[LAU] Concert with low delay realtime linux audio spatial processing

Fons Adriaensen fons at linuxaudio.org
Thu Sep 22 15:01:14 UTC 2011

On Thu, Sep 22, 2011 at 08:38:16AM +0200, Giso Grimm wrote:
> I plan to record the microphone signals, from which I can then recreate
> the Ambisonics mix. However, it is difficult to create a stereo downmix
> of this (I tried a headphone downmix via dummy head HRTFs, which is fine
> but don't allow you to turn your head, I don't think it is possible to
> create acceptable loudspeaker downmixes).

You could try the virtual stereo mic in Tetraproc. Usually a pair of
hypercardioids at an angle of 120 degrees or so gives good results.
Of course you get some 'front preference', with back sources becoming
more reverberant. But that works better than trying to preserve 
'uniform' covering of all directions in stereo.

Anyway, good luck with the concert !

Yesterday evening we had an 'electronic concert' in the Sala Bianca.
One of the pieces was produced at the LABEL (studio of la Casa della
Musica). It uses 38 (Ardour) tracks, spatialised into 3rd order AMB
in the studio. For the concert the panning data from the ambisonic
mixer were transferred to the 'WFS mixer' (which doesn't actually
mix the signals, it just controls the position of sources in the WFS
system), and the 38 tracks rendered individually. We could also
render the Ambisonic version using just 8 virtual speakers. Both
versions worked well. Those who heard both (before the concert)
preferred the WFS version. 



More information about the Linux-audio-user mailing list