[vorbis] joint stereo implementation

[Roland Nagtegaal]

Karol Pietrzak wrote:
> 
> On 14 Jun 2001, Segher Boessenkool wrote:
> 
> > A good joint stereo algorithm can throw away about 40% of the bits,
> > on "average" music streams.  So 128kbps -> 80kbps, 96kbps -> 64kbps.
> 
> Out of curiousity, how will the encoder handle this?  I mean, if
> "-b 128" is specified to oggenc, will it lower bitrate hover
> around ~80kbps (same quality as current encoder) or will it
> hover around ~128kbps (with "40% better quality", so to speak)?
> --
> noodlez:   Karol Pietrzak
> GPG/PGP-KeyID: 0x3A1446A0
> 
> --- >8 ----
> List archives:  http://www.xiph.org/archives/
> Ogg project homepage: http://www.xiph.org/ogg/
> To unsubscribe from this list, send a message to 'vorbis-request at xiph.org'
> containing only the word 'unsubscribe' in the body.  No subject is needed.
> Unsubscribe messages sent to the list will be ignored/filtered.

I have a question:

When doing channel coupling with N channels using relative strengths of
spectral bands (at least that how I understand it works), would it also
be
possible to have relative time-shifts, in order to enhance spatial
resolution?

Say, you have a guitar playing at 2 o'clock, then there is a difference
of 
0.3 millisec between sound arriving left and right when using two
channels.

I think you can calculate such differences by doing a convolution
integral, 
or fft the signals and simply multiply. (we do this in radio astronomy)

Maybe a system that does channel coupling in amplitude per freq. band
and in
temporal difference per freq. band is not patented yet?

Roland

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.