Improving Madvr HDR to SDR mapping for projector - Page 166 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 7314Likes
Reply
 
Thread Tools
post #4951 of 7899 Old 02-07-2019, 12:40 PM
AVS Forum Special Member
 
RapalloAV's Avatar
 
Join Date: Jan 2008
Location: Auckland, New Zealand
Posts: 4,354
Mentioned: 29 Post(s)
Tagged: 0 Thread(s)
Quoted: 2228 Post(s)
Liked: 773
1st post…
Madvr HDR to SDR mapping: already great, soon even better for projector?
Madvr HDR to SDR Mapping with special focus on projector.
It makes every projector or TV compatible with HDR even if they were not initially "marketed" for that.

Im sorry if I seem a bit dumb on all this as I’m very new to HDR. There are so many posts in this thread I don’t know where to start
I have been using madvr for years, currently with version v0.92.17 and the 1080Ti card.
I have owned JVC projectors for years but have always found trying to use HDR was too daunting!
Since I purchased the new JVC NX9 projector I have been loving HDR films using their auto tone mapping, it works very well.
But Im told using madvr would be better….

My questions are:
1. Would madvr be better than my NX9 for HDR films.
2. Would madvr be just as easy and automatic as the NX9 is for HDR films?

I see the first post on this page isn’t really a sticky of “how to do it” as it looks like the madvr versions have changed many times since that post on the 02/03/2018.
So if my questions 1 and 2 are YES, where do I really start here…. Is there a post using the latest madvr version (without bugs) of how to setup HDR for a projector, that once set it will select what’s needed automatically for each new film in an easy way.

Sorry to ask these dumb questions but I need some help.
Many thanks in advance..

Murray Thompson
X Owner RapalloAV
Absolutly no connection with RapalloNZ
CinemascopE Home Cinema Build & 2014 rebuild, plus new LED ceiling install Christmas 2018
RapalloAV is offline  
Sponsored Links
Advertisement
 
post #4952 of 7899 Old 02-07-2019, 12:45 PM
Member
 
Join Date: Sep 2018
Posts: 52
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 28 Post(s)
Liked: 14
Quote:
Originally Posted by madshi View Post
That's your fault if you delete that folder behind madVR's back... You'll find that most/many applications don't handle it well if you tell them to use a specific folder and then delete that folder. Most of them won't recreate that folder, but instead either simply fail to work, or alternatively switch back to their default path instead.
I know it's my fault.
i made a script to use the measurements file for iso file (mount in a virtual drive, play the right mpls with mpc-be to know the right mpls file and made the madmesure. After that, I save the measurement file in a save folder with a name who contains the iso and the mpls name.
When played , I copy the file with a format madvr wants to use it.
As it's a copy, I use my ramdrive and if it's a folder, each time I reboot, folders are delete.
It's not a great deal as i use the root of my ramdrive to store the file but i only would like to inform you of the problem
xxxx5 is online now  
post #4953 of 7899 Old 02-07-2019, 12:53 PM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,850
Mentioned: 569 Post(s)
Tagged: 0 Thread(s)
Quoted: 2744 Post(s)
Liked: 3844
@RapalloAV , this is not really a support thread. It's a thread dedicated for improving madVR's HDR algorithms. So for support questions, please use one of the other threads on AVSForum or doom9.

But to answer quickly: YES, configured correctly, madVR should produce significantly better results than NX9's "auto" tone mapping. NX9 does choose a decent tone mapping curve, *if* the metadata is correct, but it's still a completely static curve which doesn't adjust to different movie scenes at all, plus it depends on correct metadata. madVR is fully dynamic (as good as HDR10+, or even better), and doesn't need any metadata to work well. Is it easy to use? If you already have experience with madVR, then yes. You do need a decent GPU, of course. And then just download the latest test build (see just a few posts above your's), switch HDR settings to pixel shader tone mapping, choose an appropriate "target peak nits" setting, and make sure you clear the HDR related "trade quality" options. Done. Further quality improvements available by using measurement files, but it's getting complicated there, so maybe just ignore that. Ah yes, also make sure you switch LAV Video Decoder to hardware decoding (ideally D3D11 native or copyback), otherwise it will probably be a slide show.
madshi is online now  
Sponsored Links
Advertisement
 
