[DoVi_Scripts] Multi-Function Scripts for Dolby Vision processing and a lot more...

[Lucky38]

I'm not sure what you mean. You just have to create the EAE folder and put these files from Plex and EasyAudioEncoder.exe in it.
then you can control the bitrate with the script bat file
set DDP_bitrate=1024


https://i.ibb.co/c8sWD4p/Capture-d-cran-2024-07-20-094922.png

i have only this

https://ibb.co/qY90DzmT

[Kuler087]

I did mine a very long time ago and maybe plex no longer provides some files... I'll send you the files via PM tonight when I get home

[en6ads]

Just to say I tested the latest beta 8-2-6, FEL mkv source, TIFF lzw, CRF 15. Image looks great. Very impressed. Agree TIFF file size is massive, but It's great we now have a full FEL 12-bit YCbCr to P5 10-bit ICtCp transcode method. Thank you.

Do you have any suggested FEL movies to test? Examples where FEL significantly improves from the 10-bit base layer, also to test for banding / sky comparisons.

Much appreciate your scripts.

[Kuler087]

From my experience, when FEL actually makes a difference, it’s usually by improving compression (grain) and expanding brightness (https://docs.google.com/spreadsheets/d/15i0a84uiBtWiHZ5CXZZ7wygLFXwYOd84/edit?gid=1226038728#gid=1226038728). I don’t have many clear banding examples, but if I remember correctly, Wall-E shows some early in the movie, and the FEL layer mitigates it slightly.

[TR-9970X]

https://i.ibb.co/c8sWD4p/Capture-d-cran-2024-07-20-094922.png

I just thought I'd compare the Plex Server items used, between what is in my Tools folder, and the latest update:-

EAE\avcodec-58.dll" 4,144kb
EAE\avfilter-7.dll" 7,210kb
EAE\avformat-58.dll" 5,280kb
EAE\avutil-56.dll" 838kb
EAE\EasyAudioEncoder.exe" 1,952kb
EAE\libwinpthread-1.dll" 86kb
EAE\PlexTranscoder.exe" 1,529kb
EAE\swresample-3.dll" 144kb
EAE\swscale-5.dll" 751kb

====================================

Plex Media Server\avcodec-59.dll" 5,886kb
Plex Media Server\avfilter-8.dll" 5,930kb
Plex Media Server\avformat-59.dll" 3,554kb
Plex Media Server\avutil-57.dll" 1,005kb
EasyAudioEncoder.exe" 1,952kb (missing on purpose)
Plex Media Server\libwinpthread-1.dll" 81kb
Plex Media Server\Plex Transcoder.exe" 313kb
Plex\Plex Media Server\swresample-4.dll" 149kb
Plex Media Server\swscale-6.dll" 868kb

Any point in updating ???

[Kuler087]

ec3 encoding is working? then no reason to touch anything

[RRAH]

In case you or any other user needs instructions :)

Step 1: Extract DV from original_movie.mkv
Step 2: Bake FEL to ProRes using 8-2-3. This produces a prores.mov file that is essentially the HDR12 bit master. This file will be very large (upwards of 800GB sometimes!). If storage is limited, change qscale to 0 on line 168 (-qscale:v 1 --> -qscale:v 0). This should significantly reduce the size of the prores.mov, however, it may take twice as long to encode. There is no difference in quality between qscale 1 and 0.
Step 3: Rename original_movie_RPU.bin to the same file name as the prores.mov (and same path too)
Step 4: Run 3-1 on the prores.mov, using the same MDL and scene cuts as the original DV. If you did step 3 correctly, it will automatically grab the scene cuts from the original RPU. You cannot make your own scene cuts or force FBF using this method. This will produce a generated_RPU.bin
Step 5. Use 2-3 to transfer levels 12389 from generated_RPU.bin to original_movie_RPU.bin. This produces a transferred_RPU.bin
Step 6: Inject transferred_RPU.bin into original_movie.mkv using workflow 1.

