BD Rebuilder Beta - Bug Reports Only - Page 714

jdobbs · 7th February 2012, 19:47

Thanks! Yes that helps. That gives me a little better idea as to where the error is occuring. I'll scour that section of code.

jdobbs · 7th February 2012, 19:57

@RobertM

I just send you a PM...

AmigaFuture · 7th February 2012, 20:55

It stated 3 splits. I hope you find a way to leave them both active.

I much prefer cropping and the speed. Hehe. I'll disable cropping in my next test but leave everything else as posted before. I thought about changing the number of MULTIPROCESS, but hadn't created the moments to do it yet. I noticed 40.01 also does it. I'll stay with .02 for the next tests.

The_Unknown · 7th February 2012, 20:57

The error also occurs with Harry Potter and the Halfblood Prince (also Bonus Disc).

Directly after the error:

Code:

07.02.2012  17:55    <DIR>          .
07.02.2012  17:55    <DIR>          ..
07.02.2012  17:55        13.702.765 00003.track_4113.264
07.02.2012  17:55               145 AUD_00003.meta
07.02.2012  17:55               169 VID_00003.AVS
               3 Datei(en),     13.703.079 Bytes
               2 Verzeichnis(se), 274.110.214.144 Bytes frei

INI and LOG are in attachment.

jdobbs · 7th February 2012, 23:32

Quote:

Originally Posted by The_Unknown

The error also occurs with Harry Potter and the Halfblood Prince (also Bonus Disc).

Directly after the error:

Code:

07.02.2012  17:55    <DIR>          .
07.02.2012  17:55    <DIR>          ..
07.02.2012  17:55        13.702.765 00003.track_4113.264
07.02.2012  17:55               145 AUD_00003.meta
07.02.2012  17:55               169 VID_00003.AVS
               3 Datei(en),     13.703.079 Bytes
               2 Verzeichnis(se), 274.110.214.144 Bytes frei

INI and LOG are in attachment.

Thanks. I've found the problem and fixed it for the next release. Hopefully I will post it later today. I just want to do a little testing first.

AmigaFuture · 8th February 2012, 00:47

@jdobbs

I like the other method of encoding you typed about. That seems more along the way you have DVD-Rebuilder doing it. With that, HCEnc will use multiple instances of itself depending how many Cores the CPU has which results in very high FPS also, except for the last encode. After multiple encodes and only 1 process remaining, the FPS is reflected only for that single HCEnc process. However, when Mutl-Processing, whatever it's called..I don't recall at the moment, is disabled...the FPS is high as if multi is being used but with 1 instance of HCEnc. Which I see as it using multiple cores...with 1 HCEnc. I hoped that for MULTIPROCESS, but it doesn't seem like it with BD-Rebuilder. Is it possible to do the same with the external x264? I'm not using the internal x264 any more. I am noticing the 32-bit version is running only...so I'm guessing that it's not making calls to the 64-bit version?

Current render is with MULTIPROCESS=0 and Crop for Alternate Intact Audio is active. No splitting has happened. I got 40.XX FPS 1st pas. 19.XX 2nd pass.

jdobbs · 8th February 2012, 01:49

BD-RB works the same way as DVD Rebuilder for multi-processing. The difference is that it is taking a lot more processing power for HD encoding, so the benefit only shows when you have spare processor time. On faster processors it will show more benefit.

If you use AVISYNTH as input you have to use the 32 bit version of X264. It really doesn't matter anyway -- in my testing the speed difference is negligible between 32/64 bit anyway. If you did see a difference it was more likely from LAVF.

jdobbs · 8th February 2012, 02:39

I have updated the first post of this thread with a link to a new version of BD-RB (v0.40.04). Changes for this release:

Code:

- Added support for Neuron2's DGDecNV.  Note that
  this software is not required for BD-RB to run, 
  but it will likely speed encoding in most cases. 
  DGDecNV requires a CUDA enabled video card and
  a license that can be obtained through a $15 
  donation. For more info and h/w requirements: 
  http://www.neuron2.net/dgdecnv/dgdecnv.html
- Added code to make DGDecNV indexing execute
  concurrently with TSMUXER audio extraction. This
  should, in most cases, eliminate any additional
  time needed for the indexing.  If problems occur
  however, concurrent indexing can be disabled
  through a new INDEX_WITH_EXTRACT hidden option.
  See HIDDENOPTS.TXT for more info.
- Corrected a bug that could result in a crash and
  the error "PredictAndEncode() 00006 XXXX" when
  MULTIPROCESS is active.
- Modified the log so the current date is shown
  as the job begins rather than the time-of-day.
- Updated the included version of X264.EXE to the
  latest release (r2164).
- Updated the included version of X264-64.EXE to 
  the latest release (r2164).
- Other minor corrections and cosmetic fixes.

RobertM · 8th February 2012, 03:14

Good news!

40.04 seems to have fixed my problem. I didn't let it encode the whole thing through, but it completed the extraction and then started splitting for the multiprocess with no error, so it has certainly gotten past my precious problem area.

Thanks.

AmigaFuture · 8th February 2012, 03:18

I get it about the resolution differences and that to see higher FPS with higher resolution you need to tap more power (vibration) to get higher FPS...or "seemingly" have multiple processes of a particular program. Which I noticed DVD Rebuilder didn't have to (multiple processes of HCEnc) to get higher FPS.

The last encoding wasn't split, but I verified it anyway and it was smooth. Reading your last post to me...I though I'd not activate the external LAVF and try that since I haven't for several past tests.

jdobbs · 8th February 2012, 05:33

Quote:

Originally Posted by AmigaFuture

I get it about the resolution differences and that to see higher FPS with higher resolution you need to tap more power (vibration) to get higher FPS...or "seemingly" have multiple processes of a particular program. Which I noticed DVD Rebuilder didn't have to (multiple processes of HCEnc) to get higher FPS.

The last encoding wasn't split, but I verified it anyway and it was smooth. Reading your last post to me...I though I'd not activate the external LAVF and try that since I haven't for several past tests.

Yes... it does use multiple process of HCEnc... but later versions of HCEnc incorporated internal multiprocessing so the increases weren't as great. I'm pretty familiar with how it works, having written it and all.

AmigaFuture · 8th February 2012, 06:19

My intention wasn't to express you're not aware of your own program, but to help me further understand the Encoding process and speeds, and "Is there a Bug here??". Now I get it about the version of HCEnc I use which answers other question(s) a bit more even about BD-RB. :-D I Appreciate your reply, Thank You. :-)

Please keep the "Automatically crop black borders" option. I'm cool with using slower speeds for that. Fast is good, but slow has it's...benefits also. ;-)

colinhunt · 8th February 2012, 13:35

Anyone done The Thing (2011) with PiP video/DTS Express enabled?

I ran it on 0.40.02 (multiprocess disabled) and the output plays PiP video without audio on PowerDVD 11. The original plays fine, audio and all, on PDVD11.

Slight clarification: when PiP video window opens and the secondary audio is activated, PDVD11 lowers the volume level of the main audio. On the backup, main audio level is lowered but secondary audio remains mute.

Also, on the original PDVD11 shows "DTSHD" as the codec of 2nd audio, along with the bitrate of the secondary audio. On the backup, only "DTSHD" is shown on the information OSD; bitrate is not shown.

colinhunt · 8th February 2012, 13:44

Took a closer look at the secondary audio specs on both original and backup with Mediainfo. Here's the data, with parts that differ marked in bold.

