H.264 Patcher and BD-Tools

[videofan3d]

(Project is mirrored on site videofan3d (https://sites.google.com/site/videofan3d/))

BDTools is small set of simple command-line (CLI) utilities for displaying the structure of .mpls playlist and .clpi clip information files which are part of Bluray Disc navigation.

MPLS2JSON.exe
– converts .mpls playlist, .clpi clip information file or .bdjo BD-J object file into human&machine readable JSON file, or
– converts .json file into binary .mpls, .clpi or .bdjo file. This reverse operation can be used for modification of the content of the .mpls/.clip/.bdjo file.

MPLSDump.exe
– displays simplified list of all .m2ts files referred in given .mpls playlist

h264Modify.exe
- allows changes of some parameters of H.264 elementary video-stream.
Modifies frame-rate and sample-aspect-ratio of the H.264 encoded video without re-encoding!
- allows insertion or deletion of frame packing arrangement SEI messages (side-by-side, top-above-bottom, etc.)

h264Offset3D.exe
- tool for offset-metadata manipulation (subtitles in Bluray-3D)
- allows extract, delete, replace, insert offset-metadata from/into dependent view elementary stream of Bluray-3D.

AVCHDInfo
- displays recording date/time of each videoframe of AVCHD.MTS file extracted from vendor-specific information (if available)

BDTools executables and documentation can be downloaded from:
2019-06-05: BDTools version 2.31 (https://drive.google.com/file/d/12bIXjPXPjirgNf4ahpEptSycaM6JEZLi) - bugfix in MPLS2JSON - 3D subtitle depth plane extraction
2017-08-24: BDTools version 2.30 (https://drive.google.com/file/d/0BymRNDHq74DEU0dfTHdoUFlnRjA) - added AVCHDInfo
2017-08-02: BDTools version 2.20 (https://drive.google.com/file/d/0BymRNDHq74DEZ29rN2hkemRNWUU) - added h264Offset3D
2017-07-24: BDTools version 2.10 (https://drive.google.com/file/d/0BymRNDHq74DETFMwNHAtWEdMZ0E)
2017-06-16: BDTools version 2.00 (https://drive.google.com/file/d/0BymRNDHq74DEbU0yWXRyaW1ZQjA)
2017-05-10: BDTools version 1.10 (https://drive.google.com/file/d/0BymRNDHq74DENFVDX0RaMHNXaDQ)
2017-04-29: BDTools version 1.00 (https://drive.google.com/file/d/0BymRNDHq74DEM0JrckV2YmJ1clE)

[Sharc]

Very nice, thanks! :)

Stupid question: How to prevent the windows from immediately closing after displaying the result?

I put the link to the .exe into the "Send to" folder. It works but the window immediately closes.

[sneaker_ger]

Put a batch file into the send to folder. First line to call the program (with %1 parameter), second line "pause".

@videofan3d
Thanks for sharing.

[Sharc]

Odd, the called window (with the output) still closes .....

[sneaker_ger]

@videofan3d
Why does h264patch ask for admin rights?

[videofan3d]

@videofan3d
Why does h264patch ask for admin rights?

Rename the executable to whatever else than "*patch*.exe" and it will work.

Crazy Microsoft! If the exe-filename contains the word "patch" then it is considered as dangerous and requires administrator access... WTF!!!


I will rename it in next build to something else.

Hereby I'm opening discussion/your suggestions for new name recommendations :) !!!

[sneaker_ger]

Thx.

@Sharc
Try again after renaming as explained by videofan3d.

[Sharc]

Yes, renaming the h264Patch.exe solved the issue. Thanks.

[tebasuna51]

Hereby I'm opening discussion/your suggestions for new name recommendations :) !!!

Maybe h264Mody (short of h264Modify).

Other metadata than can be changed is the Level, like CoolSoft H264 Level Editor (http://coolsoft.altervista.org/en/h264leveleditor).

[videofan3d]

Maybe h264Mody (short of h264Modify).

Other metadata than can be changed is the Level, like CoolSoft H264 Level Editor (http://coolsoft.altervista.org/en/h264leveleditor).

I will change the name to h264Modify (as I prefer regular, serious names).
I will add also change of profile_idc, constraint_setN_flag and level_idc.

Next release will be around 10 of May.