Hope this helps!

Hello, can I replace step 4 with workflow 2-5 and Resolve (export EDL, import into Resolve, run DV analysis, export DV-xml, continue with workflow 2-3)?

Thank you

[Kuler087]

Yes that's what I do since Resolve is a million times faster than CM

[ralberti]

yes, DV tone mapping is useless for 99% of the movies on a 5000nits TV. It still can be useful if your TV has issues in HDR10 like my G5.

L5 + ugoos: it doesn't matter much because even if it's missing or incorrect, the ugoos output will still be accurate. So save your HDD life and avoid the editing/injecting/remuxing

Sorry, not sure if I understand this right. Does this mean that there is no advantage in converting to DV? I am converting from HDR10 and HDR10plus mkv to DV P8 with workflow 3-1 and also converting from mkv DV P7 to P8 with workflow 4-1 (this is because my plex server is not able to play DV P7).

[Kuler087]

Right, on a 5000-nits TV, there is no advantage to converting HDR10 to Dolby Vision in 3-1(or resolve).
For a native DV Blu-ray rip, some FEL titles do restore a brighter master (https://drive.google.com/drive/u/0/folders/1FS42T95TOSpoy4xtwUBIQmziCe_R_IKe), but that benefit is lost once they are converted to Profile 8. You need a device like the Ugoos Am6B+ to get that brighter DV from FEL playback.

[ralberti]

But HDR10 is supposed to be static metadata and DV is dynamic metadata. ¿Does the conversion from HDR10 to DV (3-1) not create dynamic metadata? The thing is that if If play the same file with HDR10 or with DV P8 on the same TV y see them different. I played around with the HDR10 and DV settings of the TV to get "the best" settings for each format but still I see them different, specially with movies with high contrast changes where I feel HDR10 is either too dark (clear scenes are OK, but dark scenes too dark) or too clear( clear scenes are too light or even burned and dark scenes OK ) and DV does a better job (more details in dark and light) and I supposed it is due to dynamic metadata, but may be it is my perception.

[Kuler087]

Dolby Vision is indeed dynamic, shot-by-shot metadata, and the metadata generated in profile 3-1 is also dynamic.

However, setting aside FEL-expanded brightness, Dolby Vision is fundamentally just tone-mapping metadata. When the content’s peak and average brightness already fall within the native capabilities of the display, no tone mapping is required, which means the Dolby Vision metadata has effectively no impact and the TV just displays the content as it is, just like HDR10. There may still be some gamut mapping taking place, but its effect is negligible.

A TV may still look better in Dolby Vision than in HDR10 due to superior internal processing in its Dolby Vision mode, as is the case with LG TVs but that improvement is not driven by the metadata itself.

FYI, the dovi_script workflow 7-4 can generate Dolby Vision tone-mapping comparison videos of any movie any frame, allowing you to see exactly what the metadata actually does. Many examples are available here:
https://drive.google.com/drive/folders/1g5I-z_sJmVu-SAIPNiiSlcdMiy2ka0mf

[Gatorman3385]

I'm getting this error when trying to open a "L8.Saturation.Vector.Field.PLOT" and "HUE.Vector.Field.PLOT" png. Is there something I'm missing. Windows 11.
https://imgur.com/kbLARBZ.png

[Kuler087]

This function works only with the old Trims plotter and is disabled by default. I guess you enabled it? Does the L2/L8/L3 plot works?

[Gatorman3385]

They do work. I did enable the old Trims plotter. I think I remember noticing the new plotter didn't match your L2 plots on YouTube. Am I wrong?

[Kuler087]

Strange, there’s no obvious reason the vector plot wouldn’t work then. Can you upload the RPU? It’s possible there’s a detection error, since not every CMV 4.0 RPU actually contains this metadata.

the new plotter didn't match your L2 plots on YouTube
The difference is that the old plotter dynamically rescales the data to improve visibility when the values are very low, but the underlying data being plotted is exactly the same.

