DV tape capture audio conversion issues

[mcb7]

I originally posted this on the Videohelp conversion board but didn't get a satisfactory response (eg use a different program) so trying here in the hope I can solve the issue since the picture quality I've had from Hybrid is better than any other tool I've tried.

I have captured a number of DV tapes using my Macbook Air via the firewire interface and I am wanting to convert/compress them into mp4 files to share online. I have copied the files to my Windows PC and have tried using Hybrid with default settings to convert them but I am getting an issue with the audio.

The audio seems to be stretching out to double the length of the original clip. If I do the conversion without audio the video file is the correct length but if I add audio, the file runs for twice the length but the video is running at normal speed but stops halfway through and the audio seems to be running half speed.

Is there some settings I can change to fix this?

TIA

Since that post I have tried conversion to avi container to see if passing through the audio as pcm would help but it didn't - same result. I'm attaching the HybridDebugOutput.txt for the mp4 attempt.

I included the MeadiaInfo for the source file in the original post, if that's of any use: https://forum.videohelp.com/threads/4102...ith-Hybrid
Please, read the 'Infos needed to fix&reproduce bugs,..'-sticky before you post about a problem.

[Selur]

MediaInfo reports:

Audio #1
Count                                    : 287
Count of stream of this kind             : 3
Kind of stream                           : Audio
Kind of stream                           : Audio
Stream identifier                        : 0
Stream identifier                        : 1
StreamOrder                              : 2
ID                                       : 3
ID                                       : 3
Format                                   : PCM
Format                                   : PCM
Commercial name                          : PCM
Format settings                          : Little / Signed
Format settings, Endianness              : Little
Format settings, Sign                    : Signed
Codec ID                                 : lpcm
Duration                                 : 364757
Duration                                 : 6 min 4 s
Duration                                 : 6 min 4 s 757 ms
Duration                                 : 6 min 4 s
Duration                                 : 00:06:04.757
Duration                                 : 00:06:04.757
Source duration                          : 364717
Source duration                          : 6 min 4 s
Source duration                          : 6 min 4 s 717 ms
Source duration                          : 6 min 4 s
Source duration                          : 00:06:04.717
Source duration                          : 00:06:04.717
Bit rate mode                            : CBR
Bit rate mode                            : Constant
Bit rate                                 : 2048000
Bit rate                                 : 2 048 kb/s
Channel(s)                               : 4
Channel(s)                               : 4 channels
Channel positions                        : Front: L R
Channel layout                           : L R L R
Sampling rate                            : 32000
Sampling rate                            : 32.0 kHz
Samples count                            : 11672224
Bit depth                                : 16
Bit depth                                : 16 bits
Delay                                    : 280
Delay                                    : 280 ms
Delay                                    : 280 ms
Delay                                    : 280 ms
Delay                                    : 00:00:00.280
Delay                                    : 00:00:00.280
Delay_DropFrame                          : No
Delay, origin                            : Container
Delay, origin                            : Container
Delay relative to video                  : 0
Delay relative to video                  : 00:00:00.000
Delay relative to video                  : 00:00:00.000
Stream size                              : 93377744
Stream size                              : 89.1 MiB (7%)
Stream size                              : 89 MiB
Stream size                              : 89 MiB
Stream size                              : 89.1 MiB
Stream size                              : 89.05 MiB
Stream size                              : 89.1 MiB (7%)
Proportion of this stream                : 0.06619
Source stream size                       : 93367512
Source stream size                       : 89.0 MiB (7%)
Source stream size                       : 89 MiB
Source stream size                       : 89 MiB
Source stream size                       : 89.0 MiB
Source stream size                       : 89.04 MiB
Source stream size                       : 89.0 MiB (7%)
Source_StreamSize_Proportion             : 0.06618
Title                                    : Core Media Audio
Encoded date                             : 2023-05-21 10:39:33 UTC
Tagged date                              : 2023-05-21 10:45:38 UTC
Source_Delay                             : 320
Source_Delay_Source                      : Container

