Improving Madvr HDR to SDR mapping for projector - Page 201 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 4778Likes
Reply
 
Thread Tools
post #6001 of 6063 Old 05-27-2019, 05:04 AM
Senior Member
 
Join Date: Jan 2018
Posts: 350
Mentioned: 67 Post(s)
Tagged: 0 Thread(s)
Quoted: 188 Post(s)
Liked: 219
Quote:
Originally Posted by Neo-XP View Post
They seem "easy" to detect, based on the histogram changes:



Yeah, I noticed that myself.


Also, from that film if you go to the scene when they all go trigger happy and fire non-stop into the jungle after Blain is killed, there's a muzzle fire flash inc. 12 consecutive frames, but thankfully they're not false positives which present a dilemma inc. lum chart changes.
Fer15 is offline  
Sponsored Links
Advertisement
 
post #6002 of 6063 Old 05-29-2019, 01:04 PM
Senior Member
 
Mr. Neverbicker's Avatar
 
Join Date: Feb 2011
Posts: 218
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 36 Post(s)
Liked: 25
I installed MadmeasureHDR Optimizer and madMeasureDynamicClipping and it works great, but the dynamic tune settings (outlined in red) are not present in the HDR settings of madVR. Is there any known issues I should be aware of?
Attached Thumbnails
Click image for larger version

Name:	madV-dynamic.png
Views:	152
Size:	97.2 KB
ID:	2573176  
Mr. Neverbicker is offline  
post #6003 of 6063 Old 05-29-2019, 01:34 PM
Advanced Member
 
ddgdl's Avatar
 
Join Date: Mar 2007
Posts: 608
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 329 Post(s)
Liked: 298
Quote:
Originally Posted by Mr. Neverbicker View Post
I installed MadmeasureHDR Optimizer and madMeasureDynamicClipping and it works great, but the dynamic tune settings (outlined in red) are not present in the HDR settings of madVR. Is there any known issues I should be aware of?
Those are for the live algorithm madshi is working on. If you use measurement files you don't need to worry about those
Mr. Neverbicker likes this.
ddgdl is offline  
Sponsored Links
Advertisement
 
post #6004 of 6063 Old 05-29-2019, 05:12 PM
Senior Member
 
Mr. Neverbicker's Avatar
 
Join Date: Feb 2011
Posts: 218
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 36 Post(s)
Liked: 25
Quote:
Originally Posted by ddgdl View Post
Those are for the live algorithm madshi is working on. If you use measurement files you don't need to worry about those
Thank you very much. I was suspecting that might be the case. I've followed the separate thread for the optimizer tool and I have impressive results with optimized files. I'm aware that tool is currently further developed than the live tool. However, to account for lamp dimming and changes in personal taste, I'm interested in the following and experimenting with the live algorithm because that method provides more flexibility in making tweaks without having to regenerate optimizations for hundreds of movies. Is the live tool available for download?
Mr. Neverbicker is offline  
post #6005 of 6063 Old 05-29-2019, 06:13 PM
Advanced Member
 
tommarra's Avatar
 
Join Date: Aug 2013
Location: USA
Posts: 569
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 340
Quote:
Originally Posted by Mr. Neverbicker View Post
Thank you very much. I was suspecting that might be the case. I've followed the separate thread for the optimizer tool and I have impressive results with optimized files. I'm aware that tool is currently further developed than the live tool. However, to account for lamp dimming and changes in personal taste, I'm interested in the following and experimenting with the live algorithm because that method provides more flexibility in making tweaks without having to regenerate optimizations for hundreds of movies. Is the live tool available for download?


Scroll up ... Madshi posts the test build. Neo-XP and Fer15 post their settings that you can use as a starting point.

Unless you want to on the bleeding edge, I have had perfectly good results with build 79 some settings for which have been posted by Neo-XP




Sent from my iPhone using Tapatalk
damelon and Mr. Neverbicker like this.

