Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > Video Encoding > High Efficiency Video Coding (HEVC)

Reply
 
Thread Tools Search this Thread Display Modes
Old 2nd February 2025, 17:55   #981  |  Link
ac777103
Registered User
 
Join Date: Oct 2024
Posts: 22
thanks, while my TV supports cm v4 I tend to avoid tv-led playback as it can be glitchy (regardless of DV playback device). So I tend to switch between the ATV and the Ugoos for player-led. Either way, looks like I should stick with the CPM build.
ac777103 is offline   Reply With Quote
Old 2nd February 2025, 18:00   #982  |  Link
Vymrr
Registered User
 
Join Date: Aug 2024
Posts: 12
Quote:
Originally Posted by Kuler087 View Post
Did you enable L5? In A13, it's disabled by default.

https://www.avsforum.com/posts/63803659/
Thanks! Got it enabled but it looks like my TV blacks out my OSD and subtitles on the black bars.. got it reverted and waiting for newer cpm version

I got another question not really related to this, but concerning DV movies in general with dark scenes, the dark is really dark and it's really hard to see sometimes, is this considered normal? For instance, for web streaming services (netflix, etc), this problem doesn't seem to exist, and the picture seems to be perfectly balanced in dark scenes where I can still see details.
Vymrr is offline   Reply With Quote
Old 2nd February 2025, 18:14   #983  |  Link
Kuler087
Registered User
 
Kuler087's Avatar
 
Join Date: Aug 2021
Location: Canada
Posts: 473
There's a black crush issue with cmv2.9 content especially if the RPU MDL is 4000nits. For 1000nits RPU, it's not really noticeable in real content but it is crushed in patterns.
I did some test files about this a while ago: https://drive.google.com/drive/u/1/f...5FsewWGPBII1Ib

On my C2, HDR10 and DV dark scenes look pretty much identical as long as the RPU is not cmv2.9 4000nits.
__________________
DoVi_Scripts
Kuler087 is offline   Reply With Quote
Old 2nd February 2025, 18:15   #984  |  Link
ac777103
Registered User
 
Join Date: Oct 2024
Posts: 22
thinking some more about player-led playback of HDR10 source files processed via 3-1. The ATV supports CM v4 in player-led, per the explanation in a previous post re positive lift, does this explain my impression that the devices have different brightness profiles in player-led mode? I find them to be more similar for MEL 2.9 content, and more different for my 3-1 conversions.
ac777103 is offline   Reply With Quote
Old 2nd February 2025, 18:24   #985  |  Link
Vymrr
Registered User
 
Join Date: Aug 2024
Posts: 12
Quote:
Originally Posted by Kuler087 View Post
There's a black crush issue with cmv2.9 content especially if the RPU MDL is 4000nits. For 1000nits RPU, it's not really noticeable in real content but it is crushed in patterns.
I did some test files about this a while ago: https://drive.google.com/drive/u/1/f...5FsewWGPBII1Ib

On my C2, HDR10 and DV dark scenes look pretty much identical as long as the RPU is not cmv2.9 4000nits.
How about CM4.0? The dark scenes also appear in movies converted to CM4.0 (using your tutorials).
Vymrr is offline   Reply With Quote
Old 2nd February 2025, 18:26   #986  |  Link
Kuler087
Registered User
 
Kuler087's Avatar
 
Join Date: Aug 2021
Location: Canada
Posts: 473
Since they are two different algorithm, there will be a difference between cmv2.9 or cmv4.0 playback but not much. In a cmv4.0 workflow, the v2.9 L1+generated_L2 should roughly equal v4.0 L1+L3

The ATV seems broken with cmv4.0 in the tests I did a while ago. with the pattern specifically: https://drive.google.com/file/d/1s-3...usp=drive_link
__________________
DoVi_Scripts
Kuler087 is offline   Reply With Quote
Old 2nd February 2025, 18:28   #987  |  Link
Kuler087
Registered User
 