post #4954 of 7899 Old 02-07-2019, 12:59 PM
AVS Forum Special Member
 
RapalloAV's Avatar
 
Join Date: Jan 2008
Location: Auckland, New Zealand
Posts: 4,354
Mentioned: 29 Post(s)
Tagged: 0 Thread(s)
Quoted: 2228 Post(s)
Liked: 773
Quote:
Originally Posted by madshi View Post
@RapalloAV , this is not really a support thread. It's a thread dedicated for improving madVR's HDR algorithms. So for support questions, please use one of the other threads on AVSForum or doom9.

But to answer quickly: YES, configured correctly, madVR should produce significantly better results than NX9's "auto" tone mapping. NX9 does choose a decent tone mapping curve, *if* the metadata is correct, but it's still a completely static curve which doesn't adjust to different movie scenes at all, plus it depends on correct metadata. madVR is fully dynamic (as good as HDR10+, or even better), and doesn't need any metadata to work well. Is it easy to use? If you already have experience with madVR, then yes. You do need a decent GPU, of course. And then just download the latest test build (see just a few posts above your's), switch HDR settings to pixel shader tone mapping, choose an appropriate "target peak nits" setting, and make sure you clear the HDR related "trade quality" options. Done. Further quality improvements available by using measurement files, but it's getting complicated there, so maybe just ignore that. Ah yes, also make sure you switch LAV Video Decoder to hardware decoding (ideally D3D11 native or copyback), otherwise it will probably be a slide show.
Many thanks for that madshi, since this isnt a suport thread for HDR users more of a dedicated developers thread, can you point me to the one on avsforums that would be closer to my needs please?
However i will try to start with your instructions until I find that other thread more suited to my needs.

Murray Thompson
X Owner RapalloAV
Absolutly no connection with RapalloNZ
CinemascopE Home Cinema Build & 2014 rebuild, plus new LED ceiling install Christmas 2018
RapalloAV is offline  
post #4955 of 7899 Old 02-08-2019, 02:38 AM
Advanced Member
 
chros73's Avatar
 
Join Date: Jan 2015
Posts: 583
Mentioned: 12 Post(s)
Tagged: 0 Thread(s)
Quoted: 360 Post(s)
Liked: 171
Quote:
Originally Posted by madshi View Post
I'm not good at working on 20 different things at the same time. Maybe I can look into this in the future, but not right now.
Me neither , no worries.

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 online now  
post #4956 of 7899 Old 02-08-2019, 06:42 AM
Member
 
Join Date: Nov 2016
Posts: 122
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 92 Post(s)
Liked: 56
bdmv/iso structure

@pandm1967 I am trying to get bdmv folder stucture to work. I am not that worried about the menus but I would like to avoid to convert everything in mkv. I tried your tool and it identifies the correct m2ts (biggest one) just fine and creates the measurement file. Since my movies are on a nas I can't use the hard link. So running mpc-be and starting the m2ts directly works but it dows not work starting it like Kodi in combination with mpc-be does.Is there any other way to force madvr/mpc-be to play the m2ts file which has the measurement file attached instead of the index.bdmv autmatically?
The only option I can think of atm is deleting all other files beside the main movie --> renaming it as moviename.m2ts --> renaming the measurement file accordingly --> run Soulnights optimizing --> then Kodi in combination with mps-be should run the movie (without menus and extras) with measurements correct?
Also could you add support for iso files containing a bdmv structure?
@soul nigh&Anna I know it was discussed in the other thread already but I would also like to ask for a optional m2ts/iso support as well since if I go the road I described above I could do the renaming before and just stay in your tool