Gear: Fronts: B&W 803D3 | Center: B&W HTM2D3 | Surrounds: B&W 705S2 | Backs: B&W DM601S3 | Heights: Definitive Techonlogy ProCinema 1000 | Receiver & Amps: Anthem MRX 720 + Emotiva BasX-5 | Projector: JVC NX7 | Source: HTPC with Nvidia 2070, Kodi + External Player: MPC-HC with MadVR (for Blu-ray rips), Nvidia Shield (for streaming content)
tommarra is online now  
post #6006 of 6063 Old 05-30-2019, 12:17 AM
Senior Member
 
Mr. Neverbicker's Avatar
 
Join Date: Feb 2011
Posts: 218
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 36 Post(s)
Liked: 25
Okay, so I did just that. The result is well beyond my expectation. I can't believe what I'm seeing and how easy it is. I've done extensive experimentation with the optimizer tool and now giving one shot at the live algorithm completely blew me away. I figured at least unoptimized measurement files would be required, but from what I'm seeing, they're not. So I don't have to deal with the complications of measurement files with ISO files, and the versatility is a breath of fresh air.

I hope the contributors who plug away at this thread appreciate as much as I what they've accomplished. I've always been very particular down to every detail when it comes to HT, yet coming at this with a fresh eye, I'm way more than satisfied. Wow. Thank you all!
Mr. Neverbicker is offline  
post #6007 of 6063 Old 05-30-2019, 12:16 PM
Member
 
FGEvans's Avatar
 
Join Date: Feb 2007
Posts: 179
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 80 Post(s)
Liked: 15
Sorry for what probably will be a question that will be shot down. Is there a general setting guide I can use for my jvc x7500 projector. I have read some of the thread but started getting out of my depth. I currently just use madvr with the standard send hdr data but not meta data to my projector and then switch on hdr on my projector.

I don't know what general settings to use. My lamp is around 100 hours. Do I need a light meter for example to measure my brightness. I generally find madvr easy to use but this tone mapping is over my head at the moment.
FGEvans is offline  
post #6008 of 6063 Old 05-30-2019, 02:02 PM
Senior Member
 
Mr. Neverbicker's Avatar
 
Join Date: Feb 2011
Posts: 218
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 36 Post(s)
Liked: 25
Quote:
Originally Posted by FGEvans View Post
Sorry for what probably will be a question that will be shot down. Is there a general setting guide I can use for my jvc x7500 projector. I have read some of the thread but started getting out of my depth. I currently just use madvr with the standard send hdr data but not meta data to my projector and then switch on hdr on my projector.

I don't know what general settings to use. My lamp is around 100 hours. Do I need a light meter for example to measure my brightness. I generally find madvr easy to use but this tone mapping is over my head at the moment.
Use this to estimate your Foot-Lamberts..
https://www.projectorscreen.com/proj...en-calculators
(I measured mine and it came out the same as above result)

Convert to Nits...
http://www.kylesconverter.com/lumina...mberts-to-nits

In MadVR...
1) Set "min target / real display nits" to your nits from above.

2) Experiment with "maximal target nits" and "dynamic tuning" to suit your preference. I like to lower them both quite a bit but you'll see the effect they have and find what you like.

The other settings are pretty much for the developers.

Doing this should put a smile on your face.
tommarra likes this.
Mr. Neverbicker is offline  
post #6009 of 6063 Old 05-30-2019, 02:17 PM
AVS Forum Special Member
 
Onkyoman's Avatar
 
Join Date: Aug 2007
Location: Canada
Posts: 1,644
Mentioned: 42 Post(s)
Tagged: 0 Thread(s)
Quoted: 930 Post(s)
Liked: 575
Onkyoman is online now  
post #6010 of 6063 Old 05-30-2019, 03:39 PM
Member
 
FGEvans's Avatar
 
Join Date: Feb 2007
Posts: 179
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 80 Post(s)
Liked: 15
By using the calculator I seem to get 222 nits. That's based on my projectors 1900 lumens. That can't be right can it

I have a 100 inch diagonal screen

I note the options to put into madvr. What option do I then select on the jvc projector itself. I assume I don't pick the hdr mode
FGEvans is offline  
post #6011 of 6063 Old 05-30-2019, 04:58 PM
Advanced Member
 
ddgdl's Avatar
 
Join Date: Mar 2007
Posts: 608
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 329 Post(s)
Liked: 298
Quote:
Originally Posted by FGEvans View Post
By using the calculator I seem to get 222 nits. That's based on my projectors 1900 lumens. That can't be right can it