Kuler087's Avatar
 
Join Date: Aug 2021
Location: Canada
Posts: 473
Quote:
Originally Posted by Vymrr View Post
How about CM4.0? The dark scenes also appear in movies converted to CM4.0 (using your tutorials).
As I said, as long as the RPU is not cmv2.9 and 4000nits MDL, DV dark scenes look identical to HDR10 on my C2.
__________________
DoVi_Scripts
Kuler087 is offline   Reply With Quote
Old 2nd February 2025, 19:30   #988  |  Link
ac777103
Registered User
 
Join Date: Oct 2024
Posts: 22
Quote:
Originally Posted by Kuler087 View Post
Since they are two different algorithm, there will be a difference between cmv2.9 or cmv4.0 playback but not much. In a cmv4.0 workflow, the v2.9 L1+generated_L2 should roughly equal v4.0 L1+L3

The ATV seems broken with cmv4.0 in the tests I did a while ago. with the pattern specifically: https://drive.google.com/file/d/1s-3...usp=drive_link
I've seen your screenshot of the ATV playing the Spears and Munsil color test. I have that disc and in my system the ATV does much better on that test. What we see is a combination of the player and the display and settings on both. I find my Panny OLED does a better job of rendering a 422 input from the ATV than I get from a 444 input from any Amlogic device. And 444 output from my Panny disc player beats the Ugoos on the SM test. So I avoid 444 playback from the Ugoos as I don't think it upsamples as well compared to my TV rendering 422 HDR10 (ATV) or 422 DV (ATV or Ugoos).
ac777103 is offline   Reply With Quote
Old 2nd February 2025, 19:43   #989  |  Link
Kuler087
Registered User
 
Kuler087's Avatar
 
Join Date: Aug 2021
Location: Canada
Posts: 473
Quote:
I have that disc and in my system the ATV does much better on that test. What we see is a combination of the player and the display and settings on both.
what?? the screenshot is an hdmi capture... No TV is involved...
Unless they fixed something recently, the ATV is broken in cmv4.0 at least with 4000-10000 content ( the blue channel specifically )

__________________
DoVi_Scripts
Kuler087 is offline   Reply With Quote
Old 2nd February 2025, 20:06   #990  |  Link
ac777103
Registered User
 
Join Date: Oct 2024
Posts: 22
Quote:
Originally Posted by Kuler087 View Post
Since they are two different algorithm, there will be a difference between cmv2.9 or cmv4.0 playback but not much. In a cmv4.0 workflow, the v2.9 L1+generated_L2 should roughly equal v4.0 L1+L3

The ATV seems broken with cmv4.0 in the tests I did a while ago. with the pattern specifically: https://drive.google.com/file/d/1s-3...usp=drive_link
I tried this test file and I think it's probably demonstrating how bad my Panny HZ980 is at TV-led DV.

With TV-led DV, from the Ugoos the first 3 or 4 red and green steps are a single block, no grading at all, but the blue channel has all the steps. On the ATV/Infuse all the steps are present in the red channel but there is terrible noise in the 10000 red block. Like the Ugoos, the green channel is one block until it drops to 2500 nits or so. The blue is cleaner on the Ugoos than from the ATV.

In player-led, the ATV loses the step grading in the red as well as the green at the brightest levels, it's not very good. The Ugoos renders a clean pattern with all the steps but of course it's not CMv4.

This test might confirm my impression that picture quality on my TV is degraded in TV-led mode, since neither the ATV nor the Ugoos were able to produce that good a result in TV-led mode. But it might also be showing that player-led CM v4 from the ATV isn't that great either, suggesting that I may as well stick with player-led from the Ugoos until I change my display?
ac777103 is offline   Reply With Quote
Old 2nd February 2025, 20:08   #991  |  Link
ac777103
Registered User
 
Join Date: Oct 2024
Posts: 22
apologies for the confusion, my earlier reply related to a different screenshot you'd posted.
ac777103 is offline   Reply With Quote
Old 3rd February 2025, 20:02   #992  |  Link
prudentavocado
Registered User
 
Join Date: Jan 2025
Posts: 2
Quote:
Originally Posted by Kuler087 View Post
Maybe other streaming services have a different encode but it is unlikely. This movie was mastered at 4000nits and they tone mapped a 1000nits version for the bluray release.


