Spatial sound pick-up with a low number of microphones

Abstract

Portable audio devices have become more and more popular during the last decade. Recent advances in audio compression and electronic miniaturization allow people to keep all their music and movies at hand, at all times. Nowadays people are using their smartphones everywhere as photo or video cameras. In order to provide a consumer possibility of 3D audio recording adapted to these kinds of devices, we developed a compact microphone array able to pick-up a full 3D sound scene, using less than four microphones. To compensate for the low number of microphones, which results in poor spatial selectivity, spatial post processing is applied to the microphone signals and improves the sound source localization. Another advantage is that the post processing makes the sound reproduction flexible. The 3D audio scene can be converted into any format to be rendered to any equipment and any device. The paper will describe various recording set-ups in combination with the associated post processing. As a first assessment, their performances in terms of sound localization accuracy will be compared using a new set of objective criteria descriptors.