@madshi really hoping for your magic to implement the Soulnight optimizations in the live algo which would solve all problems above as well =)

NoTechi
stevenjw likes this.
NoTechi is online now  
post #4957 of 7899 Old 02-08-2019, 07:43 AM
AVS Forum Special Member
 
Onkyoman's Avatar
 
Join Date: Aug 2007
Location: Canada
Posts: 1,917
Mentioned: 52 Post(s)
Tagged: 0 Thread(s)
Quoted: 1123 Post(s)
Liked: 661
Quote:
Originally Posted by RapalloAV View Post
Many thanks for that madshi, since this isnt a suport thread for HDR users more of a dedicated developers thread, can you point me to the one on avsforums that would be closer to my needs please?
However i will try to start with your instructions until I find that other thread more suited to my needs.
You could talk to me at the PC build thread in my signature. Start by looking at the options in the latest test build. If you want more information, I can probably answer your questions.

If want more information about HDR, try the madVR guide (how to set up), specifically start with some of the external articles linked at the end of the HDR section.
Onkyoman is online now  
post #4958 of 7899 Old 02-08-2019, 03:25 PM
Member
 
Join Date: Apr 2015
Posts: 96
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 82 Post(s)
Liked: 5
Quote:
Originally Posted by FidelioX View Post
Thank you so much guys!
I use madvr with my Sony projector.
I calibrate the hdr films using a 3D lut (created in LS) targeting DCI p3 with power gamma 2.2.

In your opinion, for this two titles, Do I've to create another lut targeting to Rec. 2020 power gamma 2.2?.
Hi guys, any feedback about this?

Thank you so much
FidelioX is offline  
post #4959 of 7899 Old 02-08-2019, 03:51 PM
Member
 
BerndFfm's Avatar
 
Join Date: Jan 2018
Location: Frankfurt / Main, Germany
Posts: 99
Mentioned: 19 Post(s)
Tagged: 0 Thread(s)
Quoted: 94 Post(s)
Liked: 104
I installed madVR and measurement tool for a home cinema dealer here, I tested it with the JVC N-9. Great combination !

I made some testings of the performance of the measurement tool :

Measuring chapter 1 of Guardians of the Galaxy Vol. 2

Hardware Decoder : D3D11 Hardware device : Automatic (Native)
SSD : time used 00:00:49.
HDD : time used 00:00:48.
USB 2.0 : time used 00:01:09.
USB 3.0 : time used 00:00:48.

Hardware Decoder : DXVA1 (copy-back) Hardware device : Automatic
SSD : time used 00:01:30.
HDD : time used 00:01:30.
USB 2.0 : time used 00:01:29.
USB 3.0 : time used 00:01:29.

Hardware Decoder : DXVA2 (native)
SSD : time used 00:01:28.
HDD : time used 00:01:27.
USB 2.0 : time used 00:01:30.
USB 3.0 : time used 00:01:29.

Hardware Decoder : NVIDIA CUVID
SSD : time used 00:01:30.
HDD : time used 00:01:29.
USB 2.0 : time used 00:01:31.
USB 3.0 : time used 00:01:30.

Bernd
mavang and chros73 like this.

Home Cinema with : Onkyo 1009, Nubert, JVC RS-500, madVR
BerndFfm is offline  
post #4960 of 7899 Old 02-09-2019, 06:11 AM
AVS Forum Special Member
 
Onkyoman's Avatar
 
Join Date: Aug 2007
Location: Canada
Posts: 1,917
Mentioned: 52 Post(s)
Tagged: 0 Thread(s)
Quoted: 1123 Post(s)
Liked: 661
Quote:
Originally Posted by FidelioX View Post
Hi guys, any feedback about this?

Thank you so much
If you already have a DCI-P3 3D LUT, there would be no point in creating a BT.2020 LUT as DCI-P3 should produce a superior result.
Onkyoman is online now  
post #4961 of 7899 Old 02-09-2019, 07:23 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,306
Mentioned: 355 Post(s)
Tagged: 0 Thread(s)
Quoted: 5608 Post(s)
Liked: 5910
Quote:
Originally Posted by Onkyoman View Post
If you already have a DCI-P3 3D LUT, there would be no point in creating a BT.2020 LUT as DCI-P3 should produce a superior result.
It depends.