[videofan3d]

BDTools 1.10 released:

h264Patch
- renamed to h264Modify (due to Windows antivirus protection)
- added options -profile_idc, -constraint_flags, -level_idc
- renamed options -frame_rate, -sar_idc (for consistency with BD-documentation)

MPLS2JSON
- added experimental support for UHD Bluray (4K)

[DenisRodman]

videofan3d, Many thanks.

MPLS2JSON.exe -a great utility. Can you add conversion support also for file *.clpi ?

And what do you think about adding support for the new version of 300? Sample ver.300 (https://www.sendspace.com/file/osqvog)

[videofan3d]

videofan3d, Many thanks.

MPLS2JSON.exe -a great utility. Can you add conversion support also for file *.clpi ?

And what do you think about adding support for the new version of 300? Sample ver.300 (https://www.sendspace.com/file/osqvog)

re *.clip ... maybe in the future (for BD it is only matter of coding work)

Re: version 300 (UHD) ... some experimental support is already there - try it.
(To finalize it I need to get the complete UHD-BD specification.)

[DenisRodman]

(... I need to get the complete UHD-BD specification.)
Maybe these documents will help BD-ROM Version 3.0 (https://www.sendspace.com/file/edyggg) I found them on the internet.

I checked your new experimental version for all three types of files *.mpls, MovieObject.bdmv, index.bdmv Conversion and reverse :thanks:

[mkver]

Thanks for the tool. May I ask for another feature in h264modify? Can you allow it to change the SPS and PPS id-values (both of the actual SPS and PPS and of the frames using them)? This would improve the ability to join different videos (with incompatible SPS and PPS) without reencoding, because one would only need to set the SPS and PPS id to different values before joining them. (Yes, I know that this is unnecessary for Blu-ray authoring because on a Blu-ray the SPS and PPS are simply repeated and can therefore be easily overwritten, but for mp4 and mkv it would allow to use the more space-efficient "out-of-band" way of transmitting the SPS and PPS.)

[videofan3d]

Thanks for the tool. May I ask for another feature in h264modify? Can you allow it to change the SPS and PPS id-values (both of the actual SPS and PPS and of the frames using them)? This would improve the ability to join different videos (with incompatible SPS and PPS) without reencoding, because one would only need to set the SPS and PPS id to different values before joining them. (Yes, I know that this is unnecessary for Blu-ray authoring because on a Blu-ray the SPS and PPS are simply repeated and can therefore be easily overwritten, but for mp4 and mkv it would allow to use the more space-efficient "out-of-band" way of transmitting the SPS and PPS.)

Hi,

I'm working now on MPLS2JSON for .clpi files (requested somewhere above).
Once finished I will check also if SPS/PPS matter is somehow easily doable.

[hubblec4]

Hi videofan3d

The MPLS2JSON is a nice tool.
I have made a small test and found a small issue.

I copyed a mpls file to the BD-Tools folder and use this commandline

F:\Download\BDTools_1.10\MPLS2JSON.exe 00001.mpls
Failed to open 00001.mpls
failed signature match:
found ""
expected "MPLS"
or "MOBJ"
or "INDX"
Parse failed: 00001.mpls

Then I entered a full path to the MPLS and it works.
F:\Download\BDTools_1.10\MPLS2JSON.exe F:\Download\BDTools_1.10\00001.mpls

[videofan3d]

Hi videofan3d

The MPLS2JSON is a nice tool.
I have made a small test and found a small issue.

I copyed a mpls file to the BD-Tools folder and use this commandline

F:\Download\BDTools_1.10\MPLS2JSON.exe 00001.mpls
Failed to open 00001.mpls
failed signature match:
found ""
expected "MPLS"
or "MOBJ"
or "INDX"
Parse failed: 00001.mpls

Then I entered a full path to the MPLS and it works.
F:\Download\BDTools_1.10\MPLS2JSON.exe F:\Download\BDTools_1.10\00001.mpls

Actually, what was your "current" directory when you performed the first command?
If it was NOT "F:\Download\BDTools_1.10", then it is logical that process could not find the first parameter.
MPLS2JSON doesn't change current directory, this is standard behavior of all command line programs in any
operation system (Unix, Windows DOS-prompt, Linux, ...)
You have specified full-path-qualification for executable program, thus operating system found it. But then executable MPLS2JSON tried to find 00001.mpls in current (and unchanged) directory, but it was not there. So it failed. In the second case you fully qualified the 00001.mpls file, thus process found it.

