I using z270 camcorder with 4 channel (HDV format) audio recording that connect 4 wireless microphones.
It is very helpful for one man shooting like me with no soundmen.

But there are very BIG PROBLEM when edit the footages: NO SOFTWARE CAN READ 4 CHANNEL AUDIO IN MY WINDOWS PC.

I using Edius 5.11 for my editing tool, Edius cannot read 4 channel Audio,
so I try another software like, Premiere, Soundforge, Vegas, Windows Media Player(also WMP classic) etc,
All of them show that the M2T file only has “Mpeg1 Layer2 Stereo” Audio channel (it’s means there are 2 channel only)

Inside S270 manual, Sony says the Audio of HDV 4 channel recording using Mpeg2 Layer2 format, I asked Japan Sony
But they do not give me any suggestion.

I know Mpeg2 Layer2 multichannel Audio is the extension of Mpeg1 Layer2 audio. but I don’t know is it a codec problem or something else.

Now, I using a very primeval method to capture my 3 and 4 channel audio:

1. Use M2T files to editing (M2T files which recorded on the CF card)
2. 1, 2 Channel are OK
3. Find out the DV TAPE footage same as M2T file
4. Use S270 Down convert function, Set iLink output 3,4 channel audio to DV format
5. Connect IEEE1394 to PC than capture the 3,4 audio

I had tried capture tape HDV footage use IEEE1394 to my PC , but after capture, the files also has 2 channel only.

Is there any good idea? I don’t want to use this kind of primeval method cause its waste my time to capture the footage again.

Thanks

Herro

I'm having the same problem. I've got a Canon XL H1... I recorded using four channels. I recorded to both tape and a Sony DR60 Hard drive.

If I play the hard drive back to the camera or another camera (Sony Z1, V1, FX1), the cameras see the extra channels and allow me to listen to them and mix them. This proves that the data is sent via firewire and is present in the files I have on my hard drive. It also indicates that the stream is compliant with the HDV spec.

The data may be present in the files on my PC... but no editing software seems to be able to read the additional channels. I've tried Premiere CS4, Final Cut, Cinelerra, VLC, Vegas, TMPGENC... how frustrating... I am going to go and read the HDV specification now and write a programme to extract this damn audio.

Why is it that such a cool function like four channel audio has only been implemented in the hardware and no where else.

*howls*

:(

SONY never talk about this! I just find another small program to read a source, but nothing finding until now, if i found, i will post here!

HDV 4 channel audio capture [Archive] - Grass Valley Desktop Solutions Forums (http://ediusforum.grassvalley.com/forum/archive/index.php/t-363.html)

Apple - Support - Discussions - Canon HDV XL H1 4 channel audio ... (http://discussions.apple.com/thread.jspa?messageID=9952934)

(Appologies if pointing "outside' is a no - no, can't find an answer here).


CS

Cool Hirata.

Chris... it shouldn't be necessary to recapture the footage using anything other than firewire.

Firewire carries a 25mb/sec data stream. Inside the stream is an MPEG2 Transport stream and inside that is a video stream and an audio stream. The bitrate of the video stream doesn't change, neither does the bitrate of the audio stream (384kbps). When you switch to four channels you simply divide the available bandwidth up between four channels instead of the normal two. When you capture footage using firewire, the entire stream is sent... including all four channels.

It shouldn't be/isn't necessary to re-capture anything. If I stream my m2t HDV files back to my camera, I can select and listen to the 3rd and 4th channels or the 1st and 2nd channels or I can choose to mix them... I can define how much of each I want to hear. Not only that... but streaming my Canon's m2t files back to a Z1, V1 and FX1 via firewire also works... those cameras allow me to select the channels I want to hear... those aren't even four channel cameras.

MPEG2 uses payload IDs inside the TS... and it seems that all four channels are held within the same PID instead of being broken into two seperate PIDs... I'm going to try and split the data a little later today by using a demuxer and then a specialised second demuxer. Standard behaviour is to put mono or stereo tracks into seperate PIDs, HDV doesn't seem to do this. The software only expects two channels and doesn't have any more PIDs to work with... so it drops the extra data.


UPDATE :
There's a reason the specification states that the codec used changes from MPEG1 Layer I to MPEG2... and that's because MPEG1 doesn't support multi-channel.

"
MPEG-2/Multichannel has the following properties:
# up to 5 full range audio channels and an LFE-channel (Low Frequency Enhancement <> subwoofer)
# sample rates the same as those of MPEG-1
"
MPEG AUDIO FRAME HEADER (http://www.mpgedit.org/mpgedit/mpeg_format/mpeghdr.htm)


UPDATE :

Demux complete and I've found the problem!

The frame header coming out of the camera is set incorrectly.

My camera is sending this : 11111111 11111100 11100100 00000100
.......................................SYNC BYTE..........................channels

The two bolded bits define the channel structure as stereo... which is why editing applications drop the additional channels. It should be possible to fix the problem by changing the frame headers.


UPDATE :

This is evil!!!!

From WikiPedia : http://en.wikipedia.org/wiki/MPEG_Multichannel

"
MPEG Multichannel is an extension to the MPEG-1 Layer II audio compression specification, as defined in the MPEG-2 standard ISO/MPEG IS-11172,[1] which allows it provide up to 5.1-channels (surround sound) of audio. To maintain backwards compatibility with the older 2-channel (stereo) audio specification, it uses a channel matrixing scheme, where the additional channels are mixed into the two backwards compatible channels. Extra information in the data stream (ignored by older hardware) contains signals to process extra channels from the matrix.[2]"

This explains everything!

Turns out some guy made a programme in 2003 for Linux that can encode and decode MPEG2 multichannel audio streams.

After a bit of tweaking and compilation :

"
A MPEG 2 Layer-1/2 bitstream will be decoded
Output file written in AIFF format
the bit stream file audio.mpg is a BINARY file
HDR: s=FFF, id=1, l=2, ep=0, br=E, sf=1, pd=0, pr=0, m=0, js=0, c=0, o=1, e=0
layer=II, MPEG1-bitrate=384, sfrq=48.0, mode=stereo, sblim=27, jsbd=27, MPEG1-ch=2
mc-channel=2, MPEG2-ext-length=0 kbit/s
initialising 4 output files
"

and it's found the 4 channels...

Okay... more tweaking... the use of an API for PCM encoding... etc...

I now have four files for every m2t file on my hard drive... channel 1, 2, 3 and 4.

And it sounds perfect.


Why has no one else has done this ?


Next step is to build it into an easy to use application that can do entire folders filled with m2t files.

yay, lunch time!

HI, Aaron Scheiner:

You are a star! How you can know so much thing inside the file, enviable!!

>>>Turns out some guy made a programme in 2003 for Linux that can encode and decode MPEG2 multichannel audio streams.<<<

>>>Okay... more tweaking... the use of an API for PCM encoding... etc...
I now have four files for every m2t file on my hard drive... channel 1, 2, 3 and 4.
And it sounds perfect.<<<<

Simplely, how can I get the 3,4 channel sound, do I need a Linux system?

HI, Aaron Scheiner:

I follow your advise to check the API for PCM encoding, but nothing found, will you please give me some more hints. Thanks

I use Blackmagic HD Extreme capture card, HD-SDI interface, embedded audio. The motion JPG avi files give me 2 stereo tracks and Sony Vegas 10c has no issues what-so-ever. I never had any luck with 4 ch audio from firewire and Sony Vegas.

See if this helps.

http://www.dvinfo.net/forum/what-happens-vegas/143038-multi-channel-audio-capture.html