If the native gamut covers more than 100% of P3 and if there is content over 100% of P3 in those BT2020 titles, it might be worth it to have a BT2020 LUT for these titles, provided the calibration software doesn't cause posterization with a wider gamut or that extending the gamut cover doesn't cost too much brightness.

For example, on the JVC rs4500/Z1, you can cover significantly more than P3, but the cost is 40% of brightness when the filter is enabled. So most people don't use it.

The new native 4K JVCs are supposed to cover up to 108% of P3 with the filter (TBC). Whether that justifies having a separate BT2020 calibration depends on both the 3D LUT, brightness and content.

What's for sure if that if the native gamut of the display or projector doesn't cover more than P3, then there is no advantage having a BT2020 LUT, mostly downsides.

At this stage, I wouldn't worry about BT2020 because we still haven't been able to confirm that there was content beyond P3 even on these titles. It's the same as titles mastered to 4000nits that have content not going above a few hundred nits (Blade Runner 2049). Ther could be content, or there could be nothing beyond P3.

I hope that when Madshi has the time, he'll put in some kind of histogram showing how far the content goes and when it actually shoots beyond P3. It would be very useful.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
JVC Macro feature on Vertex/Vertex2/Integral2/Maestro
Manni01 is online now  
post #4962 of 7899 Old 02-09-2019, 09:38 AM
Member
 
Join Date: Apr 2018
Posts: 190
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 127 Post(s)
Liked: 17
I just noticed my gpu is spiking to %100 while playing movies. This is on the newest beta build

edit: seems to be back to normal

Last edited by toby5; 02-09-2019 at 10:39 AM.
toby5 is offline  
post #4963 of 7899 Old 02-09-2019, 02:37 PM
AVS Forum Special Member
 
RapalloAV's Avatar
 
Join Date: Jan 2008
Location: Auckland, New Zealand
Posts: 4,354
Mentioned: 29 Post(s)
Tagged: 0 Thread(s)
Quoted: 2228 Post(s)
Liked: 773
Quote:
Originally Posted by Onkyoman View Post
You could talk to me at the PC build thread in my signature. Start by looking at the options in the latest test build. If you want more information, I can probably answer your questions.

If want more information about HDR, try the madVR guide (how to set up), specifically start with some of the external articles linked at the end of the HDR section.
Onkyoman many thanks, I have posted to the other thread for your advise please. https://www.avsforum.com/forum/26-ho...l#post57576780

Murray Thompson
X Owner RapalloAV
Absolutly no connection with RapalloNZ
CinemascopE Home Cinema Build & 2014 rebuild, plus new LED ceiling install Christmas 2018
RapalloAV is offline  
post #4964 of 7899 Old 02-09-2019, 04:19 PM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,850
Mentioned: 569 Post(s)
Tagged: 0 Thread(s)
Quoted: 2744 Post(s)
Liked: 3844
So here's a big new test build:

http://madshi.net/madVRhdrMeasure41.zip

I've implemented the very same algorithms that Anna & Flo implemented in their "optimization tool" for measurement files - but I've implemented them for the "live algo" (meaning when no measurement file is available). Due to not having a measurement file available for the "live algo", of course I couldn't implement the Anna & Flo "chapter" algorithm, so I implemented that part of the algorithm differently.

All the same settings/options that Anna & Flo offer are also available in this test build (except the "chapter" related settings). Please note that this build is *not* intended for fine tuning the Anna & Flo algorithms. Of course you can do that, if you like, but it's not really what I'm interested in hearing from you in this thread at this point. The main purpose of this build is to check if all the "dynamic" improvements introduced by Anna & Flo's tool can also be made to work with the "live algo".