I will change the error message to explain more precisely the cause, but behavior itself is according OS standards.

[hubblec4]

Thanks for these explanation.

I had used the console (windows 7) to enter the command line. With a file.bat it works all fine.

[videofan3d]

BDTools 2.00 released:

MPLS2JSON - added support for .clpi clip information files


Please:
I don't have so many Blu-ray discs to check thoroughly all variants of .mpls and .clpi.
If you face some discrepancies or suspicious behavior, please provide me with relevant input file (.mpls/.clpi - both these are quite small to be send or attached) and I can fix it.

[Dion]

Thank you for this tool. I have been looking for a sar changer tool for awhile.

[DenisRodman]

Hi videofan3d, thanks for CLPI.

I checked the conversion of CLPI for new discs The.Smurfs.2.2013.COMPLETE.UHD.BLURAY and Inferno.2016.COMPLETE.UHD.BLURAY all is well :thanks: .

Can you add support for the conversion also for the *.bdjo file too?
Also for versions 200 and 300.

[hubblec4]

Thanks for your work.

[videofan3d]

Thanks for the tool. May I ask for another feature in h264modify? Can you allow it to change the SPS and PPS id-values (both of the actual SPS and PPS and of the frames using them)? This would improve the ability to join different videos (with incompatible SPS and PPS) without reencoding, because one would only need to set the SPS and PPS id to different values before joining them. (Yes, I know that this is unnecessary for Blu-ray authoring because on a Blu-ray the SPS and PPS are simply repeated and can therefore be easily overwritten, but for mp4 and mkv it would allow to use the more space-efficient "out-of-band" way of transmitting the SPS and PPS.)

Hi,

could you please describe me in more detail (example) what do you request here?
And preferably also testing scenario for these cases, how can I emulate it...

[videofan3d]

BDTools 2.10 released:

MPLS2JSON - added support for .bdjo BD-J object files and sound.bdmv (extracts also its WAVE content)

[DenisRodman]

videofan3d, That's great!

Many thanks :thanks: .

[videofan3d]

BDTools 2.20 released:

h264Offset3D
- new tool for offset-metadata manipulation (subtitles in Bluray-3D) added
- allows extract, delete, replace, insert offset-metadata from/into dependent view elementary stream of Bluray-3D.

[r0lZ]

Yeah! Thanks, especially for the new h264Offset3D !

Suggestion: Can you add the possibility to modify (or create?) the frame-packing value with h264Modify ? That value is used by many TVs and hardware players to switch automatically to the right 3D mode (SBS, T&B, Frame Alternate...) It is very important, especially if the player ignores the stereo-mode of the MKV header. The frame-packing can be set with x264's --frame-packing argument, but currently if you forgot it or if you use a wrong value, you're stuck, and you have to re-encode the whole video stream.

[videofan3d]

Yeah! Thanks, especially for the new h264Offset3D !

Suggestion: Can you add the possibility to modify (or create?) the frame-packing value with h264Modify ? That value is used by many TVs and hardware players to switch automatically to the right 3D mode (SBS, T&B, Frame Alternate...) It is very important, especially if the player ignores the stereo-mode of the MKV header. The frame-packing can be set with x264's --frame-packing argument, but currently if you forgot it or if you use a wrong value, you're stuck, and you have to re-encode the whole video stream.

Hi,

regarding frame-packing.
I'm worried it is possible to add it (correctly).

Frame-packing is inserted as a SEI message, it is not just a flag. But main point is that it has also couple of parameters which are part of encoding process (and has impact on x264 encoding: you can easily verify that output files created with and w/o --frame_packing N are from image perspective different)!
Thus after such modification the image quality would not correlate with frame_packing_arrangement SEI message.
This would not be logical to make such hack.
(And beside this I don't know these parameters so I likely cannot insert correct frame_packing_arrangement SEI message).

I will check it later once more in detail if it is doable....

[sneaker_ger]

Most complicated is --frame-packing 5. I think you can get a good idea about potential problems from reading the x264 git log: http://git.videolan.org/?p=x264%2Fx264-sandbox.git&a=search&h=HEAD&st=commit&s=frame-packing
There you will also read about the optimizations x264 does for compression and compatibility.

