This paper presents a new paradigm for spatial audio teleconferencing that exploits the corpuscular nature of general audio parametric coding for low bit-rate audio compression. The paradigm consists of several processing tools. The spatial aspect is enacted by a dynamic rendition tool that optimizes placement of speech sources within the soundscape; the temporal aspects of a phoneme protection tool misalign corruptive and vulnerable speech elements in a multitalking environment. A novel means of echo suppression eliminates the need for adaptive filtering and real-time echo cancellation, which is difficult for multichannel audio.
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.