What I'd like you to do:

1) Use the same settings you're currently using with the Anna & Flo tool. Or if you've not used that tool yet, just stick to the default settings.

2) Please start by unchecking the "detect scene changes" option and fine tune the "max brightening/darkening per second" options. What you want to achieve is a fast but completely invisible adjustment to dark vs bright scenes. Of course the goals "fast" and "completely invisible" contract each other, so please try to find the sweet spot. The sweet spot is probably the setting which is *just* slow enough to be invisible.

3) Once you've selected brightening/darkening settings that are invisible to you (but still reasonably fast), please enable scene change detection and try to find a scene change threshold which doesn't produce noticeable artifacts (e.g. flickering), but which still detects most scene changes properly. Please note that it's perfectly ok if some scene changes are "missed" by madVR, as long as the miss doesn't come with noticeable image quality deficits. So you should probably error on the side of missing some scene changes, instead of producing false positive scene change detections.

4) And of course I'd like to have your opinion on whether my current "dynamic live algo" implementation with scene change and brightening/darkening settings already works well enough, or if we need to find some more improvements to make it useable.

Of course if you find problematic scene changes which this test build struggles with, please let me know. Please list the movie/demo name and time code. Ideally a small sample would be nice, but I'll be happy about reports without samples, too, of course.
madshi is online now  
post #4965 of 7899 Old 02-09-2019, 04:25 PM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,306
Mentioned: 355 Post(s)
Tagged: 0 Thread(s)
Quoted: 5608 Post(s)
Liked: 5910
Looks like you've used a significant amount of magic dust

Sounds very promising, looking forward to testing, will report back!
mkohman likes this.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
JVC Macro feature on Vertex/Vertex2/Integral2/Maestro
Manni01 is online now  
post #4966 of 7899 Old 02-09-2019, 04:31 PM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,850
Mentioned: 569 Post(s)
Tagged: 0 Thread(s)
Quoted: 2744 Post(s)
Liked: 3844
I hope I used enough. I'm all out of stock right now, but Tinkerbell is due for another visit soon.

To be honest, I've not really a good idea right now if it works well or not. In my very *very* limited tests it appeared to work decently, but I probably haven't tested with really difficult scenes yet. So I'm really curious about (all) your test results.
madshi is online now  
post #4967 of 7899 Old 02-09-2019, 07:10 PM
Member
 
Join Date: Apr 2018
Posts: 190
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 127 Post(s)
Liked: 17
First of all I want to say thank you so much for the new build. I have been looking over a few things and here a few of my first assessments. Firstly, the new build doesn't give the same amount of contrast as the optimizer tool at the same settings. Secondly, I went through the real peak nits and it didn't change no matter the number I put in.
toby5 is offline  
post #4968 of 7899 Old 02-09-2019, 07:20 PM
AVS Forum Special Member
 
Javs's Avatar
 
Join Date: Dec 2012
Location: Sydney
Posts: 8,056
Mentioned: 497 Post(s)
Tagged: 0 Thread(s)
Quoted: 6933 Post(s)
Liked: 6640
Quote:
Originally Posted by madshi View Post
So here's a big new test build:
I've implemented the very same algorithms that Anna & Flo implemented in their "optimization tool" for measurement files - but I've implemented them for the "live algo" (meaning when no measurement file is available).
Dude... Exceptional work. Cant wait to test.


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 offline  
post #4969 of 7899 Old 02-09-2019, 07:33 PM
Member
 
Join Date: Nov 2018
Posts: 173
Mentioned: 45 Post(s)
Tagged: 0 Thread(s)
Quoted: 111 Post(s)
Liked: 98
@madshi

Using these settings:

min target: 250
maximal target: 2000
no comp limit: 250
dynamic tuning: 150
scene change threshold: 30
max brightening/s: 20
max darkening/s: 20

The default settings for brightening and darkening work really well from what I've tested, I haven't found anything that looked wrong to me.