[r0lZ]

Yes, as far as I know, only frame-packing 5 (Frame Sequential) has an impact on the way the video is encoded (because the motion vectors are computed with the even or odd frames only). And even in that case, I'm not sure it has an impact on the way the video is decoded at playback time. But perhaps I'm wrong and other frame packing methods are also optimized. It is perhaps possible, for example, to encode only once the parts of the images that are identical in the two views of a movie encoded in half-SBS, but I don't think it's the case currently.

But I didn't know that the frame-packing is stored as SEI messages, and that's obviously a good argument for not implementing it in h264Modify. Thanks anyway for what h264Modify can already do.

[videofan3d]

Yes, as far as I know, only frame-packing 5 (Frame Sequential) has an impact on the way the video is encoded (because the motion vectors are computed with the even or odd frames only). And even in that case, I'm not sure it has an impact on the way the video is decoded at playback time. But perhaps I'm wrong and other frame packing methods are also optimized. It is perhaps possible, for example, to encode only once the parts of the images that are identical in the two views of a movie encoded in half-SBS, but I don't think it's the case currently.

But I didn't know that the frame-packing is stored as SEI messages, and that's obviously a good argument for not implementing it in h264Modify. Thanks anyway for what h264Modify can already do.

I checked the format of particular SEI message - with certain assumptions it will be possible to add it (indeed only for accidentally omitted --frame-packing during encoding). I will work on this for next release...

[sneaker_ger]

indeed only for accidentally omitted --frame-packing during encoding)
What does that mean?

[videofan3d]

What does that mean?

as R0lZ said: "... if you forgot it or if you use a wrong value, you're stuck..." :)

[r0lZ]

I checked the format of particular SEI message - with certain assumptions it will be possible to add it (indeed only for accidentally omitted --frame-packing during encoding). I will work on this for next release...
Yeah! Thanks!

I suppose that the SEI message is regularly repeated during the whole movie. Right? I ask this because my current method to add the frame-packing when it is missing is simply to encode 1 second of black with the correct option, and merge the original movie after that "intro" with MkvMerge. My Samsung TV recognizes the 3D mode correctly when the movie starts and doesn't lose it during the playback, but I have the impression that my method is not correct and may not work with all players.

Anyway, we will have soon a better way to do it. :-)

[sneaker_ger]

Yes, I think it must at least be repeated for every coded video sequence (GOP). Frame-packing 5 needs SEI for every view sequence.

[DenisRodman]

Hi videofan3d,

Can you add support for the conversion for the id.bdmv?
File path CERTIFICATE\id.bdmv
Also for versions 200 and 300.

[videofan3d]

Hi videofan3d,

Can you add support for the conversion for the id.bdmv?
File path CERTIFICATE\id.bdmv
Also for versions 200 and 300.

Yes, I will add it.


Please note, that I work only with specification for version 200 (from 2011).

I didn't find official BD specification 300 yet - so I can only guess and hope that it is similar and adds only minor backward-compatible extensions (i.e. uses reserved fields).

[DenisRodman]

videofan3d,

I understand you perfectly. It is difficult for you to work without having the necessary documents. Every new version of you I check on the disks of the UHD version 300. Testing the result pleased. Considering the fact that new parameters have been added in CLPI such as HDR and others, your program very competently conversion the code and reverse. Therefore, we all understand how you work. At you it turns out all.

Good luck and let's hope for the best :thanks: .

[videofan3d]

Yes, I think it must at least be repeated for every coded video sequence (GOP). Frame-packing 5 needs SEI for every view sequence.

"frame_packing_arrangement" SEI messages can appear anytime it is needed to inform decoder about change of 3D format. They should appear in general after each Picture Parameter Set (PPS) NAL unit, and in case of mode "frame alternation" also before each frame slice NAL unit (to indicate if it is left or right eye).

These "frame_packing_arrangement" SEI messages constitute very sophisticated signaling system for 3D decoder. They have meaning "from now on use decoding method X". There is also message "stop 3D decoding".

This signaling system allows to include into one elementary stream several 3D methods and switch them accordingly on the fly.
Do you know why there is 3D method 6 = MONO?
It is intended to signal decoder that "from now on we have only 2D picture", however glasses should stay turned on and TV should display the same image into both L and R eye! (switching glasses on/off usually causes unpleasant glitches)

Apparently this was designed for broadcast but unfortunately nowadays it is over, and I doubt that decoders recognize all these 3D signaling messages....

