Spatial sound field synthesis and upmixing based on the equivalent source method

Abstract

Given scarce number of recorded signals, spatial sound field synthesis with an extended sweet spot is a challenging problem in acoustic array signal processing. To address the problem, a synthesis and upmixing approach inspired by the equivalent source method (ESM) is proposed. The synthesis procedure is based on the pressure signals recorded by a microphone array and requires no source model. The array geometry can also be arbitrary. Four upmixing strategies are adopted to enhance the resolution of the reproduced sound field when there are more channels of loudspeakers than the microphones. Multi-channel inverse filtering with regularization is exploited to deal with the ill-posedness in the reconstruction process. The distance between the microphone and loudspeaker arrays is optimized to achieve the best synthesis quality. To validate the proposed system, numerical simulations and subjective listening experiments are performed. The results demonstrated that all upmixing methods improved the quality of reproduced target sound field over the original reproduction. In particular, the underdetermined ESM interpolation method yielded the best spatial sound field synthesis in terms of the reproduction error, timbral quality, and spatial quality.

Solution Strategies for Linear Inverse Problems in Spatial Audio Signal Processing

“…The second scenario was concerned with analyses and syntheses of sound fields. Recently, an integrated array system was developed on the basis of a freefield model for spatial audio recording and reproduction [8,9]. This study extended the previous work to a reverberant environment; a live room was fitted with reflective walls.…”

“…Another example is the source signal separation problem, where an individual source signal is to be extracted from a mixed array of signals [7]. Inverse problems can also be found in source sound field synthesis (SFS) problems, where the sound field produced by secondary sources is to be matched with a target field [8,9]. Other examples include sound field control [10,11], crosstalk cancellation in binaural audio rendering [12], noise reduction in speech enhancement [13], room response equalization, and dereverberation [14,15].…”

Sparse Plane Wave Approximation of Acoustic Modes to Address Basis Mismatch

“…The discretized PW directions, S = {s r } R r=1 , are uniformly sampled on the surface of a unit sphere. Generally, R > M, and w can be solved as wLS = HH ( H HH ) −1 ψ, (3) where (•) H denotes the conjugate transpose operation, which is the analytic solution of the underdetermined least-squares (LS) evaluated using the Tikhonov regularization. However, this solution tends to generate spatial aliasing artifacts.…”

“…Recently, the reconstruction of a low-frequency sound field in an enclosed space has generated interest among researchers. It is significant for many applications, including sound field auralization [1], room equalization [2], sound field control [3][4][5], etc. The reconstruction problem usually involves interpolating the impulse responses at unmeasured positions from the impulse responses measured by a discrete set of microphones.…”

Location Optimization of Monopole Equivalent Sources in Wave Superposition Method

“…WSM has been applied to calculation of acoustic radiation. [15][16][17] However, the prediction accuracy of acoustic pressure using this method strongly depends on the locations of the equivalent sources, particularly for the radiators with complex geometric shapes. Determining the locations becomes a problem.…”