[Gatorman3385]

I see. I was plotting your "Coco 2017 UHD BD_CM_Analyze_OSC" when it failed. I tried "Knock.Knock.2015.UHD-BD_P7 FEL_(retail cmv4.0 restored)" then "L8.Saturation.Vector.Field.PLOT" and "HUE.Vector.Field.PLOT" plotted just fine. Thank you.

[ralberti]

Hi, could some one tell me please the steps and or commands to reduce the size of an mkv file with HDR (either HDR10, HDR10+ or DV) keeping HDR information? As far as I have seen I have to extract HDR Information, reencode video and inject back HDR information but I have no idea how to do that :(.

[Kuler087]

Workflow 8-2-1

[Kuler087]

3.1.0 Released:

https://github.com/R3S3t9999/DoVi_Scripts/releases

[TR-9970X]

3.1.0 Released:

https://github.com/R3S3t9999/DoVi_Scripts/releases

What a nice New Year surprise :)

:thanks:

PS:-

I now have a couple of questions about the new stuff...

Not sure if you covered this in the new tutorial.

With Workflow 6-8, you show "cavernize folder", and that is a hyperlink to where to download it, so the question is, which version to download, the 2.0.2 for Windows, or the portable type.

Then, how to add it to the tools folder ?? (create a new folder called "cavernize", and add the portable version in there ????)

Then workflow 9-9, where do you place that "hevc_hdr_editor.exe" ??

Apologies.

[Kuler087]

With Workflow 6-8, you show "cavernize folder", and that is a hyperlink to where to download it, so the question is, which version to download, the 2.0.2 for Windows, or the portable type.

Then, how to add it to the tools folder ?? (create a new folder called "cavernize", and add the portable version in there ????)

I don't remember but I think the portable version will work, just place the cavernize folder in the script tools folder.

Then workflow 9-9, where do you place that "hevc_hdr_editor.exe" ??

In the tools folder

[TR-9970X]

I don't remember but I think the portable version will work, just place the cavernize folder in the script tools folder.


In the tools folder

:thanks: others may also need to know this. :)

[Kuler087]

PS: Right after the 3.1.0 release, I had an idea to auto-detect (very quickly) when DV FEL changes the BL brightness. This new feature will therefore remain beta-only (https://drive.google.com/file/d/128gq8aDUTKA_aT7SQsM9dkjA1EP1sosR/view?usp=drive_link) for a while until I have enough updates to release a regular version.

This new option was added to workflow 4-1 (P7 to P8), workflow 1 (editor), and batch info workflow 2-7.

The option checks for brightness differences and, based on my testing, it appears very accurate. However, there may be false positives due to BL artifacts or heavier grain compression, and it may miss some expanded-brightness releases since it only checks two random frames.

There are two categories:

- “FEL expanded brightness” (99.9% guaranteed): detects any difference greater than 60 nits peak or 2nits average

- “FEL possibly expanded brightness”: detects any difference greater than 15 nits

This feature is very fast and only takes a few seconds to run. It can be disabled at line 335 by setting:
set check.brighter.fel=NO

2-7 results of my whole library:

Profile 7 FEL total: 542 (36 percent)
168 titles expand the HDR10 BL brightness via FEL
67 titles possibly expand the HDR10 BL brightness via FEL
Profile 7 MEL total: 26 (1 percent)
Profile 8 total: 996 (66 percent)
Profile 5 total: 97 (6 percent)
CMv4.0 total: 985 (66 percent)
CMv2.9 total: 507 (33 percent)
DoVi Files Total: 1492

https://i.ibb.co/3ydCLMkB/1.gif

https://i.ibb.co/XrTFCFDR/2.gif

[KoVaR]

Hi
Most of the things I try to do end in error:
"GPU decoder: Failed to create video decoder [801]"

What is that ?

Google has 1 result for this error, some Russian forum post from 2019

[Kuler087]

No idea. Which workflow?