The scene detection however doesn't work in highly dynamic scenes. If you look at the sandstorm scene in Fury Road you can see lots of problems. In particular at 00:27:23 when Max is on the back of the car the targets will jump from a very low value to a very high value when there is a lightening flash, the overall effect of this is that the lightening becomes only a change in colour and not a change in brightness (with a low minimum/real value at least, with 250 it doesn't look quite as bad). IMO this whole sequence (with max on the back of the car before it cuts to inside the car) needs to be treated as a single scene as far as dynamic targets is concerned using an average target so that the dark parts look dark and the flashes look bright. I understand that in the live algo you can't use soulnight's chapter stuff but if it's helpful at all these are the settings I used to get this scene working:

Scene merge: 350
minimum chapter duration: 24
chapter merge: 200

minimum chapter duration in particular is important for this scene as I tested from 0 to 48 in 6 frame increments and found that 24 was the lowest value to work. The lightening flashes in this scene are the main problem as they are a massive change in frame brightness and any setting that is used for the scene detection threshold, if it could stop these flashes from triggering a scene change, would break any other legitimate scene change.

Another example in this scene is at 00:27:48 when they are driving towards the vortex. After every lightening flash the new target makes a noticeable change in the brightness. There are many other problem scene changes in this sequence but those are two of the more obvious ones
Manni01 likes this.
Dexter Kane is offline  
post #4970 of 7899 Old 02-09-2019, 07:46 PM
Member
 
Join Date: Mar 2018
Posts: 54
Mentioned: 19 Post(s)
Tagged: 0 Thread(s)
Quoted: 45 Post(s)
Liked: 67
Quote:
Originally Posted by NoTechi View Post
@pandm1967 I am trying to get bdmv folder stucture to work. I am not that worried about the menus but I would like to avoid to convert everything in mkv. I tried your tool and it identifies the correct m2ts (biggest one) just fine and creates the measurement file. Since my movies are on a nas I can't use the hard link. So running mpc-be and starting the m2ts directly works but it dows not work starting it like Kodi in combination with mpc-be does.Is there any other way to force madvr/mpc-be to play the m2ts file which has the measurement file attached instead of the index.bdmv autmatically?
The only option I can think of atm is deleting all other files beside the main movie --> renaming it as moviename.m2ts --> renaming the measurement file accordingly --> run Soulnights optimizing --> then Kodi in combination with mps-be should run the movie (without menus and extras) with measurements correct?
Also could you add support for iso files containing a bdmv structure?
Although I don't use Kodi in combination with mpc-be, I don't think you should delete all files beside the main movie of UHD folder, maybe you can try this : when mpc-be starts playing, press Shift+f10, find out the main media file, e.g 00800.mpls, then copy the existing measurement file in stream folder to playlist\00800.mpls.measurements.

You can read the further explanation here : https://www.avsforum.com/forum/26-ho...l#post57574174

For adding support for UHD ISO clones, it is not easy to add measurement file to ISO on the fly.
NoTechi likes this.
pandm1967 is offline  
post #4971 of 7899 Old 02-09-2019, 08:10 PM
AVS Forum Special Member
 
Javs's Avatar
 
Join Date: Dec 2012
Location: Sydney
Posts: 8,056
Mentioned: 497 Post(s)
Tagged: 0 Thread(s)
Quoted: 6933 Post(s)
Liked: 6640
Madshi.

Ive been testing Avengers Infinity War. Ive found from 25:30 to 28:00 minutes to be perfect for testing scene changes. Iron man walks into a space ship at 26min and there is a panning shot which goes from the min target (120 for my TV) to the high 300's and there is a lightning VFX effect there which is heavily affected by settings. Also if you keep watching there are a few more scenes right up to when you first see the Guardians, again, there is a low nits space pan to a super bright star.

Ive found myself wanting to figure a much faster adaptation to brightness for those, but then the Achilles heel seems to be a shot of Banner picking up a cell phone off the floor, tough to find a setting where that doesnt quickly pump brightness.

