On the Analysis and Synthesis of Local Sound Fields for Personal Audio

The aim of multizone sound reproduction is to independently synthesize multiple acoustic scenes in different areas of the reproduction environment. Thereby, individual audio contents may be simultaneously provided to multiple persons. In this thesis, a novel algorithm for this purpose is derived, and practical aspects, such as imperfections of the utilized transducer arrays or sensor noise, are investigated. Furthermore, an iterative filter design technique with a spatiotemporal weighting function is presented, which is especially beneficial in the case of reverberant reproduction environments. As high-quality sound reproduction is closely related to active noise control, this work also contributes to the analysis of noise fields evoked by spatially extended sources.