BD Rebuilder Beta - Bug Reports Only

[jdobbs]

I have just updated the link of the first page of this thread to a new version of BD-RB (v0.29.04). Changes for this version:

- Made one more adjustment to prevent "GUI Issue,
  no hWnd returned".  I think this may be gone for
  good now.  It finally became clear as to what was
  causing this to randomly occur.
- Fixed an error in which progressive output could
  under some circumstances be created at an illegal
  (29.97fps, 25fps) framerate.
- Changed the default processing for interlaced
  sources.  All sources that are interlaced now are
  kept interlaced in the output.  This means the
  "KEEP_INTERLACING" INI setting has been removed
  (as unnecessary)
- Added a "DEINTERLACING=" hidden option, see the
  description in HIDDENOPTS.TXT for details.
- Changed the type of AVISYNTH preprocessing used to 
  ensure comb-free deinterlacing.  Interlaced sources 
  should be more crisp, and free of jagged edges now.
- Corrected a "Selected source is not BD format" error 
  that could happen on very small sources.
- Added a "quick search" function to identify sources
  that are Bottom Field First (BFF) interlaced.
- Added code to scan all M2TS files and find audio
  streams that have delayed PTS times (for correction).
- Other minor corrections and cosmetic fixes.

[chudm]

Originally Posted by jdobbs

I have just updated the link of the first page of this thread to a new version of BD-RB (v0.29.04). Changes for this version:

- Made one more adjustment to prevent "GUI Issue,
  no hWnd returned".  I think this may be gone for
  good now.  It finally became clear as to what was
  causing this to randomly occur.
- Fixed an error in which progressive output could
  under some circumstances be created at an illegal
  (29.97fps, 25fps) framerate.
- Changed the default processing for interlaced
  sources.  All sources that are interlaced now are
  kept interlaced in the output.  This means the
  "KEEP_INTERLACING" INI setting has been removed
  (as unnecessary)
- Added a "DEINTERLACING=" hidden option, see the
  description in HIDDENOPTS.TXT for details.
- Changed the type of AVISYNTH preprocessing used to 
  ensure comb-free deinterlacing.  Interlaced sources 
  should be more crisp, and free of jagged edges now.
- Corrected a "Selected source is not BD format" error 
  that could happen on very small sources.
- Added a "quick search" function to identify sources
  that are Bottom Field First (BFF) interlaced.
- Added code to scan all M2TS files and find audio
  streams that have delayed PTS times (for correction).
- Other minor corrections and cosmetic fixes.

great news man! thx!

[PurpleMan]

- Added code to scan all M2TS files and find audio
streams that have delayed PTS times (for correction).

Is this refering to the Batman Begins timeshift bug? If so, could you elaborate what exactly was the problem and how we're working around it?

Thanks for the excellent work!

[jdobbs]

Originally Posted by PurpleMan

Is this refering to the Batman Begins timeshift bug? If so, could you elaborate what exactly was the problem and how we're working around it?

Thanks for the excellent work!

Yes. In looking at the M2TS file, I noticed that the first Presentation Time Stamp (PTS) for the audio track doesn't start until 83ms after the first PTS for the video. That's where the sync error came from. I just added a routine that scans all M2TS files looking for the starting PTS for video/audio tracks and accounted for any delta. If there is one, I then put a corrective "timeshift=XXms" in the meta when demuxing the audio.

Interestingly, this is pretty rare -- I could only find one other disc of my standard working/test set that had an offset, and it wasn't enough to matter (about 6ms). But you have to cover all scenarios. The new scan will probably add a couple of seconds to the initial scan of most discs when you select the source path.

As long as I was scanning anyway -- I also added code that looks for the TFF flag on video streams -- so I can sense when a source was BFF (but not pulldown)... which, by the way, turned out to be part of the the issue with the "Making Of" extras on "The Matrix".

[Capsbackup]

Originally Posted by jdobbs

Yes. In looking at the M2TS file, I noticed that the first Presentation Time Stamp (PTS) for the audio track doesn't start until 83ms after the first PTS for the video. That's where the sync error came from. I just added a routine that scans all M2TS files looking for the starting PTS for video/audio tracks and accounted for any delta. If there is one, I then put a corrective "timeshift=XXms" in the meta when demuxing the audio.