Can you confirm what the numerical values in max brightening per second mean? if I type 10, does it mean a max of 1% brightening per second?

Also, is Madvr still generating its own new measurement file when using the live algo as it goes if it doesn't find one? I have purposefully removed the reference to my measurements folder.

Here is an example of where I feel we want faster reactions to brightness. If I leave it on 20 for max brightening and darkening I get this:





if I raise it to 250 I get this by the time the panning shot is settled (goes from 120 nits to 424) and the highlights are there, and more HDR effect.



To others: Guys I think these few minutes in Infinity War are seriously torture tests for this stuff, it would be great if more of you would use these few minutes of the film to test... I think if we can get these scenes right, it will work for everything. There is very fast succession of peak nits changes in these scenes.

Settings:

stevenjw, Manni01 and Dexter Kane like 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 offline  
post #4972 of 7899 Old 02-09-2019, 08:26 PM
AVS Forum Special Member
 
Javs's Avatar
 
Join Date: Dec 2012
Location: Sydney
Posts: 8,056
Mentioned: 497 Post(s)
Tagged: 0 Thread(s)
Quoted: 6933 Post(s)
Liked: 6640
After playing around with the above scene more and skipping right through Avengers Infinity War and also watching the scene where Thor is at the starforge... These settings are looking pretty good.

Very minimal flicker. not super perfect, but the dynamic HDR is sure earning its wage!



Thanks so much Madshi, Flo and Anna.

This is pretty amazing for not using measurement files.
stevenjw, Manni01, LTD_666 and 1 others like 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 offline  
post #4973 of 7899 Old 02-09-2019, 08:56 PM
Member
 
Join Date: Nov 2018
Posts: 173
Mentioned: 45 Post(s)
Tagged: 0 Thread(s)
Quoted: 111 Post(s)
Liked: 98
Quote:
Originally Posted by toby5 View Post
First of all I want to say thank you so much for the new build. I have been looking over a few things and here a few of my first assessments. Firstly, the new build doesn't give the same amount of contrast as the optimizer tool at the same settings. Secondly, I went through the real peak nits and it didn't change no matter the number I put in.
I'm getting the opposite effect. It seems to me that either the dynamic clipping in the new test build is more aggressive than in soulnight's tool, or the clipping in the tool isn't working properly. Either way I think the madvr version looks better.
Dexter Kane is offline  
post #4974 of 7899 Old 02-09-2019, 09:06 PM
Member
 
Join Date: Nov 2018
Posts: 173
Mentioned: 45 Post(s)
Tagged: 0 Thread(s)
Quoted: 111 Post(s)
Liked: 98
Quote:
Originally Posted by Javs View Post
After playing around with the above scene more and skipping right through Avengers Infinity War and also watching the scene where Thor is at the starforge... These settings are looking pretty good.

Very minimal flicker. not super perfect, but the dynamic HDR is sure earning its wage!



Thanks so much Madshi, Flo and Anna.

This is pretty amazing for not using measurement files.
These settings seem to work pretty well. With scene detection turned off the sandstorm scene in fury road works with these settings. It's not very stable but it's not visually distracting.
Dexter Kane is offline  
post #4975 of 7899 Old 02-09-2019, 10:23 PM
AVS Forum Special Member
 
Javs's Avatar
 
Join Date: Dec 2012
Location: Sydney
Posts: 8,056
Mentioned: 497 Post(s)
Tagged: 0 Thread(s)
Quoted: 6933 Post(s)
Liked: 6640
Quote:
Originally Posted by Dexter Kane View Post
These settings seem to work pretty well. With scene detection turned off the sandstorm scene in fury road works with these settings. It's not very stable but it's not visually distracting.
Ive now settled on 25/125/30 for the live algo options.

lower max darkening seemed to tame more flickering and scene detection seemed to take care of the other darker shots that mattered anyway.

