Parametric Time-Frequency Domain Spatial Audio

3 Sound Field Analysis Using Sparse Recovery

Craig T. Jin, Nicolas Epain, and Tahereh Noohi

CARLab, School of Electrical and Information Engineering, University of Sydney, Australia

3.1 Introduction

The analysis or decomposition of a sound field into plane-wave source signals and their directions has many uses, such as in sound source tracking, direction of arrival estimation, source separation, sound field reproduction, and architectural acoustics. The history of microphone array processing is replete with numerous methods for sound field analysis. For example, there are simple beamforming methods that can provide an energy map of the sound field (Gover et al., 2002; Park and Rafaely, 2005). There are numerous time difference of arrival (TDOA) methods based on the cross-correlation of microphone signals (for a detailed review, refer to Benesty et al., 2008). There are also more elaborate techniques such as multiple signal classification (MUSIC; Schmidt, 1986) and the estimating signal parameters via rotational invariance technique (ESPRIT; Roy and Kailath, 1989), which have recently been applied to spherical microphone arrays (see EB-MUSIC: Rafaely et al., 2010; EB-ESPRIT: Sun et al., 2011). There are also direction-of-arrival techniques based on multi-dimensional maximum-likelihood estimation which have recently been proposed for spherical microphone arrays (Tervo and Politis, 2015). The previous chapter of this book presents a PWD operation based on robust signal-independent ...