[TR-9970X]

Hi
Most of the things I try to do end in error:
"GPU decoder: Failed to create video decoder [801]"

What is that ?

Google has 1 result for this error, some Russian forum post from 2019

What brand & model is your GPU ??

Maybe it's too old.

The error code 801 indicates that the GPU is not supported for the resolution being processed.
This can occur due to various reasons, such as the GPU not being able to handle the video codec or the video file format not being supported by the GPU.
To resolve this issue, you may need to check the video codec and format compatibility with your GPU, consider using software decoding, or update your GPU drivers to the latest version.

[KoVaR]

GeForce 970

How do I foce software decoding ?

[sk2316]

How to Converting a Dolby Vision Profile 8 file to Profile 7?

[en6ads]

PS: Right after the 3.1.0 release, I had an idea to auto-detect (very quickly) when DV FEL changes the BL brightness. This new feature will therefore remain beta-only (https://drive.google.com/file/d/128gq8aDUTKA_aT7SQsM9dkjA1EP1sosR/view?usp=drive_link) for a while until I have enough updates to release a regular version.

This new option was added to workflow 4-1 (P7 to P8), workflow 1 (editor), and batch info workflow 2-7.

The option checks for brightness differences and, based on my testing, it appears very accurate. However, there may be false positives due to BL artifacts or heavier grain compression, and it may miss some expanded-brightness releases since it only checks two random frames.

There are two categories:

- “FEL expanded brightness” (99.9% guaranteed): detects any difference greater than 60 nits peak or 2nits average

- “FEL possibly expanded brightness”: detects any difference greater than 15 nits

This feature is very fast and only takes a few seconds to run. It can be disabled at line 335 by setting:
set check.brighter.fel=NO

2-7 results of my whole library:

Profile 7 FEL total: 542 (36 percent)
168 titles expand the HDR10 BL brightness via FEL
67 titles possibly expand the HDR10 BL brightness via FEL
Profile 7 MEL total: 26 (1 percent)
Profile 8 total: 996 (66 percent)
Profile 5 total: 97 (6 percent)
CMv4.0 total: 985 (66 percent)
CMv2.9 total: 507 (33 percent)
DoVi Files Total: 1492


Thanks for doing this.

Someone else did something similar recently that runs in docker / linux / mac / Windows via wsl:

https://docs.doviconvert.com/before-you-start#understanding-the-enhancement-layer
https://github.com/cryptochrome/dovi_convert

He links your github page and spreadsheet.
I was going to try it out but I have been traveling over the holidays.

Edit: his version scans either 10 frames, or all frames:
https://docs.doviconvert.com/essentials/scanning#how-inspect-differs-from-scan
Interesting how you can get over 99% with just 2 frames.
Are these 2 chosen at random?

[Kuler087]

How to Converting a Dolby Vision Profile 8 file to Profile 7?

In the next Beta, there will be a way to encode profile 7 FEL/MEL but it will require DEE. It will even have the option to tone map the base layer to lower brightness like the studios often do.

profile 7 encoding mode. default =( no_mapping_with_fel ) Choices are: 'no_mapping_with_mel', 'no_mapping_with_fel', 'map_to_1000_nits_with_fel', 'map_to_1000_nits_with_mel', 'map_to_600_nits_with_fel'
set usecase=no_mapping_with_fel


Interesting how you can get over 99% with just 2 frames.
Are these 2 chosen at random?
Yes,random, and most of the time the script detects the difference on the first frame and doesn’t even need to check the second one, which is why the process is very fast.

In most cases where FEL changes the BL brightness, it isn’t limited to the highlights. The average brightness of the entire image is lifted, and this happens across nearly all shots. Because of that, detecting a small 2–3 nits difference in average brightness is quite straightforward in those situations.

[TR-9970X]

Just a trivial thing, but there are later versions of Lavfilters now, than the one you have linked to on your Github page.

https://files.1f0.de/lavf/nightly/LAVFilters-0.80.0-14.exe

[Kuler087]

Thanks, I prefer sharing only the versions I tested, but the latest version should be ok.