Original

Code:

Audio #6
Count                            : 169
Count of stream of this kind     : 6
Kind of stream                   : Audio
Kind of stream                   : Audio
Stream identifier                : 5
Stream identifier                : 6
ID                               : 6656
ID                               : 6656 (0x1A00)
Menu ID                          : 1
Menu ID                          : 1 (0x1)
Format                           : DTS
Format/Info                      : Digital Theater Systems
Commercial name                  : DTS
Format profile                   : Express
Codec ID                         : 162
Codec                            : DTS
Codec                            : DTS
Duration                         : 6180779
Duration                         : 1h 43mn
Duration                         : 1h 43mn 0s 779ms
Duration                         : 1h 43mn
Duration                         : 01:43:00.779
Bit rate mode                    : CBR
Bit rate mode                    : Constant
Bit rate                         : 192000
Bit rate                         : 192 Kbps
Channel(s)                       : 2
Channel(s)                       : 2 channels
Channel positions                : Front: L R
Channel positions                : 2/0/0
Sampling rate                    : 48000
Sampling rate                    : 48.0 KHz
Samples count                    : 296677392
Frame count                      : 193149
Resolution                       : 24
Resolution                       : 24 bits
Bit depth                        : 24
Bit depth                        : 24 bits
Compression mode                 : Lossy
Compression mode                 : Lossy
Delay                            : 11650.667
Delay                            : 11s 651ms
Delay                            : 11s 651ms
Delay                            : 11s 651ms
Delay                            : 00:00:11.651
Delay, origin                    : Container
Delay, origin                    : Container
Delay relative to video          : 0
Video0 delay                     : 0
Stream size                      : 148338696
Stream size                      : 141 MiB (0%)
Stream size                      : 141 MiB
Stream size                      : 141 MiB
Stream size                      : 141 MiB
Stream size                      : 141.5 MiB
Stream size                      : 141 MiB (0%)
Proportion of this stream        : 0.00488

Backup

Code:

Audio #3
Count                            : 169
Count of stream of this kind     : 3
Kind of stream                   : Audio
Kind of stream                   : Audio
Stream identifier                : 2
Stream identifier                : 3
ID                               : 6656
ID                               : 6656 (0x1A00)
Menu ID                          : 1
Menu ID                          : 1 (0x1)
Format                           : DTS
Format/Info                      : Digital Theater Systems
Commercial name                  : DTS
Format profile                   : Express
Codec ID                         : 162
Codec                            : DTS
Codec                            : DTS
Duration                         : 6180608
Duration                         : 1h 43mn
Duration                         : 1h 43mn 0s 608ms
Duration                         : 1h 43mn
Duration                         : 01:43:00.608
Bit rate mode                    : CBR
Bit rate mode                    : Constant
Bit rate                         : 192000
Bit rate                         : 192 Kbps
Channel(s)                       : 2
Channel(s)                       : 2 channels
Channel positions                : Front: L R
Channel positions                : 2/0/0
Sampling rate                    : 48000
Sampling rate                    : 48.0 KHz
Samples count                    : 296669184
Frame count                      : 193144
Resolution                       : 24
Resolution                       : 24 bits
Bit depth                        : 24
Bit depth                        : 24 bits
Compression mode                 : Lossy
Compression mode                 : Lossy
Delay                            : 600000.000
Delay                            : 10mn 0s
Delay                            : 10mn 0s 0ms
Delay                            : 10mn 0s
Delay                            : 00:10:00.000
Delay, origin                    : Container
Delay, origin                    : Container
Delay relative to video          : 0
Video0 delay                     : 0
Stream size                      : 148334592
Stream size                      : 141 MiB (1%)
Stream size                      : 141 MiB
Stream size                      : 141 MiB
Stream size                      : 141 MiB
Stream size                      : 141.5 MiB
Stream size                      : 141 MiB (1%)
Proportion of this stream        : 0.00742

Note: the original is VC-1 encoded for both main and PiP video.

Note 2: the answer (or part of it) might lie in the length of Delay. Original has 11 seconds, backup has 10 minutes.

jdobbs · 8th February 2012, 16:37

Quote:

Originally Posted by colinhunt

Note 2: the answer (or part of it) might lie in the length of Delay. Original has 11 seconds, backup has 10 minutes.

