Vorbis - Jul 2009 - Square Polar Stereo Coupling

Hello,

I've read http://xiph.org/ogg/vorbis/doc/stereo.html and the Vorbis specs,
but I haven't got a clear concept of how does the Square Polar Mapping
works. I mean, in the "Ogg Vorbbis stereo-specific channel coupling
discussion" document, it's described how the decoupling process works
for
Square Polar Mapping (and it is fairly clear explained), but it does not
describe how the coupling process works. I've found another thread of May
2003 asking the same question but it neither gives a clear answer, and
googling it I've had the same (no) success.

In short, what I'd like to know is how, having several (two for simplicity)
uncoupled channels, how does the square polar mapping does to get the
associated square polar notation?

Thanks.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.xiph.org/pipermail/vorbis/attachments/20090716/fc62e3a7/attachment.htm

On Thu, Jul 16, 2009 at 12:36 PM, Jes?s D?az
Vico<jesus.diaz.vico at gmail.com> wrote:
> Hello,
>
> I've read http://xiph.org/ogg/vorbis/doc/stereo.html and the Vorbis
specs,
> but I haven't got a clear concept of how does the Square Polar Mapping
> works.
The idea is to turn two magnitude vectors into a magnitude vector with
most of the energy and an angle vector with much smaller values.

With more than two channels, the coupling happens in a tree until, for
example, you end up with one magnitude vector and n-1 angle vectors.

Monty