[sk2316]

In the next Beta, there will be a way to encode profile 7 FEL/MEL but it will require DEE. It will even have the option to tone map the base layer to lower brightness like the studios often do.


Good Job:)

[Colek]

Hello,

I wanted to make an 10-bit x265 encode of UHD disc - I've used the 8-2-3 to bake FEL into lossless ProRes file, but honestly I am unsure how to proceed now with this file I've received in the end, this is my first time dabbling in DV and encoding this stuff.

I've loaded the ProRes file into AviSynth like this:

LWLibavVideoSource("G:\DoVi_Scripts\Movie_prores.422_DV-FEL_to_HDR10.mov")
Crop(0, 42, -0, -42)
ConvertBits(10, dither=1)
ConvertToYUV420()


I assume this will get me HDR10 file - but what parameters do I provide to x265 for master-display and max-cll? How do I proceed with injecting (and converting) RPU now to Profile 8?

I want to work on lossless files to minimize any potential quality loss.

Thank you!

[Kuler087]

You dont have to encode the file to ProRes.
Just use workflow 8-2-1 and the script will do everything for you (bake fel and encode HDR10 x265 with static metadata+P8).

[Colek]

You dont have to encode the file to ProRes.
Just use workflow 8-2-1 and the script will do everything for you (bake fel and encode HDR10 x265 with static metadata+P8).

I see it now, didn't notice before, that you can do 2-pass encode with your script. Thank you, I'll try it :)

[Rootdown4594]

Tell me if this workflow is flawed, please...

Using 8-2-1 to bake FEL, I set a crf of 0 and an encode speed of faster. Gives me about a 100gig file.

I do this because I ultimately want to encode that baked file to AV1 and I want to give the AV1 encoder (AV1-SVT-HDR 10bit) the best possible quality file.

Is there anything wrong with this approach?

[Kuler087]

Tell me if this workflow is flawed, please...

Using 8-2-1 to bake FEL, I set a crf of 0 and an encode speed of faster. Gives me about a 100gig file.

I do this because I ultimately want to encode that baked file to AV1 and I want to give the AV1 encoder (AV1-SVT-HDR 10bit) the best possible quality file.

Is there anything wrong with this approach?



Yes, don’t do that.

dovi_script does not support AV1 (I could add support eventually), but you can still use it to generate the FEL Baker AviSynth script and then encode to AV1 using a GUI like StaxRip.

1- Run 8-2-1 and cancel the process as soon as the encode starts.
2- Go to the dovi_script TEMP folder and import ''script.avs'' into StaxRip. From there, you can bake the FEL and encode directly to AV1.

*It’s also possible that StaxRip now supports FEL directly, but I haven’t followed its development recently.

[Kuler087]

(Latest beta) (https://drive.google.com/file/d/128gq8aDUTKA_aT7SQsM9dkjA1EP1sosR/view?usp=drive_link) Pretty cool new addition to workflow 7-2 for users who prefer exporting SDR tone-mapped screenshots.

The workflow now includes an option to use the Dolby Vision 100-nits trim pass. Because the colorist creates these trim passes manually and they are delivered for the official SDR releases, the HDR-to-SDR mapping is far more accurate than any generic tone-mapping algorithm like Libplacebo, madVR, which tend to darken too much to preserve highlight details. This method will match the official SDR release most of the time, at least for titles mastered/delivered in cmv4.0, which is the case for the majority of movies today.