Sorry, but that isn't it. That's just the "start" point that TSMUXER always uses. You'll find that all other video and audio streams also start at that point. Commercial discs often use 11.650 seconds as their start point -- probably because they use the same authoring software. If you open up the disc with BDEDIT you'll see that the "delay" isn't a delay at all -- it is the same for all the streams.

colinhunt · 8th February 2012, 17:55

Quote:

Originally Posted by jdobbs

Sorry, but that isn't it. That's just the "start" point that TSMUXER always uses. You'll find that all other video and audio streams also start at that point. Commercial discs often use 11.650 seconds as their start point -- probably because they use the same authoring software. If you open up the disc with BDEDIT you'll see that the "delay" isn't a delay at all -- it is the same for all the streams.

Okey-dokey. It's something else, then. Let me know if there's something I can do to help with debugging.

soneca · 8th February 2012, 17:55

I expected a better difference in speed, the first pass time was higher using DGDecNV, and the second pass, the difference was very small.
In this case I do not know if it's worth sharing the job with the video card(GTS450/1GB).

Code:

[02/08/12] BD Rebuilder v0.40.04 (beta)
[00:27:28] Source:  CLEOPATRA_DISC_1
  - Input BD size: 31,29 GB
  - Approximate total content: [01:59:34.584]
  - Target BD size: 7,84 GB
  - Windows Version: 6.1 [7601]
  - MOVIE-ONLY mode enabled
  - RESIZE 1080p to 720p enabled
  - SHUTDOWN at completion enabled
  - Quality: Highest (Very Slow), Two Pass
  - X264 Tweak(s) enabled
  - Decoding/Frame serving: DGDecNV
  - Audio Settings: AC3=0 DTS=0 HD=0 Kbs=640
  - Subtitle [9] is defaulted ON.
  - Audio [1] set to default.
[00:27:28] PHASE ONE, Encoding
 - [00:27:28] Processing: VID_00001 (1 of 1)
 - [00:27:28] Extracting A/V streams [VID_00001]
 - [00:32:54] Reencoding video [VID_00001]
   - Source Video: MPEG-4 (AVC), 1920x1080
   - Rate/Length: 23,976fps, 172.018 frames
   - Convert: 1280x720, 23,976fps, 172.018 frames
   - Bitrate: 7.271 Kbs
 - [00:32:54] Reencoding: VID_00001, Pass 1 of 2
 - [01:20:54] Reencoding: VID_00001, Pass 2 of 2
 - [02:42:39] Video Encode complete
 - [02:42:39] Processing audio tracks
   - Track 4352 (eng): Reencoding audio to AC3...
   - Track 4357 (por): Keeping original audio
 - [02:45:50] Multiplexing M2TS
[02:50:01]PHASE ONE complete
[02:50:01]PHASE TWO - Rebuild Started
 - [02:50:01] Rebuilding AVCHD file Structure
[02:54:50] - Encode and Rebuild complete
 - Computer SHUTDOWN initiated.
----------------------
[02/08/12] BD Rebuilder v0.40.04 (beta)
[11:29:32] Source:  CLEOPATRA_DISC_1
  - Input BD size: 31,29 GB
  - Approximate total content: [01:59:34.584]
  - Target BD size: 7,84 GB
  - Windows Version: 6.1 [7601]
  - MOVIE-ONLY mode enabled
  - RESIZE 1080p to 720p enabled
  - Quality: Highest (Very Slow), Two Pass
  - X264 Tweak(s) enabled
  - Decoding/Frame serving: DirectShow
  - Audio Settings: AC3=0 DTS=0 HD=0 Kbs=640
  - Subtitle [9] is defaulted ON.
  - Audio [1] set to default.
[11:29:36] PHASE ONE, Encoding
 - [11:29:36] Processing: VID_00001 (1 of 1)
 - [11:29:36] Extracting A/V streams [VID_00001]
 - [11:34:58] Reencoding video [VID_00001]
   - Source Video: MPEG-4 (AVC), 1920x1080
   - Rate/Length: 23,976fps, 172.018 frames
   - Convert: 1280x720, 23,976fps, 172.018 frames
   - Bitrate: 7.271 Kbs
 - [11:34:58] Reencoding: VID_00001, Pass 1 of 2
 - [12:09:51] Reencoding: VID_00001, Pass 2 of 2
 - [13:39:09] Video Encode complete
 - [13:39:09] Processing audio tracks
   - Track 4352 (eng): Reencoding audio to AC3...
   - Track 4357 (por): Keeping original audio
 - [13:42:17] Multiplexing M2TS
[13:45:36]PHASE ONE complete
[13:45:36]PHASE TWO - Rebuild Started
 - [13:45:36] Rebuilding AVCHD file Structure
[13:50:25] - Encode and Rebuild complete
[13:50:25] JOB: CLEOPATRA_DISC_1 finished.

