Fanrestore - Fan Restoration Forum

Full Version: FANRES: Dolby Vision P5 layer was converted to P8
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Pages: 1 2
Code:
FANRES: Dolby Vision P5 layer was converted to P8

I keep seeing this on recent FraMeSToR Dolby Vision releases but don't know what it means.

Anyone enlighten me?

Thanks.
Fanres is just being used as a generic name for fan restoration(possible in reference to a tor site) but not a reference to this site specifically.

If someone is converting DV here to something else I am unaware but would love to be able to do it myself.
So

fanres=(fan restoration) : postit=(small paper sheet that has an adesive on back)

that's great! Big Grin
Sorry I should have been more specific. I was actually referring to the conversion of Dolby Vision P5 layer to P8. What exactly is that?
Found here https://forum.makemkv.com/forum/viewtopi...start=5325:

Quote:the best DV is FEL P7 (dual-layer) but not all the movies are FEL. Note that your TV internal app won't play FEL movies properly because it lack of a 2nd decoder but your x700 will play properly pretty much anything you throw at him

P7 fel(always dual layer) = 1 HDR10 base layer 10bits + 12bits data in EL(2 extra bits) + dynamic metadata in RPU
P7 mel (always dual layer) = 1 HDR10 base layer 10bits + blank EL (no extra bit) + dynamic metadata in RPU
P8 (always mel and single layer) = 1 HDR10 base layer 10bits + dynamic metadata in RPU
P5 (always single layer) = none-tonemapped DV base layer(green and purple colors) + dynamic metadata in RPU
P4 (dual layer,FEL or MEL) = 1 SDR base layer + FEL or MEL + dynamic metadata in RPU

so it goes like this:
P7 fel > P7 mel = P8 = P5


*Profile 7 can be merged in a single track (makemkv) but it's still dual layer DV.
*Profile 8 can also have an HLG base layer (iphone recording) so this format is compatible with HLG, SDR and DV.
*profile 5 DV : I'm not sure if it's 12bits or 10 bits because there's this description about my profile 5 patterns: ''10bits encoded remapped 12bits''. Not really sure what it means but it sounds like profile 5 could be 12bits as well.


not that I understood too much, though... Big Grin

A Dolby Vision expert is needed!
Judging by the release notes it seems they're taking the DV data from WEB-DLs and syncing it with the BR HDR releases.

Code:
This is a HYBRID DV REMUX, including the WEB DV metadata converted to P8 and synced (Thanks Visi0n!).
(2021-12-27, 12:21 PM)Hitcher Wrote: [ -> ]Judging by the release notes it seems they're taking the DV data from WEB-DLs and syncing it with the BR HDR releases

If that's true, that's pretty big news. Especially since many distributors will only use HDR10 on disc but DV on streaming. I wonder if it's a simple manor of frame syncing the sources and running a simple script. Any information on how they did it?
Anyone ever figure out how this works?

Been thinking about the issue lately and how it might be cool to add a DV layer from Web-DLs to UHD BDs. I would assume that the inherent details in the "luma" would have to be the same encode but it seems that's not the case based off the previous work this guy has done. Bare minimum it would be nice to create Web-DL's with both a HDR layer and a DV layer instead of having to choose just one. Like a UHD BD, have the TV recognize a DV layer but fall back to HDR if the TV doesn't support it. I know you can add a DVD layer in MKVmerge but not sure how to strip the DV layer out of an existing stream.

But I'm just thinking out loud.
FraMeSToR are releasing loads of these encodes but you'd need to be in the scene to find out how they're doing it.
There is a tool called dovi_tool that can be found on github that allows the manipulation of HEVC streams containing Dolby Vision data. With it you can export the Dolby Vision layer from one file and apply it to another.
It's a pretty straightforward process if you already have both files and both sources match.
I have once applied a Dolby Vision layer from a web source to a uhd-bd and the final result looked fine to me. I would not be surprised if this method is not 100% accurate considering the base layer might not be the exact same but I don't see what else could be done better with what we have.
If I remember correctly, this is what I did:
  1. Demux both files to get the raw HEVC stream (use the enhancement layer from the BD or the merged HEVC stream if it has been converted to single layer)
  2. Use dovi_tool to extract the rpu from the HEVC stream containing Dolby vision
  3. Inject rpu to the BD HEVC stream using dovi_tool
  4. Remux everything back to a format supporting Dolby Vision, I used MKV in my case

The end result is a Dolby Vision Profile 8 file that is also HDR10 compatible. The exact same procedure could be done with a HDR10 web source instead of a BD to create a file containing both HDR10 and Dolby Vision.
Pages: 1 2