Example comparison:
DV trim pass vs Libplacebo vs studio original SDR release
https://slow.pics/c/2ZrwYO9i

more examples:
Babe 1995 L8 trim: https://slow.pics/c/KBk1g3gg
War of the Worlds L2 trim: https://slow.pics/c/sTwj3ifW


Notes:
- The process is slightly slower than using libplacebo, but still fast enough to be practical and the mapping quality 100% worth it.
Pipeline DV input: Index video → extract RPU → ProRes (qscale=0) → cm_offline → ffmpeg lossless RGB 16-bit screenshots
Pipeline HDR10 input: Index video → ProRes (qscale=0) → cm_analyze → cm_offline → ffmpeg lossless RGB 16-bit screenshots

- If the input does not contain Dolby Vision, the script falls back to cm_analyse to generate Dolby’s default SDR tone mapping.

-This new function made me discover a bug in workflow 8-2-4 trim pass delivery, the output has slightly raised black in the latest dovi_scripts.

EDIT: I forgot to say, you need to add these two files in your tool folder.
https://drive.google.com/drive/folders/1msSCOHJm4kzemykWjL0kmwJ4JGvDZ4px?usp=drive_link

https://i.ibb.co/TqxHhk3j/Windows-Terminal-SRk-MK0-AFeu.gif

[Rootdown4594]

Yes, don’t do that.

dovi_script does not support AV1 (I could add support eventually), but you can still use it to generate the FEL Baker AviSynth script and then encode to AV1 using a GUI like StaxRip.

1- Run 8-2-1 and cancel the process as soon as the encode starts.
2- Go to the dovi_script TEMP folder and import ''script.avs'' into StaxRip. From there, you can bake the FEL and encode directly to AV1.

*It’s also possible that StaxRip now supports FEL directly, but I haven’t followed its development recently.


I get this error when I drag and dop the scipt.avs file into Straxrip and select automatic:

Script Error

Python exception: name 'core' is not defined

Traceback (most recent call last):
File "src/cython/vapoursynth.pyx", line 3393, in vapoursynth._vpy_evaluate
File "src/cython/vapoursynth.pyx", line 3394, in vapoursynth._vpy_evaluate
File "D:\DoVi\DoVi.Scripts\temp.folder31\script.avs_temp\script_source.vpy", line 1, in <module>
clip = core.avisource.AVISource(r"D:\DoVi\DoVi.Scripts\temp.folder31\script.avs")
^^^^
NameError: name 'core' is not defined

[Kuler087]

You have to change your staxrip project to avisynth.

https://i.ibb.co/ycXqhZ5M/Stax-Rip-VG378fcn-Op.png

[Rootdown4594]

Thank you

[dkangel]

hello
i'm trying to inject your L5 offsets data in the rpu of interstellar.
copy the text into a samename.json file
everything good except at the end, i've this message

01/13/2026/20:29:54.198000000 metafier: ERROR Illegal Per-frame Level Data: 5
Parsing RPU file...

Summary:
Frames: 243218
Profile: 8
DM version: 1 (CM v2.9)
Scene/shot count: 2358
RPU mastering display: 0.0050/4000 nits
RPU content light level (L1): MaxCLL: 3056.40 nits, MaxFALL: 588.84 nits
L6 metadata: Mastering display: 0.0050/4000 nits. MaxCLL: 0 nits, MaxFALL: 0 nits
L5 offsets: top=0..280, bottom=0..280, left=0, right=0
L2 trims: 100 nits

do you know what it is ? ERROR illegal per frame

thank you

[Kuler087]

It means that some L5 change happens on a frame that is not a scene cut.
Is this the json edit you've used? I just checked my RPU, and everything is fine.