I have a 100 inch diagonal screen

I note the options to put into madvr. What option do I then select on the jvc projector itself. I assume I don't pick the hdr mode
This is the wrong thread for these kinds of questions. A good place to start for your exact situation is this thread, which is about madvr + the jvc n and rs series:
MadVR HTPC build for JVC NX7 projector - a build! https://www.avsforum.com/forum/26-home-theater-computers/3053112-madvr-htpc-build-jvc-nx7-projector-build.html
Danonano and alps006 like this.
ddgdl is offline  
post #6012 of 6063 Old 05-30-2019, 11:35 PM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,011
Mentioned: 307 Post(s)
Tagged: 0 Thread(s)
Quoted: 5360 Post(s)
Liked: 5412
@madshi
The BT2020 flag seems to be broken. BT2020 is on even for rec-709 content. Have you changed anything recently that might have caused this? I started noticing over the last couple of weeks ago.
I'm still on windows 1809 and it does it with the reliable 385.28 driver on my 1080ti, so I don't think it's the OS or GPU drivers, unless there was some kind of OS update for 1809 that broke this.

JVC Autocal Software V11 Calibration for 2019 Models (Google)
Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
Manni01 is offline  
post #6013 of 6063 Old 05-31-2019, 02:26 AM
Member
 
FGEvans's Avatar
 
Join Date: Feb 2007
Posts: 179
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 80 Post(s)
Liked: 15
Quote:
Originally Posted by ddgdl View Post
This is the wrong thread for these kinds of questions. A good place to start for your exact situation is this thread, which is about madvr + the jvc n and rs series:
MadVR HTPC build for JVC NX7 projector - a build! https://www.avsforum.com/forum/26-home-theater-computers/3053112-madvr-htpc-build-jvc-nx7-projector-build.html
cheers
FGEvans is offline  
post #6014 of 6063 Old 05-31-2019, 02:49 AM
Member
 
Join Date: Sep 2018
Posts: 47
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 27 Post(s)
Liked: 11
Quote:
Originally Posted by Manni01 View Post
@madshi
The BT2020 flag seems to be broken. BT2020 is on even for rec-709 content. Have you changed anything recently that might have caused this? I started noticing over the last couple of weeks ago.
I'm still on windows 1809 and it does it with the reliable 385.28 driver on my 1080ti, so I don't think it's the OS or GPU drivers, unless there was some kind of OS update for 1809 that broke this.

use a profile rule :

if (filePath="*BT2020*") or (hdr) "BT2020" else "BT709"

to "detect" bt2020
probably not the only way to do that but it works
xxxx5 is offline  
post #6015 of 6063 Old 05-31-2019, 03:11 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,011
Mentioned: 307 Post(s)
Tagged: 0 Thread(s)
Quoted: 5360 Post(s)
Liked: 5412
Quote:
Originally Posted by xxxx5 View Post
use a profile rule :

if (filePath="*BT2020*") or (hdr) "BT2020" else "BT709"

to "detect" bt2020
probably not the only way to do that but it works
Thanks but that's a workaround. I'm hoping for a fix. It used to work fine and I'm already using the (hdr) flag in my profile rules.

JVC Autocal Software V11 Calibration for 2019 Models (Google)
Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
Manni01 is offline  
post #6016 of 6063 Old 05-31-2019, 06:15 AM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 836
Mentioned: 134 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 687
@madshi

The results from the previous test are unfortunately wrong . 0 was entered for some cells instead of the formula, bad copy/paste from me I think. Sorry.

I'm adding new scenes because I got some issues while watching Backdraft, Metric2 is too high on some scenes while Metric1 is under the threshold, causing new false detections.

This can be improved by setting a Metric1 limit for a scene to be detected.

