Posted by madshi View Post
So work on shot detection will soon be concluded (for now).
All for the better. It was good enough already in the past. And resetting at each shot change is not always wanted anyway.

Let's move to something more interesting.
Manni01 likes this.
Soulnight is offline  
post #5874 of 6789 Old 05-13-2019, 02:23 AM
Home Theater Enthusiast
 
SOWK's Avatar
 
Join Date: Dec 2004
Location: WI
Posts: 5,388
Mentioned: 138 Post(s)
Tagged: 0 Thread(s)
Quoted: 1508 Post(s)
Liked: 645
Quote:
Originally Posted by madshi View Post
Oh man, got the results from Fer15 and they're completely different to Neo-XP's results. Basically for Neo-XP the "temporal motion smooth + 0.1 * prvIgnPix" algo produces the best results by a noticeable margin. But for Fer15, this algo shares the last place. Now I'm screwed...



I am not sure it matters, but Neo-XP's best mode is temporal motion smooth + 1.0 * prvIgnPix, and not temporal motion smooth + 0.1 * prvIgnPix


Dont know if that changes what you were discussing with Fer15 or not but just wanted to make sure we are all testing the same thing for the most accurate results.

---------------------------------------------------------------------
Wisconsin Home Theater Enthusiasts Meet
SOWK Home Theater
SOWK is offline  
post #5875 of 6789 Old 05-13-2019, 02:28 AM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,739
Mentioned: 526 Post(s)
Tagged: 0 Thread(s)
Quoted: 2605 Post(s)
Liked: 3476
Yes, I know. I was looking at "0.1 * prvIgnPix" because that was a mode which worked reasonably well for both Neo-XP and Fer15.
madshi is offline  
post #5876 of 6789 Old 05-13-2019, 09:57 AM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 993
Mentioned: 151 Post(s)
Tagged: 0 Thread(s)
Quoted: 700 Post(s)
Liked: 875
@madshi I don't have the exact same value as Fer15, and for all algos (including Metric1).


Is this normal? They are not that far, but maybe far enough to produce a difference.


Was the sub of prev frame disabled in Fer15 tests?
Neo-XP is offline  
post #5877 of 6789 Old 05-13-2019, 10:06 AM
Senior Member
 
Join Date: Jan 2018
Posts: 363
Mentioned: 70 Post(s)
Tagged: 0 Thread(s)
Quoted: 200 Post(s)
Liked: 240
Quote:
Originally Posted by Neo-XP View Post
@madshi I don't have the exact same value as Fer15, and for all algos (including Metric1).


Is this normal? They are not that far, but maybe far enough to produce a difference.


Was the sub of prev frame disabled in Fer15 tests?




Yes, "disable sub of prev frame" was checked in my tests


edit:


BTW, did you include any of my scenes that you hadn't tested previously into your tests, and if so, did it affect the outcome?
Fer15 is offline  
post #5878 of 6789 Old 05-13-2019, 10:08 AM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,739
Mentioned: 526 Post(s)
Tagged: 0 Thread(s)
Quoted: 2605 Post(s)
Liked: 3476
I assume that the differences could be due to:

1) Decoder (very unlikely, unlike one of you is using native DXVA decoding).
2) Playback resolution.
3) Output gamut.

This is just a guess, though. Which playback resolutions and output gamut have you been using in your tests?
madshi is offline  
post #5879 of 6789 Old 05-13-2019, 10:13 AM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 993
Mentioned: 151 Post(s)
Tagged: 0 Thread(s)
Quoted: 700 Post(s)
Liked: 875
@Fer15 Hum... then there is a problem.

Just for the first real scene I tested, this is what I got :

Code:
Fer15: Atomic Blonde 6124 -> 28,58 10,88 11,79 16,22 10,88 10,88 11,79 16,22
Me:    Atomic Blonde 6124 -> 28,61 12,14 12,72 17,91 12,14 12,14 12,72 17,91
Since your results were lower, I though you forgot to disable the sub, but no.
Neo-XP is offline  
post #5880 of 6789 Old 05-13-2019, 10:22 AM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,739
Mentioned: 526 Post(s)
Tagged: 0 Thread(s)
Quoted: 2605 Post(s)
Liked: 3476
Well, at least we're on the way to maybe find out why your results are diverging, which is good news. Can both of you report which playback resolution and output gamut you've been testing with? Then maybe we can find out which of these makes the difference here (if any). Or maybe there's yet another reason that's not clear to me right now.
madshi is offline  
Sponsored Links
Advertisement
 
Reply Digital Hi-End Projectors - $3,000+ USD MSRP

Tags
dynamic tone mapping , hdr , madvr , sdr , ton mapping

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off