https://i.ibb.co/rGSD3Hn9/v8-Jl-QFNy-Iw.png

{
"mode": 0,
"active_area": {
"presets": [
{
"id": 0,
"left": 0,
"right": 0,
"top": 280,
"bottom": 281
},
{
"id": 1,
"left": 0,
"right": 0,
"top": 0,
"bottom": 0
}
],
"edits": {
"0-2117": 0,
"2118-2589": 1,
"2590-4429": 0,
"4430-4588": 1,
"4589-4657": 0,
"4658-4736": 1,
"4737-9349": 0,
"9350-12547": 1,
"12548-20071": 0,
"20072-20228": 1,
"20229-26005": 0,
"26006-27624": 1,
"27625-29207": 0,
"29208-29379": 1,
"29380-32410": 0,
"32411-33439": 1,
"33440-52177": 0,
"52178-52270": 1,
"52271-59929": 0,
"59930-61037": 1,
"61038-61098": 0,
"61099-61199": 1,
"61200-61306": 0,
"61307-61411": 1,
"61412-62296": 0,
"62297-62454": 1,
"62455-62779": 0,
"62780-62895": 1,
"62896-62984": 0,
"62985-63072": 1,
"63073-63310": 0,
"63311-63407": 1,
"63408-65118": 0,
"65119-65269": 1,
"65270-65475": 0,
"65476-65745": 1,
"65746-65790": 0,
"65791-65886": 1,
"65887-65951": 0,
"65952-66059": 1,
"66060-66179": 0,
"66180-66318": 1,
"66319-66770": 0,
"66771-66917": 1,
"66918-67974": 0,
"67975-68163": 1,
"68164-69114": 0,
"69115-69530": 1,
"69531-69652": 0,
"69653-69843": 1,
"69844-70575": 0,
"70576-70774": 1,
"70775-71548": 0,
"71549-71652": 1,
"71653-71835": 0,
"71836-71865": 1,
"71866-71945": 0,
"71946-72079": 1,
"72080-72124": 0,
"72125-72403": 1,
"72404-76948": 0,
"76949-77258": 1,
"77259-78833": 0,
"78834-79911": 1,
"79912-82014": 0,
"82015-82265": 1,
"82266-82856": 0,
"82857-82985": 1,
"82986-83345": 0,
"83346-83474": 1,
"83475-84866": 0,
"84867-84986": 1,
"84987-85142": 0,
"85143-85255": 1,
"85256-85479": 0,
"85480-85719": 1,
"85720-85877": 0,
"85878-86018": 1,
"86019-86094": 0,
"86095-86191": 1,
"86192-86423": 0,
"86424-86551": 1,
"86552-86615": 0,
"86616-86700": 1,
"86701-86840": 0,
"86841-86893": 1,
"86894-87014": 0,
"87015-87113": 1,
"87114-87457": 0,
"87458-87556": 1,
"87557-88210": 0,
"88211-88456": 1,
"88457-88841": 0,
"88842-89377": 1,
"89378-94039": 0,
"94040-94168": 1,
"94169-94328": 0,
"94329-94490": 1,
"94491-94574": 0,
"94575-94638": 1,
"94639-95152": 0,
"95153-95210": 1,
"95211-95399": 0,
"95400-95516": 1,
"95517-96608": 0,
"96609-96705": 1,
"96706-96820": 0,
"96821-96911": 1,
"96912-96953": 0,
"96954-97019": 1,
"97020-97372": 0,
"97373-97397": 1,
"97398-97441": 0,
"97442-97491": 1,
"97492-97540": 0,
"97541-97620": 1,
"97621-97663": 0,
"97664-97727": 1,
"97728-97770": 0,
"97771-97869": 1,
"97870-97918": 0,
"97919-98070": 1,
"98071-98112": 0,
"98113-98203": 1,
"98204-98607": 0,
"98608-105200": 1,
"105201-110488": 0,
"110489-110578": 1,
"110579-110783": 0,
"110784-110833": 1,
"110834-110870": 0,
"110871-110910": 1,
"110911-110956": 0,
"110957-111007": 1,
"111008-111039": 0,
"111040-111106": 1,
"111107-111135": 0,
"111136-111158": 1,
"111159-111181": 0,
"111182-111443": 1,
"111444-128219": 0,
"128220-129099": 1,
"129100-130228": 0,
"130229-130512": 1,
"130513-135990": 0,
"135991-137775": 1,
"137776-142345": 0,
"142346-142941": 1,
"142942-150171": 0,
"150172-151096": 1,
"151097-154631": 0,
"154632-156627": 1,
"156628-156817": 0,
"156818-157207": 1,
"157208-157847": 0,
"157848-159661": 1,
"159662-159951": 0,
"159952-161259": 1,
"161260-161479": 0,
"161480-162074": 1,
"162075-162476": 0,
"162477-163947": 1,
"163948-165059": 0,
"165060-170806": 1,
"170807-170925": 0,
"170926-175489": 1,
"175490-175689": 0,
"175690-176841": 1,
"176842-177228": 0,
"177229-200823": 1,
"200824-201104": 0,
"201105-204170": 1,
"204171-204414": 0,
"204415-204477": 1,
"204478-204553": 0,
"204554-204700": 1,
"204701-204873": 0,
"204874-205839": 1,
"205840-206259": 0,
"206260-206443": 1,
"206444-206504": 0,
"206505-206829": 1,
"206830-206849": 0,
"206850-206892": 1,
"206893-207033": 0,
"207034-207082": 1,
"207083-207172": 0,
"207173-207243": 1,
"207244-207302": 0,
"207303-207343": 1,
"207344-207380": 0,
"207381-207621": 1,
"207622-207637": 0,
"207638-207790": 1,
"207791-207882": 0,
"207883-208172": 1,
"208173-208290": 0,
"208291-208976": 1,
"208977-209112": 0,
"209113-209376": 1,
"209377-209558": 0,
"209559-210406": 1,
"210407-210990": 0,
"210991-212925": 1,
"212926-213076": 0,
"213077-214127": 1,
"214128-214235": 0,
"214236-214672": 1,
"214673-214762": 0,
"214763-214840": 1,
"214841-215120": 0,
"215121-215403": 1,
"215404-215498": 0,
"215499-215613": 1,
"215614-215687": 0,
"215688-217571": 1,
"217572-217634": 0,
"217635-217691": 1,
"217692-217781": 0,
"217782-217977": 1,
"217978-218091": 0,
"218092-218575": 1,
"218576-218628": 0,
"218629-218697": 1,
"218698-218858": 0,
"218859-222520": 1,
"222521-222594": 0,
"222595-222722": 1,
"222723-222737": 0,
"222738-223265": 1,
"223266-225789": 0,
"225790-227094": 1,
"227095-229653": 0,
"229654-229822": 1,
"229823-233190": 0,
"233191-235847": 1,
"235848-243202": 0
}
}
}