This is what I use now :


And this is what I propose:

Line1: There is definitely no scene change if metric1 is below [7.4] OR if metric2 is below [1.0]

Line2: Otherwise use weighted avrg. M1 thres: [11.6] - M2 thres: [3.7] - M1 weight %: 50
SOWK, carpantata and Manni01 like this.
Neo-XP is offline  
post #6017 of 6063 Old 05-31-2019, 06:24 AM
Home Theater Enthusiast
 
SOWK's Avatar
 
Join Date: Dec 2004
Location: WI
Posts: 5,817
Mentioned: 125 Post(s)
Tagged: 0 Thread(s)
Quoted: 1364 Post(s)
Liked: 2111
@Neo-XP , Thank you very much for your continued work. I appreciate your updates and I know many others that do as well.
RioBar4U and alps006 like this.

Wisconsin Home Theater Enthusiasts Meet: Main Thread l Gear For Sale
SOWK Home Theater
SOWK is online now  
post #6018 of 6063 Old 05-31-2019, 06:28 AM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,573
Mentioned: 457 Post(s)
Tagged: 0 Thread(s)
Quoted: 2337 Post(s)
Liked: 3038
Quote:
Originally Posted by Neo-XP View Post
@madshi

The results from the previous test are unfortunately wrong . 0 was entered for some cells instead of the formula, bad copy/paste from me I think. Sorry.

I'm adding new scenes because I got some issues while watching Backdraft, Metric2 is too high on some scenes while Metric1 is under the threshold, causing new false detections.

This can be improved by setting a Metric1 limit for a scene to be detected.

This is what I use now :


And this is what I propose:

Line1: There is definitely no scene change if metric1 is below [7.4] OR if metric2 is below [1.0]

Line2: Otherwise use weighted avrg. M1 thres: [11.6] - M2 thres: [3.7] - M1 weight %: 50
How's the number of misses now, when using "OR" vs "AND" in line 1? Can you share a new sheet?
madshi is offline  
post #6019 of 6063 Old 05-31-2019, 08:23 AM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 836
Mentioned: 134 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 687
@madshi

I removed all the false positives from "flash" scenes from before, because they were not making sense anymore to me: not detecting some of them as new scenes was actually worse for image quality, and the others were not making any difference.

Doc: https://www.mediafire.com/file/hef4bv5b0ddhtic/NeoV23.xlsx/file

One interesting thing: the excel solver tool is proposing a better solution to this problem by decreasing Metric1 and increasing Metric2 like this:

Manni01 likes this.

Last edited by Neo-XP; 05-31-2019 at 09:06 AM.
Neo-XP is offline  
post #6020 of 6063 Old 05-31-2019, 08:39 AM
Member
 
FGEvans's Avatar
 
Join Date: Feb 2007
Posts: 179
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 80 Post(s)
Liked: 15
when using the madvr tone mapping option do i still tell madvr to pass through the HDR meta data - i assume not as that will then automatically turn on the hdr profile (it does on my projector)

also do i pick a custom picture option on the projector and then use that for the tone mapping. Also what lamp level do most people use ie low or high.

i also assume i dont have to worry about using the measurement tool that is referred earlier in the tone mapping post
FGEvans is offline  
post #6021 of 6063 Old 05-31-2019, 12:15 PM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 836
Mentioned: 134 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 687
@madshi All those give me the lowest number of misses, even after adding more scenes:

Metric1: 8.9 / Metric2: 6.0
Metric1: 9.0 / Metric2: 5.8-5.9
Metric1: 9.1 / Metric2: 5.7-5.8
Metric1: 9.2 / Metric2: 5.6-5.7
Metric1: 9.3 / Metric2: 5.5-5.6
Metric1: 9.4 / Metric2: 5.4

So this works great, without having to change the logic:



Doc (please double check): https://www.mediafire.com/file/5j3cd...oV24.xlsx/file

PS: I'm going to watch Inferno now, let's see how it goes...

Last edited by Neo-XP; 05-31-2019 at 01:24 PM.
Neo-XP is offline  
post #6022 of 6063 Old 05-31-2019, 05:38 PM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 836
Mentioned: 134 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 687
Only one problem found with Inferno, this one:

With "disable sub of prev frame" OFF




With "disable sub of prev frame" ON



But looking at Metric1 for this case and others similar ones where a new scene is needed for 1 frame, it seems easy to fix without looking ahead.

