Low Delay Audio Streaming for a 3D Audio Recording System

This paper presents a prototype of a 3D audio recording system named AudioSense which uses Wireless Acoustic Sensors to capture spatial audio. The sound is recorded in real-time by microphones embedded in each sensor device and streamed to a Processing Unit for 3D audio compression. One of the key problems in systems which stream audio data is end-to-end latency. This paper is focused on analyzing a set of chosen parameters of the Opus codec in order to obtain the minimal delay. Experimental results on the prototype system have shown that it is possible to achieve below 10ms of end-toend audio delay with the use of the Opus codec.