Jan writes:
>The abillity to perform quad sound with some
>multichannel card is one thing... But what I was
>heading for is 3d spacial audio (height included) as
>discribed in the ambisonics homepage, which is more
>than distribution of sounds in space just by a quad
>mix, based on amplitude (like panning in the stereo-
>field):

Ambisonics uses 4 channels (WXYZ) to _encode_ a periphonic (3d) soundfield.
The Z component encodes the height information. To decode ambisonics you
need at least 8 speakers to effectively reproduce height information + a
suitable decoder (this can be done with an 8 buss desk). The frequency
dependent component of ambisonics is used only in the decoder  to improve
localisation, _not_ to synthesize height information without extra speakers
to reproduce the height information.

So your options for csound are:

1. Create a four channel B-Format (WXYZ) file and decode it with separate
software / hardware.
2. Perform 1., then write a separate csound orc to decode bformat into
however many speakers/sound card outputs you have.
3, Perform 1&2 at once (as in my previously posted example).

I believe that it is theoretically possible to convert an ambisonic signal
to binaural using HRTFs for headphone playback, but I'm not sure. I guess it
might be possible to synthesize/embed the height info in a quad decode -
best to ask on the sursound list.

Ross.