jdobbs · 8th February 2012, 18:02

Quote:

Originally Posted by soneca

I expected a better difference in speed, the first pass time was higher using DGDecNV, and the second pass, the difference was very small.
In this case I do not know if it's worth sharing the job with the video card(GTS450/1GB).

Results can vary, I guess. But I can tell you that on my system I'm seeing increases in speed of 15-25%, and my video card is far from top-of-the-line (GT520). So it's definitely worth it to me. I can't guarantee everyone will see those kinds of increases -- for all I know they may see more. But I think its safe to say the most will see some increase.

Have you turned INDEX_WITH_EXTRACT off? I've never seen pass 1 be slower with DGDecNV...

colinhunt · 8th February 2012, 18:04

What's the working relationship of LAVF and DGDecNV? Does one override the other or..?

jdobbs · 8th February 2012, 18:21

Quote:

Originally Posted by colinhunt

What's the working relationship of LAVF and DGDecNV? Does one override the other or..?

LAVF, DGDecNV, and DirectShow all perform the same function, so you have to pick only one of the three.

7th February 2012, 19:47	#14261 \| Link
jdobbs Moderator Join Date: Oct 2001 Posts: 20,975	Thanks! Yes that helps. That gives me a little better idea as to where the error is occuring. I'll scour that section of code. __________________ Help with development of new apps: Donations. Website: www.jdobbs.net

7th February 2012, 19:57	#14262 \| Link
jdobbs Moderator Join Date: Oct 2001 Posts: 20,975	@RobertM I just send you a PM... __________________ Help with development of new apps: Donations. Website: www.jdobbs.net

7th February 2012, 20:55	#14263 \| Link
AmigaFuture Registered User Join Date: Apr 2003 Location: Within the main Source. Posts: 895	It stated 3 splits. I hope you find a way to leave them both active. I much prefer cropping and the speed. Hehe. I'll disable cropping in my next test but leave everything else as posted before. I thought about changing the number of MULTIPROCESS, but hadn't created the moments to do it yet. I noticed 40.01 also does it. I'll stay with .02 for the next tests. __________________ Life is not a journey to the grave; but rather to skid out broadside, thoroughly used, torn and warn and loudly proclaim; WOW; What a ride!!! Soon, I'm going to do it AGAiN in different skin!!

8th February 2012, 00:47	#14266 \| Link
AmigaFuture Registered User Join Date: Apr 2003 Location: Within the main Source. Posts: 895	@jdobbs I like the other method of encoding you typed about. That seems more along the way you have DVD-Rebuilder doing it. With that, HCEnc will use multiple instances of itself depending how many Cores the CPU has which results in very high FPS also, except for the last encode. After multiple encodes and only 1 process remaining, the FPS is reflected only for that single HCEnc process. However, when Mutl-Processing, whatever it's called..I don't recall at the moment, is disabled...the FPS is high as if multi is being used but with 1 instance of HCEnc. Which I see as it using multiple cores...with 1 HCEnc. I hoped that for MULTIPROCESS, but it doesn't seem like it with BD-Rebuilder. Is it possible to do the same with the external x264? I'm not using the internal x264 any more. I am noticing the 32-bit version is running only...so I'm guessing that it's not making calls to the 64-bit version? Current render is with MULTIPROCESS=0 and Crop for Alternate Intact Audio is active. No splitting has happened. I got 40.XX FPS 1st pas. 19.XX 2nd pass. __________________ Life is not a journey to the grave; but rather to skid out broadside, thoroughly used, torn and warn and loudly proclaim; WOW; What a ride!!! Soon, I'm going to do it AGAiN in different skin!!