[dkangel]

it's this one

https://justpaste.it/bhv73 from your excell sheet

my last frame is 243218

EDIT : ok if i change the last line with "235848-243217": 0

[Kuler087]

Yeah, my RPU frame count comes from the WEB-DL before it’s injected into the Blu-ray remux, which automatically duplicates frames at the end to match the video length.

Other than that, there’s no difference with that JSON, so you shouldn’t be getting a Metafier error. Are you using an older version of DoVi_meta.exe? I know that a while ago, it had an issue when converting RPU to XML, which caused this kind of error in Metafier.

[dkangel]

no everything is up to date, dovi script 3.1.0 with last tools directory

it seems ok now with changing the number of the last frame

if i put 243217

01/13/2026/21:43:58.811000000 metafier: INFO No issues found in the Metadata
Parsing RPU file...

Summary:
Frames: 243218
Profile: 8
DM version: 1 (CM v2.9)
Scene/shot count: 2335
RPU mastering display: 0.0050/4000 nits
RPU content light level (L1): MaxCLL: 3056.40 nits, MaxFALL: 588.84 nits
L6 metadata: Mastering display: 0.0050/4000 nits. MaxCLL: 0 nits, MaxFALL: 0 nits
L5 offsets: top=0..280, bottom=0..280, left=0, right=0
L2 trims: 100 nits

[Kuler087]

This movie trim pass is useless, FYI. You'll get better quality if you generate new CMV4.0 metadata.

[dkangel]

ok i will do that with resolve, is it possible to import your l5 metadata in resolve to avoid all the cuting stuff ?