With plex. The newer GTV chromecast probably dropped P7 support so you need to use the older device.
Anyway, don't use this device, the BL and EL are out of sync and it's doing ''fake'' TV-LED.
Get a Ugoos, you wont regret it, especially with the latest development that allows unaltered RPU processing and Level 5 + positive lift support.



Yes, these groups bake FEL properly and can be played on any device. Their older encodes might not be though.


no
Thanks

Oh I see.....So, "List of DV movies that can't be converted to profile 8 Dolby Vision" doesn't necessarily mean there is a technical hurdle to where nobody can convert it. If it's one of the proper groups then they could actually do it. (as long as it's not an older encode)


That's a big difference I was missing; I was avoiding all those releases in that red section because I thought it was an impossibility. Why include the BD-FEL ones in the red section then? Seems like you could move those to the green list if we just stick to the proper encoder groups.
prudentavocado is offline   Reply With Quote
Old 3rd February 2025, 20:12   #993  |  Link
Kuler087
Registered User
 
Kuler087's Avatar
 
Join Date: Aug 2021
Location: Canada
Posts: 473
The list is for REMUX rip and I say:

Quote:
Can only be played on a FEL device and can't be converted to P8 without baking FEL into BL
The list about the proper encoder groups is just a bonus. I really don't care about encodes, I only watch/keep REMUX.
__________________
DoVi_Scripts
Kuler087 is offline   Reply With Quote
Old 4th February 2025, 16:12   #994  |  Link
Kuler087
Registered User
 
Kuler087's Avatar
 
Join Date: Aug 2021
Location: Canada
Posts: 473
3.0.6 Released: https://github.com/R3S3t9999/DoVi_Scripts/releases

Quote:
** LATEST Tools pack + bat files + GUI (bat files only link below)**
[GoogleDrive](https://drive.google.com/file/d/1S4d...usp=drive_link)
[MEGA](https://mega.nz/file/BO0TFK5Y#m3jW75...mvntq_OFpUJWZM)
(case sensitive) PASSWORD FOR THE TOOLS PACK IS THE ZIP FILENAME= DoVi.Scripts

**CHANGES:**

- (7-4) NEW workflow that can quickly and automatically create 2min (10 frames) DV L1 vs L2 metadata comparison [videos like these](https://drive.google.com/drive/folde...iSlcdMiy2ka0mf)
- (7-5) NEW workflow that can quickly and automatically create 2min (10 frames) FEL vs MEL comparison[ videos like these.](https://drive.google.com/drive/u/0/f...BIQmziCe_R_IKe)
- (8-2) NEW: Can now create DV/HDR10 samples of a single frame in loop (line 180, default = NO)
you must [add these new files](https://drive.google.com/drive/folde...usp=drive_link) to your tools folder. (or re-download the tools pack)
- (6-2) Sped up the P5 input measurement 2.2X (if your cpu can handle it)
- (6-2) Sped up when you opt to crop and use prores
- (3-1) Added a workaround for mkv/ts/m2ts files that fail with madmeasureHDR.exe
- (6-3)(6-4) Added more stats to the L1 plot
- (1) RPU input + ''e'' selection can now loop the RPU frame(s) to check/info.
- (7-1) OSD will now display the frame(s) Level 1 rpu values in nits
- (7-2) OSD can now display the frame(s) Level 1 rpu values in nits if DV is present (line 123 , default = NO)
- (7-2) Can now accept up to 20 inputs
- (7-2) Manual mode will now ask for the frame to export before indexing.
__________________
DoVi_Scripts
Kuler087 is offline   Reply With Quote
Reply

Tags
batch scripts, dolby vision processing, dovi_scripts

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

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

Forum Jump


All times are GMT +1. The time now is 21:36.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2025, vBulletin Solutions Inc.