Idea:

Line0: Before doing the sub of prev frame and only if the previous frame was detected as the beginning of a new scene -> if Metric1 is above [50] AND Metric2 is above [1], then there is a scene change

Line1: If Metric2 is below [1], then there is no scene change

Line2: Otherwise use weighted avrg. M1 thres [9.6] / M2 thres [8.8] / M1 weight % [50]



Doc: https://www.mediafire.com/file/qohen...oV25.xlsx/file

PS: When looking ahead, a simple way to handle this would be to create a new scene only if Line0 is matched, otherwise it is most probably a muzzle flash or a scene that is best ignored.
Manni01 and Javs like this.

Last edited by Neo-XP; 05-31-2019 at 11:23 PM.
Neo-XP is offline  
post #6023 of 6063 Old 06-01-2019, 12:03 AM
AVS Forum Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 7,573
Mentioned: 457 Post(s)
Tagged: 0 Thread(s)
Quoted: 2337 Post(s)
Liked: 3038
Thanks! Hmmmmm... How did you go from this:

Quote:
Originally Posted by Neo-XP View Post
Metric1: 8.9 / Metric2: 6.0
Metric1: 9.0 / Metric2: 5.8-5.9
Metric1: 9.1 / Metric2: 5.7-5.8
Metric1: 9.2 / Metric2: 5.6-5.7
Metric1: 9.3 / Metric2: 5.5-5.6
Metric1: 9.4 / Metric2: 5.4
... to this?
Quote:
Originally Posted by Neo-XP View Post
Line2: Otherwise use weighted avrg. M1 thres [9.6] / M2 thres [8.8] / M1 weight % [50]
It seems like a pretty dramagic increase for Metric2 for the weighted average?

I'm not sure if we should worry too much about scenes like the one in Inferno, which should clearly be easy to fix by looking ahead.
Manni01 and omarank like this.
madshi is offline  
post #6024 of 6063 Old 06-01-2019, 12:59 AM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 836
Mentioned: 134 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 687
Quote:
Originally Posted by madshi View Post
Thanks! Hmmmmm... How did you go from this:


... to this?

It seems like a pretty dramagic increase for Metric2 for the weighted average?
Yes, adding new scenes from Fury Road and Inferno did this.

I'm going to watch some movies today, so this can change again a bit, but I don't think it will change much because Inferno was a hell of a movie for scene detection .

Quote:
Originally Posted by madshi View Post
I'm not sure if we should worry too much about scenes like the one in Inferno, which should clearly be easy to fix by looking ahead.
It can already improve a lot actually. I'm thinking for instance about the sandstorm sequence from Fury Road which is a big mess with the sub of prev frame enabled.

Also, some important scenes are missed because there is a big change just before the cut, so I'm wondering if this:

Line0: Before doing the sub of prev frame and only if the previous frame was detected as the beginning of a new scene -> if Metric1 is above [50] AND Metric2 is above [1], then there is a scene change

Shouldn't become:

Line0: Before doing the sub of prev frame -> if Metric1 is above [50] AND Metric2 is above [1], then there is a scene change

Because I haven't found any false positive with more than 50 for Metric1 and 1 for Metric2 at the same time.

The same logic seems to work when looking ahead:

If FALL goes up and down quickly (within a few frames), if Metric1 is below [50] and Metric2 is below [1] on FALL changes, then there is no scene change

Of course, the value of 50 is a guess at the moment, we will need to add a margin of error (it would be more like ~60-65) to be safe.

Testing the sandstorm sequence from Fury Road should be a good starting point to find the sweet spot.

Edit: But maybe we should leave this for later and only focus on finding the best Metric1 and Metric2 thresholds for "regular" scenes for now.