Audio #2
Count                                    : 285
Count of stream of this kind             : 3
Kind of stream                           : Audio
Kind of stream                           : Audio
Stream identifier                        : 1
Stream identifier                        : 2
ID                                       : 1-0
ID                                       : 1-0
Format                                   : PCM
Format                                   : PCM
Commercial name                          : PCM
Format settings                          : Big / Signed
Format settings, Endianness              : Big
Format settings, Sign                    : Signed
Muxing mode                              : DV
Muxing mode, more info                   : Muxed in Video #1
Duration                                 : 364760
Duration                                 : 6 min 4 s
Duration                                 : 6 min 4 s 760 ms
Duration                                 : 6 min 4 s
Duration                                 : 00:06:04.760
Duration                                 : 00:06:04.760
Bit rate mode                            : CBR
Bit rate mode                            : Constant
Bit rate                                 : 768000
Bit rate                                 : 768 kb/s
Encoded bit rate                         : 0
Encoded bit rate                         : 0 b/s
Channel(s)                               : 2
Channel(s)                               : 2 channels
Sampling rate                            : 32000
Sampling rate                            : 32.0 kHz
Samples count                            : 11672320
Bit depth                                : 12
Bit depth                                : 12 bits
Delay                                    : 280
Delay                                    : 280 ms
Delay                                    : 280 ms
Delay                                    : 280 ms
Delay                                    : 00:00:00.280
Delay                                    : 00:00:00.280
Delay_DropFrame                          : No
Delay, origin                            : Container
Delay, origin                            : Container
Delay relative to video                  : 0
Delay relative to video                  : 00:00:00.000
Delay relative to video                  : 00:00:00.000
Stream size                              : 35016960
Stream size                              : 33.4 MiB (2%)
Stream size                              : 33 MiB
Stream size                              : 33 MiB
Stream size                              : 33.4 MiB
Stream size                              : 33.39 MiB
Stream size                              : 33.4 MiB (2%)
Proportion of this stream                : 0.02482
Encoded stream size                      : 0
Encoded stream size                      : 0.00 Byte (0%)
Encoded stream size                      :  Byte0
Encoded stream size                      : 0.0 Byte
Encoded stream size                      : 0.00 Byte
Encoded stream size                      : 0.000 Byte
Encoded stream size                      : 0.00 Byte (0%)
StreamSize_Encoded_Proportion            : 0.00000

Is it correct that the audio is 4 channel? (I ask, since that is rather uncommon for dv video, especially with just 32kHz sampling rate)
Strange is that mediainfo shows 3 audio channels, while Hybrid seems to detect only one.
Okay, Hybrid assumes there is only 1 audio channel, since mplayer only detects one.
mplayer reports a bunch of '[dvvideo @ 00007ff704aa6000]Concealing bitstream errors' and reports one time one audio and another time two audio channels.

I assume the different audio count will cause problems later, since it seems like Hybrid tried to extract an audio stream which Hybrid assumes to be 4 channels (same as ffmpeg), but during processing ffmpeg complains:

[pcm_s16le @ 00007ff704aa6000]Channel layout 'stereo' with 2 channels does not match specified number of channels 4: ignoring specified channel layout

=> Do you have a source you can share with me which triggers this problem?
I would need something to play with to come up with a workaround for this.

Cu Selur

[mcb7]

Thanks for looking into this. I've put a smaller clip on my Google drive here (let me know if you can't access):
https://drive.google.com/file/d/1vRPWBMh...drive_link

I have now noticed that VLC is the only thing I have installed that can play the original files correctly, when I was sure Windows Media player was working a few weeks ago. WMP plays with the same audio issue which I don't think is half speed because I played the converted mp4 in VLC at x2 speed and it still sounded strange. I'm wondering if it's the meta data that is wrong, maybe the endian-ness of the samples or incorrect bits/sample or rate? Also strangley, Windows TV & Films app will play the audio correctly but won't recognise the video.

[annael]

Just a little note on Mini DV audio: when using 48khz, DV is only 2 channels, but when set to 32khz, you get 4 channels of audio. The additional channels were reserved for "audio dubbing".

I remember having lots of problems with 32khz audio back in the mini DV days. Most apps would capture audio incorrectly when rewrapping to .mov or .avi. I had to capture to a .dv file or capture via analogue outputs to get good sync. I'm not saying this is the issue here, just an historical FYI ;-)

[mcb7]

Thanks for the info, I might try to see if the capture software I used on the Macbook has any settings.

[Selur]

First: Sorry, totally forgot about this.
Second: Problem is neither ffmpeg, nor mplayer or mencoder handle this audio properly.
Mplayer can properly playback the audio but, like ffmpeg, creates a slowed down track on extraction.

Even a simple:

ffmpeg -i "1999-04-11 20_04_29-1.mov" -vn -acodec copy "test.wav"

produces a slowed down audio output.

Sadly atm. I have no clue how to get a normal audio stream out of this and thus have no clue how to add a workaround for this in Hybrid.
I could playback the file correctly with Audacity, so one could extract the audio as wav using audacity.
Funnily, unlike everything else, Audacity identifies the audio as 4 separate identical mono 32kHz 32bit audio. So maybe the issue is with a wrong handling of the sample rate in ffmpeg.

Cu Selur

[mcb7]

Thanks for looking at this again Selur. 

As it happens I've found a work around - the software on the Macbook that I was capturing with had cut the viedo into clips which was great for the most part but in some cases it had cut the same scene into multiple parts so I used Avidemux to join the .mov files so as not to lose any quality with re-encoding. I took the default mkv container option and it transfers the audio into a format that Hybrid (and everything else) can understand. 

It's not ideal as I now have an extra step to do in the processing, plus I am needing double the disk space but it works.