[geheim]

Hi videofan3D,

so I did some tests with h264offset3D and it works pretty fine. The only problem is that after inserting the original offsets back into the mvc stream after reencoding, and remuxing, I get blocking artefacts when playing on my Sony standalone player. This doesn't happen without inserting the offsets.
h264offset3d seems to brake something in the stream. Do you have any idea what it could be?? Thanks for creating this nice tool at all :)

[sneaker_ger]

"frame_packing_arrangement" SEI messages can appear anytime it is needed to inform decoder about change of 3D format. They should appear in general after each Picture Parameter Set (PPS) NAL unit, and in case of mode "frame alternation" also before each frame slice NAL unit (to indicate if it is left or right eye).

These "frame_packing_arrangement" SEI messages constitute very sophisticated signaling system for 3D decoder. They have meaning "from now on use decoding method X". There is also message "stop 3D decoding".

This signaling system allows to include into one elementary stream several 3D methods and switch them accordingly on the fly.
Do you know why there is 3D method 6 = MONO?
It is intended to signal decoder that "from now on we have only 2D picture", however glasses should stay turned on and TV should display the same image into both L and R eye! (switching glasses on/off usually causes unpleasant glitches)

Apparently this was designed for broadcast but unfortunately nowadays it is over, and I doubt that decoders recognize all these 3D signaling messages....
Some players are very picky about correct frame packing SEIs. For your tool 6/mono wouldn't really make sense. As you say it only makes sense if you intend to switch from 3D to 2D. And even then I believe it is not strictly needed. It could only be useful for broadcast/live encoding when encoder doesn't yet know when the switch will occur and may need to do it in the middle of a GOP on top of that.

AVC/HEVC specs cover endless use-cases. Much of it is irrelevant in the real world. It's way too complicated.

[videofan3d]

Hi videofan3D,

so I did some tests with h264offset3D and it works pretty fine. The only problem is that after inserting the original offsets back into the mvc stream after reencoding, and remuxing, I get blocking artefacts when playing on my Sony standalone player. This doesn't happen without inserting the offsets.
h264offset3d seems to brake something in the stream. Do you have any idea what it could be?? Thanks for creating this nice tool at all :)

Hi, is there any chance to create small sample and upload it to web? (both with and w/o inserted offset_metadata). I need to check the structure...

[geheim]

Hi, is there any chance to create small sample and upload it to web? (both with and w/o inserted offset_metadata). I need to check the structure...

Of course, I'll create a small sample later today :)

[geheim]

Hi,

over the last few days I did quit a few tests with h264offset3D. And I can say that it works absolutely stable and exactly as it is meant to work. Thanks @videofan3D!

As for muxing frim-encoded streams with integrated offsets, I had some difficulties. TSMuxer muxes them perfectly fine, however the output can't be played on every standalone player because TSMuxer uses the out-of-mux method for 3D folders. This is of Course not a problem of Frim or h264offset3d.
A solution is to only mux with TSMuxer using it's BluRay ISO output option. But you can have only one title per Disc using this method.

For people like me, who want to have some menus and generate them with Scenarist the situation is a bit more complicated.
Scenarist muxes FRIM encodes withOUT offsets perfectly. Hoewever you need to uncheck "Spec check mode" in MUI generator.
You'll get a 3D bluray out of Scenarist that plays in every single standalone player because Scenarist is a professional tool and designed for perfect compatibility.
But as FRIM encodes have broken offset structure, subtitles can't have any depth, not even if you want to set Offsets manually, because scenarist doesn't offer the option to manage Offsets at all with FRIM encodes.

This is where videofan3d's new h264offset3d gets important. With this tool you can extract offsets from original files or generate a text file with offsets manually and intergrate them into the mvc stream after encoding with FRIM. This works perfectly.
Now even Scenarist recognizes the FRIM encoded file as valid (no need to disable "spec check mode").
In Scenarist you can now manage offsets in every possible way.

However, the resulting disc shows some block artefacts every 5-10 seconds during playback on standalone players.
This didn't happen without integrating offsets.
Luckily there is an easy fix for this using jdobbs' BD-Rebuilder. Just process the disc you got out of Scenarist with BD-Rebuilder. If you set the output size accordingly, BD-Rebuilder won't reencode any video files. It only remuxes the disc using the in-mux file structure, which is also compatible with every standalone player.