Really pretty happy here.
stevenjw, NoTechi and Dexter Kane like 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 offline  
post #4976 of 7899 Old 02-10-2019, 02:01 AM
Member
 
sat4all's Avatar
 
Join Date: Apr 2013
Posts: 105
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 74 Post(s)
Liked: 30
I guess the new algo default settings are optimized for 200 nits SDR output, Any suggestion for a starter settings if i'm using output in HDR format, my real display peak nit is 700 nits.
Sorry, never used the measurement tool before so i'm a little behind



Thanks Madshi, Flo and Anna.

LG OLED65C8 / Denon AVR-X3200W / KEF E305+ONKYO SKH-410 / Synology DS2415+ / Logitech Harmony 950
ZOTAC MAGNUS EN1060K: Win 10 x64 + Kodi DSPlayer x64

Last edited by sat4all; 02-10-2019 at 02:10 AM.
sat4all is online now  
post #4977 of 7899 Old 02-10-2019, 02:09 AM
Member
 
Join Date: Nov 2018
Posts: 173
Mentioned: 45 Post(s)
Tagged: 0 Thread(s)
Quoted: 111 Post(s)
Liked: 98
Quote:
Originally Posted by sat4all View Post
I guess the new algo default settings are optimized for 200 nits SDR output, Any suggestion for a starter settings if i'm using output in HDR format, my real display peak nit is 700 nits.
Sorry nevery used the measurement tool before so i'm a little behind



Thanks Madshi, Flo and Anna.
If you're outputting in HDR then you don't want to use dynamic targets
sat4all likes this.
Dexter Kane is offline  
post #4978 of 7899 Old 02-10-2019, 02:31 AM
Member
 
sat4all's Avatar
 
Join Date: Apr 2013
Posts: 105
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 74 Post(s)
Liked: 30
So i should avoid the latest build because i don't see any setting where to switch dynamic targets off?
If yes, @madshi please consider that in future builds.

Thanks

LG OLED65C8 / Denon AVR-X3200W / KEF E305+ONKYO SKH-410 / Synology DS2415+ / Logitech Harmony 950
ZOTAC MAGNUS EN1060K: Win 10 x64 + Kodi DSPlayer x64
sat4all is online now  
post #4979 of 7899 Old 02-10-2019, 05:47 AM
AVS Forum Special Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 1,136
Mentioned: 198 Post(s)
Tagged: 0 Thread(s)
Quoted: 859 Post(s)
Liked: 1184
Quote:
Originally Posted by madshi View Post
So here's a big new test build
This is amazing, I'm having a hard time seeing any difference between the "live" and stored measurements algo

The default settings are pretty good, but unfortunately there is still a difference for content with and without black bars to set the threshold (similar to the scene detection for the tone mapping target).

Two scenes I used for the threshold (most significant to me):

Pan, frame 93956 (01:05:19), real cut

- Lum changes (black bars included): 19.44%
- Lum changes (black bars excluded): 26.12%

The Meg, frame 1249 (00:00:52), false cut

- Lum changes (black bars included): 14.83%
- Lum changes (black bars excluded): 19.60%

Therefore, I got two answers for the threshold:

- With black bars included: 17.5%
- With black bars excluded: 22.5%

As for the max brightening/darkening per second, 50 is working fine for me on Infinity War panning shot at 26min.
I tried with 100, but there was some bad flickering on Passengers scene from frame 61249 (00:42:35).

I also looked at the sandstorm scene in Fury Road, but I don't see any problem there if I don't do a frame-to-frame analysis.

Edit: In addition to not having to measure titles anymore or having to try to manually find a static target (which didn't work very well anyway), I don't notice any performance drop

Last edited by Neo-XP; 02-10-2019 at 06:12 AM.
Neo-XP is offline  
post #4980 of 7899 Old 02-10-2019, 06:05 AM
Member
 
Join Date: Nov 2016
Posts: 122
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 92 Post(s)
Liked: 56
Might be a silly question but is there a setting to force the live algo so I don't have to remove/rename the measurement files for testing?


NoTechi
NoTechi is online now  
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