Please post here for issues related to UHD discs
-
speeddemon
- Posts: 77
- Joined: Wed Oct 16, 2019 3:44 am
Post
by speeddemon » Fri Jul 07, 2023 11:02 pm
RESET_9999 wrote: ↑Fri Jul 07, 2023 10:37 pm
IMO, It's better to generate a new RPU in resolve because:
1- if you use hdr10plus, your min_pq will always be 0
2- HDR10plus avg_pq is VERY different than what Dolby Vision produces. DV has some specific calculations for the avg_pq.
3- with Resolve, you can make true CMV4.0 metadata (with L3) which improves tone mapping.
Thanks! Understood.
Another question... have you done much research yet comparing Resolve Analyzed output with typical metadata that includes L2/L8 trims?
-
RESET_9999
- Posts: 2388
- Joined: Mon Aug 05, 2019 7:12 pm
Post
by RESET_9999 » Fri Jul 07, 2023 11:22 pm
speeddemon wrote: ↑Fri Jul 07, 2023 11:02 pm
Another question... have you done much research yet comparing Resolve Analyzed output with typical metadata that includes L2/L8 trims?
Yes
For some reason, resolve does not export the L8 trims similar to when you set your project to cmv2.9 which does not export L2.
So for the resolve cmv4.0 output, a cmv4.0 device won't have any trims effect: only L1 + L3 but for a cmv2.9 device, automatic L2 trims will brighten everything below the tv capabilities and darken every thing above.
BTW, I think only one time I've seen an actual CMV4.0 rpu with active L8... strangely, everything else doesn't have any L8 trim or it's set to 2048 no effect.
There's a bunch of test files in my GD in which you can test pretty much everything DV related.
https://drive.google.com/drive/u/1/fold ... V0E3VVkhiC
-
speeddemon
- Posts: 77
- Joined: Wed Oct 16, 2019 3:44 am
Post
by speeddemon » Sat Jul 08, 2023 12:24 am
RESET_9999 wrote: ↑Fri Jul 07, 2023 11:22 pm
So for the resolve cmv4.0 output, a cmv4.0 device won't have any trims effect: only L1 + L3 but for a cmv2.9 device, automatic L2 trims will brighten everything below the tv capabilities and darken every thing above.
Is the L2 part actually true for both CMv2.9 and CMv4 analysis when outputting both when outputting CMv2.9 or CMv4? Or, do you just strip the L2 trims when processing with your script anyways?
-
RESET_9999
- Posts: 2388
- Joined: Mon Aug 05, 2019 7:12 pm
Post
by RESET_9999 » Sat Jul 08, 2023 1:03 am
speeddemon wrote: ↑Sat Jul 08, 2023 12:24 am
part actually true for both CMv2.9 and CMv4 analysis when outputting both when outputting CMv2.9 or CMv4? Or, do you just strip the L2 trims when processing with your script anyways?
trims are reset to 0 or removed depending on your script configuration. You can set the script to keep the trims but by default, they are overwritten as explained a couple of days ago in this thread
-
speeddemon
- Posts: 77
- Joined: Wed Oct 16, 2019 3:44 am
Post
by speeddemon » Sat Jul 08, 2023 1:09 am
RESET_9999 wrote: ↑Sat Jul 08, 2023 1:03 am
speeddemon wrote: ↑Sat Jul 08, 2023 12:24 am
part actually true for both CMv2.9 and CMv4 analysis when outputting both when outputting CMv2.9 or CMv4? Or, do you just strip the L2 trims when processing with your script anyways?
trims are reset to 0 or removed depending on your script configuration. You can set the script to keep the trims but by default, they are overwritten as explained a couple of days ago in this thread
And you recommend the trims being overwritten for the reasons you stated earlier, correct?
But, just to make sure I understand correctly, can you confirm the following with regards to Resolve?
- CMv2.9 Analysis & CMv2.9 output = L1
- CMv4 Analysis & CMv2.9 output = L1+L2
- CMv4 Analysis & CMv4 output = L1+L3 in CMv4 or L1+L2 in CMv2.9 Fallback
-
RESET_9999
- Posts: 2388
- Joined: Mon Aug 05, 2019 7:12 pm
Post
by RESET_9999 » Sat Jul 08, 2023 1:12 am
speeddemon wrote: ↑Sat Jul 08, 2023 1:09 am
And you recommend the trims being overwritten for the reasons you stated earlier, correct?
But, just to make sure I understand correctly, can you confirm the following with regards to Resolve?
- CMv2.9 Analysis & CMv2.9 output = L1
- CMv4 Analysis & CMv2.9 output = L1+L2
- CMv4 Analysis & CMv4 output = L1+L3 in CMv4 or L1+L2 in CMv2.9 Fallback
That's correct.
Actually, the L8 100nits trim is exported but it's set to 0 and obviously not useful to DV playback.