Interestingly, this is pretty rare -- I could only find one other disc of my standard working/test set that had an offset, and it wasn't enough to matter (about 6ms). But you have to cover all scenarios. The new scan will probably add a couple of seconds to the initial scan of most discs when you select the source path.

As long as I was scanning anyway -- I also added code that looks for the TFF flag on video streams -- so I can sense when a source was BFF (but not pulldown)... which, by the way, turned out to be part of the the issue with the "Making Of" extras on "The Matrix".

I agree, there are not many discs I have come across that have audio delays. However, could you clarify the correction.
If the original .mpls/m2ts shows a delay, for example Batman Begins 83ms, should BD-RB's backup also show a delay in the same mpls/m2ts?
If one chooses to keep HD audio and do not reencode ac3, will this delay be maintained per example above? Or is there another method for you to correct/address the delay?
I guess I am just curious how to determine that the delay is correct besides the visual check I always perform anyway.

[jdobbs]

Originally Posted by Capsbackup

I agree, there are not many discs I have come across that have audio delays. However, could you clarify the correction.
If the original .mpls/m2ts shows a delay, for example Batman Begins 83ms, should BD-RB's backup also show a delay in the same mpls/m2ts?
If one chooses to keep HD audio and do not reencode ac3, will this delay be maintained per example above? Or is there another method for you to correct/address the delay?
I guess I am just curious how to determine that the delay is correct besides the visual check I always perform anyway.

No. Not normally. The audio stream is corrected to include the delay. For example, if you have an 83ms delay you would put 83ms of silence at the beginning of the audio stream. I'm actually letting TSMUXER do the adjustment during the demux by using the timeshift command -- but that would be my guess as to how it does it. Since the delay is added during the demux, it would carry through any reencode, or be there already if you do keep the HD original.

[Capsbackup]

Originally Posted by jdobbs

No. Not normally. The audio stream is corrected to include the delay. For example, if you have an 83ms delay you would put 83ms of silence at the beginning of the audio stream. I'm actually letting TSMUXER do the adjustment during the demux by using the timeshift command -- but that would be my guess as to how it does it. Since the delay is added during the demux, it would carry through any reencode, or be there already if you do keep the HD original.

Interesting. Is there any way to tell if tsMuxer applies the delay to the demux?
My experiment with Batman Begins, doing just this and remuxing it back into BD-RB would still not bring the audio in sync.
Only after editing the BD-RB meta file for this m2ts file to "time shift=250",( 83ms was not enough) then rebuilding that m2ts thru the command line, and then substituting this new m2ts into the BD-RB rebuild was I able to get the audio in sync.
I noticed that when checking this new mpls/m2ts created with BD-RB with tsMuxeR, it now also showed the 250ms delay.
This worked for me for this disc, be it cumbersome, audio was now in sync.
So, not to be a bother, but this method of timeshift within the BD-RB meta file also works, and seems to reflect the original mpls/m2ts as well.

[jdobbs]

Originally Posted by Capsbackup

Interesting. Is there any way to tell if tsMuxer applies the delay to the demux?
My experiment with Batman Begins, doing just this and remuxing it back into BD-RB would still not bring the audio in sync.
Only after editing the BD-RB meta file for this m2ts file to "time shift=250",( 83ms was not enough) then rebuilding that m2ts thru the command line, and then substituting this new m2ts into the BD-RB rebuild was I able to get the audio in sync.
I noticed that when checking this new mpls/m2ts created with BD-RB with tsMuxeR, it now also showed the 250ms delay.
This worked for me for this disc, be it cumbersome, audio was now in sync.
So, not to be a bother, but this method of timeshift within the BD-RB meta file also works, and seems to reflect the original mpls/m2ts as well.

Hmmm... I'm doing "Batman Begins" right now -- I'll let you know how it turns out. If it doesn't work like I thought, I can add the delay during the remux (which is what I'm guessing you did).

