[Soulnight]

With this new version, using the default settings (other than 105 real display nits and 125 dynamic tuning) I am seeing some huge fluctuations in brightness.

For example, Oceans Eight, chapter 6 (Vogue offices) there is a conversation that takes place between two women. Every time it changes to the women who is conducting the interview, the brightness changes. One time it is dark, the next it is bright, then dark, then bright etc.

There was zero change in dynamic target nits algo.
Dynamic clipping only was changed.

So when you say in this new version, what version are you comparing it too?

For me, it only seems you may not like the default value for scene/chapter detection and rolling avg.

If stability is your priority, you should try @Dexter Kane settings in that matter:

Rolling avg: 120
Merge scene: 350
Minimum chapter duration: 24frames
Merge chapter: 200



This is only a matter of taste.

 

[Ian_Currie]

There was zero change in dynamic target nits algo.
Dynamic clipping only was changed.

So when you say in this new version, what version are you comparing it too?

For me, it only seems you may not like the default value for scene/chapter detection and rolling avg.

If stability is your priority, you should try @Dexter Kane settings in that matter:

Rolling avg: 120
Merge scene: 350
Minimum chapter duration: 24frames
Merge chapter: 200



This is only a matter of taste.

I was comparing it to 3.5.6. I was just pointing it out because with every new version I always use the defaults for scene/chapter detection and never noticed anything like this before.

If it doesn't worry you it doesn't worry me I know I can control it... and I will try Dexter's.

Thanks, Flo.

 

[Neo-XP]

@Neo-XP

Please could you post your current settings for scene/chapter change with the tool?

I'd like to try your approach here as well after becoming a convert with the live algo in the other thread

I need to recalc my test files for my new peak nits (120nits), so it would be a good opportunity to try.

I don't really use the tool anymore, but the last settings I used were these:



But now I prefer the result with the live algo, based on lum changes and not on FALL change.

I hope you will also test the different brightness adaptation speeds with the live algo and report on the other thread

 

[Soulnight]

But now I prefer the result with the live algo, based on lum changes and not on FALL change.

Can you define "lum changes"? Or @madshi maybe?

 

[madshi]

He means madVR's "scene change detection" metric. The live algo uses the same metric which madMeasureHDR uses to split the movie into scenes. However, for dynamic HDR the live algo uses a higher threshold. If you'd like to test the same metric with your optimization tool, I can email you the algorithm details.

 

[Soulnight]

He means madVR's "scene change detection" metric. The live algo uses the same metric which madMeasureHDR uses to split the movie into scenes. However, for dynamic HDR the live algo uses a higher threshold. If you'd like to test the same metric with your optimization tool, I can email you the algorithm details.

Thanks. Yes, please. I would like to test that.

 

[Dexter Kane]

With this new version, using the default settings (other than 105 real display nits and 125 dynamic tuning) I am seeing some huge fluctuations in brightness.

For example, Oceans Eight, chapter 6 (Vogue offices) there is a conversation that takes place between two women. Every time it changes to the women who is conducting the interview, the brightness changes. One time it is dark, the next it is bright, then dark, then bright etc.

Maybe give my chapter settings a try:

Rolling average: 120
Scene merge: 350
Minimum chapter duratuon: 24
Chapter merge: 200

 

[Dexter Kane]

There was zero change in dynamic target nits algo.
Dynamic clipping only was changed.

So when you say in this new version, what version are you comparing it too?

For me, it only seems you may not like the default value for scene/chapter detection and rolling avg.

If stability is your priority, you should try @Dexter Kane settings in that matter:

Rolling avg: 120
Merge scene: 350
Minimum chapter duration: 24frames
Merge chapter: 200



This is only a matter of taste.

Beat me to it

 

[Jacob B]

Situation:
I have a mediaserver and two different screens with each their own HTPC: a dedicated home theater with a JVC RS540/7900 and the living room with a LG OLED 65B7. I recently upgraded both HTPCs with new madvr 4K capable GPUs. I currently use the HDR pass through mode in madVR (and stream from Apple TV 4K as well).

Problem:
I don’t mind the “hassle” of measuring 4K HDR movies in advance, so I would like to use measurement files together with the clipping optimized tool. However, the clipping settings would be quite different for playback on the OLED and on the JVC...

Question:
can the clipping optimizer handle different sets of measurement files in a practical manner?
Or am I better off using a pragmatic solution such as using the new live algo in the living room and measurement files in the theater?
Which solution requires most of the GPU (theater GPU is RTX2060 and much faster than living room GTX 1050ti).

Thanks for the great work - programmers and testers alike 👍🏆🙂

 

[Ian_Currie]

Maybe give my chapter settings a try:

Rolling average: 120
Scene merge: 350
Minimum chapter duratuon: 24
Chapter merge: 200

Yup, your setting are what I'll try next... however I'm now not using measurement files (or rather the latest live algo from madshi based on Flo & Anna's work) and I'm getting superb results. I'm both impressed and puzzled. It's a very different 'look' (for the same dynamic tuning value) that is more like what I was trying to achieve.

 

[chros73]

Just a note that the fix in madVRhdrMeasure46 affects measurement files.

 

[Dexter Kane]

@Soulnight

Now that we may have to remeasure everything again can I make a suggestion for your tool? Would it be possible to have the tool write a file when it begins working on a movie to say that it's currently processing this movie, so that you could run multiple instances of the tool on different machines on the same directory without them trying to process the same movie at the same time?

 

[arcspin]

Hi
I need a quick refresher course.

To use MadMeasureDynamicClipping V3.6.2 to remeasure movie files using madVRhdrMeasure46 (edit) 47.

Should I click "Restore original measurements" or "Delete all measurement" before I start the measurements?


Thanx,

 

[madshi]

New test build 47:

http://madshi.net/madVRhdrMeasure47.zip

Unfortunately I have to recommend that you trash all measurement files created by builds 41-46. But all measurement files created with builds up to 40 should be fine.
@arcspin, why do you want to remeasure all your movie files? It's not needed. Just remeasure those you made with builds 41-46.

 

[arcspin]

To be honest,
I'm starting to loose track of what files I have measured with what version.

Is there an easy way to check which version of madVRhdrMeasure that has been used?

 

[madshi]

Not really, other than the date of the measurement file. Test build 41 was released just 6 days ago, so any measurement files older than that should be fine.

 

[Soulnight]

To be honest,
I'm starting to loose track of what files I have measured with what version.

Is there an easy way to check which version of madVRhdrMeasure that has been used?

In our detail log file, you have the date when the measurement file has been created. Also you can sort out the measurement files per date and remove any measurement generated after publication of beta V41.

Actually, we are going to do that ourself right now.

 

[chros73]

Just a note that the fix in madVRhdrMeasure46 affects measurement files.

No need for this, sorry about it.

 

[arcspin]

Just to clarify: and the date to look into would be the measurement file found in the original folder?

 

[Soulnight]

Just to clarify: and the date to look into would be the measurement file found in the original folder?

What we did was: restore measurement files, and then look with the search option of windows *measurements which file had been created with a verson of madmeasurehdr from beta41 or after. And delete those.