Edit2: I just got 1 less miss using LibreOffice solver instead of Exel's:



Edit3: It seems to stabilize at 10.2 for M1 and 7.2 for M2:



Doc: https://www.mediafire.com/file/cdmtv...oV26.xlsx/file

Last edited by Neo-XP; 06-01-2019 at 07:37 AM.
Neo-XP is offline  
post #6025 of 6063 Old 06-01-2019, 11:21 PM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 836
Mentioned: 134 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 687
Only one problem found while watching Transformers, this brightness jump:


Transformers - Frames 124827-124831


If we adapt immediately during fade ins:




If we block the scene detection during fade ins:



I won't add this one to the sheet, because those cases are easy to fix without increasing the thresholds, either by:

- blocking the scene detection during fade ins/out (without looking ahead)
- adapting the target immediately during fade ins/out (better for image quality, but safer when looking ahead)

Edit:

Last answer from me on build 86:



Doc: https://www.mediafire.com/file/ghc67...oV27.xlsx/file

Results:


Edit:

To summarize, I see 3 things that can be improved from here when looking ahead:

- See if it is possible somehow to detect a real scene which comes right after a big change (it looks difficult, but not impossible?)

- Adapt to the ideal target during fade ins/outs to remove brightness jumps caused by false detections and improve image quality

- Decide on which situations it is better to let a false positive to be detected as a new scene (flash only on a part of the image VS flash on the entire image for instance)
Manni01 and Mr. Neverbicker like this.

Last edited by Neo-XP; 06-03-2019 at 07:06 PM.
Neo-XP is offline  
post #6026 of 6063 Old 06-03-2019, 01:34 AM
Newbie
 
Join Date: Feb 2016
Posts: 7
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 6 Post(s)
Liked: 5
I watched Captain Marvel UHD with the last build and the above settings yesterday ... It was fine and I didn't notice anything wrong...

(That was probably the most useless post ever... )
Ian_Currie, tommarra and Neo-XP like this.
DocCharky is offline  
post #6027 of 6063 Old 06-05-2019, 11:39 AM
Advanced Member
 
tommarra's Avatar
 
Join Date: Aug 2013
Location: USA
Posts: 569
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 340
Wondering if some optimisations were done on processing speeds. The new test build is significantly faster than the previous ones. Everything looks pretty good with Neo-XPs latest settings


Sent from my iPhone using Tapatalk
Ian_Currie likes this.

Gear: Fronts: B&W 803D3 | Center: B&W HTM2D3 | Surrounds: B&W 705S2 | Backs: B&W DM601S3 | Heights: Definitive Techonlogy ProCinema 1000 | Receiver & Amps: Anthem MRX 720 + Emotiva BasX-5 | Projector: JVC NX7 | Source: HTPC with Nvidia 2070, Kodi + External Player: MPC-HC with MadVR (for Blu-ray rips), Nvidia Shield (for streaming content)
tommarra is online now  
post #6028 of 6063 Old 06-11-2019, 09:07 AM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 836
Mentioned: 134 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 687
@madshi

While watching some movies, everything seems to work well on the scene detection topic (with the optimal values from the sheet), until now.

Unfortunately, I just found an unrelated issue, with the tone mapping adaptation.

You can check it from this sample at around 00:00:03 on the clouds:

https://www.mediafire.com/file/c0h2x...ation.mkv/file

Warning: it's pretty bad

Last edited by Neo-XP; 06-11-2019 at 11:42 AM.
Neo-XP is offline  
post #6029 of 6063 Old 06-15-2019, 12:31 AM
Advanced Member
 
Neo-XP's Avatar
 
Join Date: Jun 2018
Location: Switzerland
Posts: 836
Mentioned: 134 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 687
@madshi Hi,

There is an issue with the Universal logo and the immediate target adaptation during fade ins.

I uploaded a sample here: https://www.mediafire.com/file/qbjhw..._logo.mkv/file

You can see the flickering at the top left side during the fade in.

I think it might be a good idea to implement some kind of protection to prevent the target going up and down from one frame to the next during fade ins, similar to the darkness adaptation delay (of 2s IIRC).

Edit: or better, just say that the target can only go UP during fade ins and DOWN during fade outs. That would fix it safely and easily.

PS: I am available if some testing is needed
SOWK, Icaro, Manni01 and 4 others like this.

Last edited by Neo-XP; 06-15-2019 at 12:54 AM.
Neo-XP is offline  
post #6030 of 6063 Old 06-19-2019, 08:07 AM
Member
 
Join Date: Dec 2016
Posts: 92
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 72 Post(s)
Liked: 10
How do I install a new build? Probably a dumb question, but not sure how.
Klaus.S 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