[Vorbis] Multi-Track-Single-Logical-Stream Ogg-Vorbis (mogg) vs Multi-Track-Multi-Logical-Stream Ogg-Vorbis

[Conrad Parker]

On 18 May 2013 00:52, Bob Ingraham <bobi at ingrahams.us> wrote:
> Hi All,
>
> I am in the midst of developing a library for easily recording multi-track
> audio for a court recording system. Well, it can be used for anything, but
> this is its first intended use.
>
> The idea is to simultaneously capture/compress/archive 8 to 16 mono
> sources (microphones) in an Ogg container using Vorbis. Then later, be
> able to play back only selected logical streams, or
> multiple-logical-streams via down-mixing.
>
> Now, the way I've written the encoding logic (with liboggz and fishsound,)
> is to compress each mono source (microphone) with Vorbis and store it in
> Ogg as its own logical stream.
>
> This works is currently working as I had hoped, because I end-up with an
> Ogg file with 8 logical streams, one
> Vorbis-encoded-logical-stream-per-microphone.
>
>
> THE ISSUE:
>
> So, I've now seen "mogg" files and have examined one and found out that
> they do NOT have multiple logical streams but INSTEAD have multiple
> channels in a single logical stream.
>
>
> So I am confused: I thought that Vorbis only could encode up to 2 channels
> max (stereo), but I've found, for example, a 12-channel Vorbis-encoded
> song in a single-logical-stream Ogg container.
>
> The whole purpose of me going to the pain of separately encoding each of
> my mono channels and then stuffing each in its own logical stream was to
> preserve channel separation for later editing and/or selective
> down-mixing.
>
> THE QUESTIONS:
>
> 1. So, did I NOT need to compress each mono channel separately with its
> own Vorbis instance and use separate logical streams in Ogg?
>
> 2. For preserving individual track integrity (and editability and
> down-mixing,) which approach is better? (mogg vs multi-logical-stream)?
>
>
> 3. How does Vorbis handle multiple mono tracks (say 8+)? Does it encode
> each channel separately and independently?
>
>
> I guess I'm asking, given what I'm trying to achieve, if I chose the
> correct approach (encode each track independently and stored in its own
> logical stream,) or if I just made extra work for no reason?
>
> Thanks for any insight from the experts!

Hi,

could you paste the output of oggz-info on one of those files?

Conrad.