Panaudia is an ambisonic mixer and a binaural renderer


Ambisonics, invented by Michael Gerzon and others at the University of Oxford in the early 1970s, is a technique for encoding three-dimensional sound fields. It was a technology ahead of its time and lost out commercially to simpler surround sound techniques that were easier to work with in cinemas.

Today ambisonic audio is increasingly popular. Technology is catching up, and when paired with binaural rendering it provides a way to offer individualised, immersive audio for users of AR and VR spaces.

Panaudia is both an ambisonic mixer and a binaural renderer. It takes in audio and a small amount of positional metadata from every user, and does an individualised second degree ambisonic mix for each of the others in the space. Then, taking into account head rotation, it renders this into binaural stereo and returns this to the user.

Panaudia's mixing maintains highly accurate spatialisation locally for each user. To allow it to scale to large numbers Panaudia uses a number of optimisations that reduce the accuracy of the rendering of distant, fainter, sounds. For a few hundred users our rendering engine can create accurately spatialised many-to-many mixes for each user, of every other user in a Space. As the numbers go up into the high hundreds and thousands the rendering of more distant users will become slightly less accurate.