-
speeddemon
- Posts: 77
- Joined: Wed Oct 16, 2019 3:44 am
Post
by speeddemon » Sat Jul 08, 2023 2:52 pm
RESET_9999 wrote: ↑Sat Jul 08, 2023 1:12 am
speeddemon wrote: ↑Sat Jul 08, 2023 1:09 am
And you recommend the trims being overwritten for the reasons you stated earlier, correct?
But, just to make sure I understand correctly, can you confirm the following with regards to Resolve?
- CMv2.9 Analysis & CMv2.9 output = L1
- CMv4 Analysis & CMv2.9 output = L1+L2
- CMv4 Analysis & CMv4 output = L1+L3 in CMv4 or L1+L2 in CMv2.9 Fallback
That's correct.
Actually, the L8 100nits trim is exported but it's set to 0 and obviously not useful to DV playback.
Which of these typically look most accurate to the original Profile 7?
-
valington
- Posts: 44
- Joined: Tue May 02, 2023 1:39 pm
Post
by valington » Sat Jul 08, 2023 4:45 pm
RESET_9999 wrote: ↑Sat Jul 08, 2023 1:12 am
speeddemon wrote: ↑Sat Jul 08, 2023 1:09 am
And you recommend the trims being overwritten for the reasons you stated earlier, correct?
But, just to make sure I understand correctly, can you confirm the following with regards to Resolve?
- CMv2.9 Analysis & CMv2.9 output = L1
- CMv4 Analysis & CMv2.9 output = L1+L2
- CMv4 Analysis & CMv4 output = L1+L3 in CMv4 or L1+L2 in CMv2.9 Fallback
That's correct.
Actually, the L8 100nits trim is exported but it's set to 0 and obviously not useful to DV playback.
Reset may be you could do one of your scripts without DRAG AND DROP where we could just put file into the folder and choose programm we need
-
RESET_9999
- Posts: 2388
- Joined: Mon Aug 05, 2019 7:12 pm
Post
by RESET_9999 » Sat Jul 08, 2023 9:47 pm
speeddemon wrote: ↑Sat Jul 08, 2023 2:52 pm
Which of these typically look most accurate to the original Profile 7?
depends on how the original p7 is I guess? (cmv2.9 or cmv4.0 / trims or not)
the best you can hope as I said, is cmv4.0 without trims and no point using anything else
-
RESET_9999
- Posts: 2388
- Joined: Mon Aug 05, 2019 7:12 pm
Post
by RESET_9999 » Sat Jul 08, 2023 9:48 pm
valington wrote: ↑Sat Jul 08, 2023 4:45 pm
Reset may be you could do one of your scripts without DRAG AND DROP where we could just put file into the folder and choose programm we need
this won't happen.
Drag and drop is a much better approach because your input can be anywhere in your PC (different drives)
-
dapope
- Posts: 39
- Joined: Fri Oct 09, 2020 4:54 am
Post
by dapope » Sun Jul 09, 2023 10:42 am
@RESET_9999
Salut, sorry for the maybe silly question but how do I get a DV P8 when baking in FEL using your 6-9 workflow? In changelog you say you added in option to export in DV but I still only get option for HDR or SDR.
-
RESET_9999
- Posts: 2388
- Joined: Mon Aug 05, 2019 7:12 pm
Post
by RESET_9999 » Sun Jul 09, 2023 11:29 am
dapope wrote: ↑Sun Jul 09, 2023 10:42 am
@RESET_9999
Salut, sorry for the maybe silly question but how do I get a DV P8 when baking in FEL using your 6-9 workflow? In changelog you say you added in option to export in DV but I still only get option for HDR or SDR.
P7 rpu will be converted to P8 automatically so once the encoding is completed, just remux the hevc with mkvtoolnix or tsmuxer and DV P8 will be present.
-
dapope
- Posts: 39
- Joined: Fri Oct 09, 2020 4:54 am
Post
by dapope » Sun Jul 09, 2023 11:56 am
RESET_9999 wrote: ↑Sun Jul 09, 2023 11:29 am
dapope wrote: ↑Sun Jul 09, 2023 10:42 am
@RESET_9999
Salut, sorry for the maybe silly question but how do I get a DV P8 when baking in FEL using your 6-9 workflow? In changelog you say you added in option to export in DV but I still only get option for HDR or SDR.
P7 rpu will be converted to P8 automatically so once the encoding is completed, just remux the hevc with mkvtoolnix or tsmuxer and DV P8 will be present.
Ok, thanks very much! Is there an easy way to tell for myself what FEL titles have this issue?
-
dapope
- Posts: 39
- Joined: Fri Oct 09, 2020 4:54 am
Post
by dapope » Sun Jul 09, 2023 12:09 pm
Thanks but I meant to check myself in case there was a title you hadn't looked at. I guess you are pretty thorough, though.