8th February 2012, 01:49	#14267 \| Link
jdobbs Moderator Join Date: Oct 2001 Posts: 20,975	BD-RB works the same way as DVD Rebuilder for multi-processing. The difference is that it is taking a lot more processing power for HD encoding, so the benefit only shows when you have spare processor time. On faster processors it will show more benefit. If you use AVISYNTH as input you have to use the 32 bit version of X264. It really doesn't matter anyway -- in my testing the speed difference is negligible between 32/64 bit anyway. If you did see a difference it was more likely from LAVF. __________________ Help with development of new apps: Donations. Website: www.jdobbs.net

8th February 2012, 03:14	#14269 \| Link
RobertM Registered User Join Date: Apr 2011 Posts: 429	Good news! 40.04 seems to have fixed my problem. I didn't let it encode the whole thing through, but it completed the extraction and then started splitting for the multiprocess with no error, so it has certainly gotten past my precious problem area. Thanks.

8th February 2012, 03:18	#14270 \| Link
AmigaFuture Registered User Join Date: Apr 2003 Location: Within the main Source. Posts: 895	I get it about the resolution differences and that to see higher FPS with higher resolution you need to tap more power (vibration) to get higher FPS...or "seemingly" have multiple processes of a particular program. Which I noticed DVD Rebuilder didn't have to (multiple processes of HCEnc) to get higher FPS. The last encoding wasn't split, but I verified it anyway and it was smooth. Reading your last post to me...I though I'd not activate the external LAVF and try that since I haven't for several past tests. __________________ Life is not a journey to the grave; but rather to skid out broadside, thoroughly used, torn and warn and loudly proclaim; WOW; What a ride!!! Soon, I'm going to do it AGAiN in different skin!!

8th February 2012, 06:19	#14272 \| Link
AmigaFuture Registered User Join Date: Apr 2003 Location: Within the main Source. Posts: 895	My intention wasn't to express you're not aware of your own program, but to help me further understand the Encoding process and speeds, and "Is there a Bug here??". Now I get it about the version of HCEnc I use which answers other question(s) a bit more even about BD-RB. :-D I Appreciate your reply, Thank You. :-) Please keep the "Automatically crop black borders" option. I'm cool with using slower speeds for that. Fast is good, but slow has it's...benefits also. ;-) __________________ Life is not a journey to the grave; but rather to skid out broadside, thoroughly used, torn and warn and loudly proclaim; WOW; What a ride!!! Soon, I'm going to do it AGAiN in different skin!! Last edited by AmigaFuture; 8th February 2012 at 08:49.

8th February 2012, 13:35	#14273 \| Link
colinhunt Registered User Join Date: Dec 2002 Posts: 1,022	Anyone done The Thing (2011) with PiP video/DTS Express enabled? I ran it on 0.40.02 (multiprocess disabled) and the output plays PiP video without audio on PowerDVD 11. The original plays fine, audio and all, on PDVD11. Slight clarification: when PiP video window opens and the secondary audio is activated, PDVD11 lowers the volume level of the main audio. On the backup, main audio level is lowered but secondary audio remains mute. Also, on the original PDVD11 shows "DTSHD" as the codec of 2nd audio, along with the bitrate of the secondary audio. On the backup, only "DTSHD" is shown on the information OSD; bitrate is not shown.

8th February 2012, 18:04	#14279 \| Link
colinhunt Registered User Join Date: Dec 2002 Posts: 1,022	What's the working relationship of LAVF and DGDecNV? Does one override the other or..?