[Capsbackup]

Originally Posted by jdobbs

Hmmm... I'm doing "Batman Begins" right now -- I'll let you know how it turns out. If it doesn't work like I thought, I can add the delay during the remux (which is what I'm guessing you did).

Correct. I inserted/edited BD-RB's meta file for that m2ts with "timeshift=250ms" in the appropriate location within that meta file.
However, just editing and saving this meta file inside the Workfiles folder was not working. Upon rebuild, BD-RB would change/correct my edits to back to the original meta. I had to run that edited meta from the command line, and then substitute this new m2ts/clpi into the BD-RB Workfiles folder for a new rebuild. ( Using a creative rebuild, canceling the rebuild just after the final m2ts is created, substituting the new m2ts/clpi for the ones BD-RB just rebuilt, and then selecting rebuild again and letting BD-RB just integrate all those m2ts/clpi/mpls files into the new BluRay structure.
EDIT:
I will be curious if you have audio sync this first time try. Though 83ms is the original delay, this was not enough to correct the async for me. Only after several tests/burns to BD-RE did I finally get audio sync, but it took 250ms. Definately an odd disc!

[jdobbs]

Originally Posted by Capsbackup

Correct. I inserted/edited BD-RB's meta file for that m2ts with "timeshift=250ms" in the appropriate location within that meta file.
However, just editing and saving this meta file inside the Workfiles folder was not working. Upon rebuild, BD-RB would change/correct my edits to back to the original meta. I had to run that edited meta from the command line, and then substitute this new m2ts/clpi into the BD-RB Workfiles folder for a new rebuild. ( Using a creative rebuild, canceling the rebuild just after the final m2ts is created, substituting the new m2ts/clpi for the ones BD-RB just rebuilt, and then selecting rebuild again and letting BD-RB just integrate all those m2ts/clpi/mpls files into the new BluRay structure.
EDIT:
I will be curious if you have audio sync this first time try. Though 83ms is the original delay, this was not enough to correct the async for me. Only after several tests/burns to BD-RE did I finally get audio sync, but it took 250ms. Definately an odd disc!

I just tested it. 83ms is the correct value on my system, it synched up perfectly (after I did it right)... but in the release I did it ass-backwards. The "timeshift" parameter has to be (as you did it) on the mux, not the demux.

I've corrected it for the next release. Luckily there aren't very many discs out there that need it.

[Capsbackup]

Originally Posted by jdobbs

I just tested it. 83ms is the correct value on my system, it synched up perfectly (after I did it right)... but in the release I did it ass-backwards. The "timeshift" parameter has to be (as you did it) on the mux, not the demux.

I've corrected it for the next release. Luckily there aren't very many discs out there that need it.

Great! Thankfully there are not many discs made with audio delays of significance, but when there is, you have it covered.

[samtroy]

Originally Posted by jdobbs

No bug that I know of. Are you using BATCH mode? In BATCH, the settings are stored when you add it to the queue, and any changes made later has no effect on the BATCH encode.

Also, are you sure two of the tracks aren't secondary audio tracks? You said "similar names", which would be the case if that were so.

[edit] I just went back and looked at the code and it is implemented correctly. I did a couple of tests, and they appear to have worked correctly also.

No, I didn't use batch mode but the normal movie-only mode. But I will do some more tests later on and let you know more specifics tomorrow.

Thanks for checking and generally for your awsome program .

[chudm]

hi guys tried to do a rip and cant do it, says target is too small to continue, i wrote on the ini the force encode =1 option, but still had the same problem with the lastest 2 beta, any suggestion?

[jdobbs]

Originally Posted by chudm

hi guys tried to do a rip and cant do it, says target is too small to continue, i wrote on the ini the force encode =1 option, but still had the same problem with the lastest 2 beta, any suggestion?

It means that all the audio and subtitle selections you have made combined are already too big to fit on the output disc size you've chosen -- even without the video. The only way you can fix it is to either choose a bigger output size or select fewer audio/subs to keep. FORCE_ENCODE won't help becaus you can't reencode to a negative size... this is most likely seen when LPCM or lossless tracks exist and you're trying to output to BD-5/9.

