Dolby Vision now possible through MP4 Mux.

[RESET_9999]

All the metadata levels are generated automatically by the Dolby Algo but then indeed, the trims are adjusted manually by the colorist for most of the movies. The 100nits trim is also adjusted manually because it's going to be delivered for the SDR bluray. The colorist can also blend shots together or copy L1/L2 from one shot to another.

I wouldn't be surprised if the movies with only a 100nits L2 trim were not edited manually though.

Would it then be better/cleaner to generate a Resolve rpu in all these cases? (Just keeping the L1 metadata ).

or better , use my script 3-1 and use the original scene cuts.

I personally go even further, I generate new L1 for movies without 600-1000trims that were done with the old cmv2.9 algo because the latest version of the algo with tunning is a lot superior IMO.
you can read about it here: https://professionalsupport.dolby.com/s ... uage=en_US

And only keep a WEB DV if the MDl are close and also has L2 600 or 1000 or 2000 nits trims?

static MDL metadata is not really important, what matters is the PQ and nits measurement comparisons as I do in my excel sheet.
https://docs.google.com/spreadsheets/d/ ... edit#gid=0

[HotFudge]

Would it then be better/cleaner to generate a Resolve rpu in all these cases? (Just keeping the L1 metadata ).
or better , use my script 3-1 and use the original scene cuts.

How do I use the original scene cuts with your script 3-1

I personally go even further, I generate new L1 for movies without 600-1000trims that were done with the old cmv2.9 algo because the latest version of the algo with tunning is a lot superior IMO.

looks like Resolve the way to go. Do you discard the L2 100nit when you generate L1 Resolve metadata? or only discard the L2 600-1000 trims?

Cheers

[RESET_9999]

How do I use the original scene cuts with your script 3-1

its all automated but read the instructions on the github webpage for more info: MODE.H (3)

looks like Resolve the way to go. Do you discard the L2 100nit when you generate L1 Resolve metadata? or only discard the L2 600-1000 trims?

The script discard 600-1000nits and keep 100nits by default but you can control which trims is kept or removed


::Choose if you want to remove the 100 L2 trim pass. (only valid for 1-1xml and 3-1) (require the dolby metafier) (default=NO)
set remove_100=NO
::Choose if you want to remove the 600 L2 trim pass.(only valid for 1-1xml and 3-1) (require the dolby metafier)(default=YES)
set remove_600=YES
::Choose if you want to remove the 1000 L2 trim pass.(only valid for 1-1xml and 3-1) (require the dolby metafier) (default=YES)
set remove_1000=YES

[HotFudge]

How do I use the original scene cuts with your script 3-1

its all automated but read the instructions on the github webpage for more info: MODE.H (3)

looks like Resolve the way to go. Do you discard the L2 100nit when you generate L1 Resolve metadata? or only discard the L2 600-1000 trims?

The script discard 600-1000nits and keep 100nits by default but you can control which trims is kept or removed


::Choose if you want to remove the 100 L2 trim pass. (only valid for 1-1xml and 3-1) (require the dolby metafier) (default=NO)
set remove_100=NO
::Choose if you want to remove the 600 L2 trim pass.(only valid for 1-1xml and 3-1) (require the dolby metafier)(default=YES)
set remove_600=YES
::Choose if you want to remove the 1000 L2 trim pass.(only valid for 1-1xml and 3-1) (require the dolby metafier) (default=YES)
set remove_1000=YES

Cheers buddy

Reading the article this looks great! have your tools been updated with this new analysis tuning as it says its in Resolve 18.6 and later?

Which tools have implemented the new analysis tuning?
As of September 2023, the new analysis tuning is supported in DaVinci Resolve Studio version 18.6 and later. Other partner products such as Filmlight Baselight and Colorfront Transkoder will implement the new analysis tuning in their upcoming versions.

Are those changes considered as new Dolby Vision CM version?
No. This update only affects how the L1 metadata is generated and can be considered an update to the existing CM version 4.0.

[RESET_9999]

yes, my script use the same Dolby algo. You can control the tuning in the script main settings

:: choose the L1 analysis tuning... Choices are: old=0 - Most Highlight Detail/Most Mapping=1(default) - More Highlight Detail/More Mapping=2 - Balanced=3 - Less Highlight Detail/Less Mapping=4 - Least Highlight Detail/Least Mapping=5
set L1_Tuning=1

[HotFudge]

yes, my script use the same Dolby algo. You can control the tuning in the script main settings

:: choose the L1 analysis tuning... Choices are: old=0 - Most Highlight Detail/Most Mapping=1(default) - More Highlight Detail/More Mapping=2 - Balanced=3 - Less Highlight Detail/Less Mapping=4 - Least Highlight Detail/Least Mapping=5
set L1_Tuning=1

[Hellboy.]

Is possible to remove DV "Dolby Vision Profile 8 Level 6" from a Remux without encoding the movie?
Thanks.

[RESET_9999]

yes, workflow 8-1-2

[HotFudge]

Hi

When creating a ProRes 422HQ with the DoVi Script when would you change the cropping option to yes?

Cheers

[RESET_9999]

never i guess.

[HotFudge]

never i guess.

heh cheers!

[HotFudge]

Hi

I have a film where the top black bar is 281 and the bottom is 280.

How do I crop this correctly for dolby vision analysis?
If I under crop I will get wrong min values in my analysis and if I over crop I will not anaylze the whole active area?

Cheers

[RESET_9999]

use 281/280, L5 works just fine with non-even or non mod 2 crop.
in fact, when you select 2.39(not manually) in resolve you get L5 : 276/277

[HotFudge]

use 281/280, L5 works just fine with non-even or non mod 2 crop.
in fact, when you select 2.39(not manually) in resolve you get L5 : 276/277

AVSPMod show the Black bar area for my Inception clip test is Top 281 Bottom 280.

When I use Top 281 Bottom 280 using DoVi script the resulting rpu has L5 Top value of 280 and L5 bottom value of 281...(Should this not be the other way around?)

In resolve I tried to use Top 280 and Bottom 1879 in blanking but this also made the L5 values Top 280 and Bottom 281....

I could just edit the L5 values but I am unsure if the DV analyze process was based on an incorrect active area.

Where am I going wrong?

Cheers

[RESET_9999]

In the script 3-1 it works properly, I just tested it.
If you do the xml in resolve , you have to edit L5 when you inject in 1-1 because otherwise, the dovi_tool cannot know which side is 281 or 280 but the analysis itself in resolve is correct.