As Offsets are only assigned within the mpls files (thanks @videofan3D), the offsets you assigned in Scenarist will be there even after remuxing. Now the discs work perfectly fine with menus and 3d subtitle offsets at the same time. And we are able to make perfect backups of original discs with FRIM and h264offset3D as well.

I hope my post will help some people in creating their backups or own movies with Scenarist, Frim and h264offset3D.

Big thanks to @videofan3D for creating these wonderful tools.

[videofan3d]

BDTools 2.30 released:

MPLS2JSON
- added support for id.bdmv files
- sound.bdmv - WAVE files are extracted to the same directory as .json file

h264Modify
- added -frame_packing for insertion or deletion of frame packing SEI messages:
CHECKER or 0 - for checkerboard
COLINT or 1 - column interleave
ROWINT or 2 - row interleave
SBS or 3 - side by side
TAB or 4 - top above bottom
MONO or 6 - mono - 2D frame
TILE or 7 - tile format
(frame alternation (mode 5) is not supported yet)

h264Offset3D
- minor bugfixing - no new features

AVCHDInfo
- new tool. Displays recording date/time of each frame of AVCHD.MTS file extracted from vendor-specific information (if available) - in textual or tab-separated format.
Vendor-specific information contains sometimes additional data like camcorder model, recording aperture, gain, GPS info, etc. However this format is not standardized nor unified and each vendor like SONY, CANON, PANASONIC etc. uses its own coding. Further vendor-specific data might be possibly extracted in future versions of AVCHDInfo.

[Yanada]

I downloaded ver 2.30 of your awesome tools and found that MPLS2JSON seems to still be the previous version (2.20) in the 2.30 download. The release notes say that id.bdmv support was added in 2.30...

Also the help info says that it can show info for clpi files, but when I try it it states failed signature match: found ""

This is one of the most useful tools I've found so far! Keep up the great work!


As a side request, could you add the Windows standard help argument "/?" as an alias to the Linux standard "--help"? Thanks!

[videofan3d]

I downloaded ver 2.30 of your awesome tools and found that MPLS2JSON seems to still be the previous version (2.20) in the 2.30 download. The release notes say that id.bdmv support was added in 2.30...

Also the help info says that it can show info for clpi files, but when I try it it states failed signature match: found ""

This is one of the most useful tools I've found so far! Keep up the great work!


As a side request, could you add the Windows standard help argument "/?" as an alias to the Linux standard "--help"? Thanks!



There might be slight mismatch in numbering, but "MPLS to JSON - version 2.20 (build: Aug 14 2017)" support id.bdmv - you can test it.

Regarding clip and signature match: found "" - please give me the particular .clip file, I will check it.

Regarding /? ... try -help
(I don't like Microsoft's /?, it confuses regular "questions mark" meaning. Standard UNIX uses only single "minus")

[Yanada]

Oh, interesting... I'm not sure which file I tried it on. CLPI is indeed working on other files... curious (on my end).

Please ignore that problem I guess. If I can remember the file again I'll send it.

:thanks:

[tebasuna51]

A spanish user ask me about a problem playing (with a BDplayer) a custom BD with a h264 stream like this:

General
Complete name : D:\tmp\Muestra.h264
Format : AVC
Format/Info : Advanced Video Codec
File size : 132 MiB
Overall bit rate mode : Variable

Video
Format : AVC
Format/Info : Advanced Video Codec
Format profile : High@L4.1
Format settings, CABAC : Yes
Format settings, RefFrames : 2 frames
Bit rate mode : Variable
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate : 23.976 (24000/1001) FPS
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Progressive

Text
ID : CC1
Format : EIA-608
Muxing mode : SCTE 128 / DTVCC Transport
Bit rate mode : Constant
Stream size : 0.00 Byte (0%)
CaptionServiceName : CC1

Maybe the problem is the subs included: https://en.wikipedia.org/wiki/EIA-608

Is that stream allowed in BluRay specs?
Do you know a tool to modify the h264 deleting the EIA-608 subs?

Sample (https://mega.nz/#!tRMnnawL!u1Zirc0w15jwVl7pFAwoD3w-l7gZc1lCiPCp8yeIy1k)