[chudm]

Originally Posted by jdobbs

It means that all the audio and subtitle selections you have made combined are already too big to fit on the output disc size you've chosen -- even without the video. The only way you can fix it is to either choose a bigger output size or select fewer audio/subs to keep. FORCE_ENCODE won't help becaus you can't reencode to a negative size... this is most likely seen when LPCM or lossless tracks exist and you're trying to output to BD-5/9.

im doing full backup to bd25

[Sharc]

Originally Posted by jdobbs

I just tested it. 83ms is the correct value on my system, it synched up perfectly (after I did it right)... but in the release I did it ass-backwards. The "timeshift" parameter has to be (as you did it) on the mux, not the demux.

I've corrected it for the next release. Luckily there aren't very many discs out there that need it.

Looks like this would also solve the sync problem which I reported earlier for "DOUBT". It was actually the only disc I ever had a sync problem with

[Race Guy]

This bug is called "The Case Of The Missing Menu Text". But I don't think it's really a bug!

I've had 3 rebuilds that have ended up "stuck in an endless loop" on the top menu, because the "button text" is not there. I'm talking about the "Play", "Audio Select", "Special Features"....., etc "buttons".

On one of them (Elton 60), I'm able to use my player's "Go-To" button & then manually select the main title & then the disc takes off & plays fine. Even the pop up menu works to navigate chapter to chapter.

The other two, (Dave Matthews Radio City & Baraka) wont allow me to do that. On the latest one to fail, (Baraka) I did a little investigation With BD Info & found the culprit. The "interactive graphics stream (IGS)" is missing from the rebuilt menu video's remux.

Here's the Original:

Name_________Time In_______Length
00012.M2TS___0:00:00.000___0:08:45.525

STREAM DIAGNOSTICS:

File___________PID___________Type___Codec
00012.M2TS___4113 (0x1011)___0xEA___VC-1
00012.M2TS___4352 (0x1100)___0x81___AC3
00012.M2TS___4353 (0x1101)___0x81___AC3
00012.M2TS___5120 (0x1400)___0x91___IGS

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

Here's the Rebuild:

Name_________Time In_______Length
00012.M2TS___0:00:00.000___0:08:45.525

STREAM DIAGNOSTICS:

File___________PID___________Type___Codec
00012.M2TS___4113 (0x1011)___0xEA___VC-1
00012.M2TS___4352 (0x1100)___0x81___AC3

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

So then I reopened the project in BD-RB & discovered that it's not really a "bug"! I have my config setup to keep all English subs only. That IGS (for whatever reason???) had 3 "languages", 2 English & 1 Spanish. With my config, Spanish is eliminated which eliminated the WHOLE IGS!

Unfortunately, I dumped my workfiles, so for the helluvit, & rebuilt that menu vid again, this time with all the audio & all the sub languages. I then replaced those "corrected" m2ts & clpi files in the structure. I burned it to RE & it appears that I "saved" the project! The menu now comes up & functions & it appears all is good now!

Is what I did there really a "good" fix jdobbs?

I think the moral of this story is "keep ALL your subs"!!!

[chudm]

Originally Posted by Race Guy

This bug is called "The Case Of The Missing Menu Text". But I don't think it's really a bug!

I've had 3 rebuilds that have ended up "stuck in an endless loop" on the top menu, because the "button text" is not there. I'm talking about the "Play", "Audio Select", "Special Features"....., etc "buttons".

On one of them (Elton 60), I'm able to use my player's "Go-To" button & then manually select the main title & then the disc takes off & plays fine. Even the pop up menu works to navigate chapter to chapter.

The other two, (Dave Matthews Radio City & Baraka) wont allow me to do that. On the latest one to fail, (Baraka) I did a little investigation With BD Info & found the culprit. The "interactive graphics stream (IGS)" is missing from the rebuilt menu video's remux.

Here's the Original:

Name_________Time In_______Length
00012.M2TS___0:00:00.000___0:08:45.525

STREAM DIAGNOSTICS:

File___________PID___________Type___Codec
00012.M2TS___4113 (0x1011)___0xEA___VC-1
00012.M2TS___4352 (0x1100)___0x81___AC3
00012.M2TS___4353 (0x1101)___0x81___AC3
00012.M2TS___5120 (0x1400)___0x91___IGS

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

Here's the Rebuild:

Name_________Time In_______Length
00012.M2TS___0:00:00.000___0:08:45.525

STREAM DIAGNOSTICS:

File___________PID___________Type___Codec
00012.M2TS___4113 (0x1011)___0xEA___VC-1
00012.M2TS___4352 (0x1100)___0x81___AC3

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

So then I reopened the project in BD-RB & discovered that it's not really a "bug"! I have my config setup to keep all English subs only. That IGS (for whatever reason???) had 3 "languages", 2 English & 1 Spanish. With my config, Spanish is eliminated which eliminated the WHOLE IGS!

Unfortunately, I dumped my workfiles, so for the helluvit, & rebuilt that menu vid again, this time with all the audio & all the sub languages. I then replaced those "corrected" m2ts & clpi files in the structure. I burned it to RE & it appears that I "saved" the project! The menu now comes up & functions & it appears all is good now!

Is what I did there really a "good" fix jdobbs?

I think the moral of this story is "keep ALL your subs"!!!

yeap had the same problems with foo fighters bluray, and this elton john bluray

[DK]

I'm having problems with Tomb Raider 1 Region B

==========================
[21:48:28] BD Rebuilder v0.29.03 (beta)
[21:48:28] BEGIN - BATCH SEQUENCE
==========================
-----------------------
[21:48:28] PROCESSING BATCH FILE [1]
-----------------------
[08:06:27] BD Rebuilder v0.29.03 (beta)
- Source: LARA_CROFT_1
- Input BD size: 17,78 GB
- Approximate total content: [01:40:45.456]
- Target BD size: 7,72 GB
- MOVIE-ONLY mode enabled
[08:06:27] PHASE ONE, Encoding
- [08:06:27] Extracting A/V streams [VID_00004]
- [08:13:43] Reencoding: VID_00004 (1 of 2)
- Analyzing 25,42 10,00 18,45 21,28 20,28 20,10 19,95 20,07 20,10 20,09 [20,09]
- [20:32:34] Extracting A/V streams [VID_00006]
- [20:32:36] Reencoding: VID_00006 (2 of 2)
- Error in attempt to multiplex: MUX_00006.meta
- Can't open file: G:\DVD\TR1\WORKFILES\VID_00006.AVS.264
[20:32:45] - Failed to build structure, aborted

-----------------------

[22:32:23] BD Rebuilder v0.29.04 (beta)
- Source: LARA_CROFT_1
- Input BD size: 17,78 GB
- Approximate total content: [01:40:45.456]
- Target BD size: 7,72 GB
- MOVIE-ONLY mode enabled
[22:32:30] PHASE ONE, Encoding
- [22:32:30] Extracting A/V streams [VID_00004]
- [22:38:39] Reencoding: VID_00004 (1 of 2)
- [11:11:51] Extracting A/V streams [VID_00006]
- [11:11:51] Reencoding: VID_00006 (2 of 2)
- Error in attempt to multiplex: MUX_00006.meta
- Can't open file: G:\DVD\TR1\WORKFILES\VID_00006.AVS.264
[11:12:03] - Failed to build structure, aborted

VID_00006.AVS.264 is not present, only:

- VID_00006.AVS
- VID_00006.AVS.MKV
- VID_00006.AVS.MKV.stats

The main movie is:
VID_00004.AVS.264 and this was created properly and completely together with the two sound streams.

Muxing video + audio streams woorked fine via tsmuxer, so the goal was achieved: a movie only backup but i have no idea why bd rebuilder got stuck with two different versions and slightly different settings:
1pass with version 0.28.03 and 2 passes with 0.29.04

[chudm]

about the problems with some musical blurays jdobbs commented once that its caused possibly by lpcm audio or something like that, hope he can fix it sometime