Investigation on the reproduction performance versus acoustic contrast control in sound field synthesis

Abstract

A sound reconstruction system is proposed for audio reproduction with extended sweet spot and reduced reflections. An equivalent source method (ESM)-based sound field synthesis (SFS) approach, with the aid of dark zone minimization is adopted in the study. Conventional SFS that is based on the free-field assumption suffers from synthesis error due to boundary reflections. To tackle the problem, the proposed system utilizes convex optimization in designing array filters with both reproduction performance and acoustic contrast taken into consideration. Control points are deployed in the dark zone to minimize the reflections from the walls. Two approaches are employed to constrain the pressure and velocity in the dark zone. Pressure matching error (PME) and acoustic contrast (AC) are used as performance measures in simulations and experiments for a rectangular loudspeaker array. Perceptual Evaluation of Audio Quality (PEAQ) is also used to assess the audio reproduction quality. The results show that the pressure-constrained (PC) method yields better acoustic contrast, but poorer reproduction performance than the pressure-velocity constrained (PVC) method. A subjective listening test also indicates that the PVC method is the preferred method in a live room.

Causal MSE-Optimal Filters for Personal Audio Subject to Constrained Contrast

“…Note that the desired system behaviour is introduced into the pre-compensator (7) exclusively via the causal factor Q(q −1 ). The Diophantine equation (10) separates the causal part, Q(q −1 ), of the compensation from the anti-causal part L * (q), which is not used in the causal solution. It can be shown that letting d 0 → ∞ produces the (time-shifted) noncausal Wiener solution, in which L * (q) = 0.…”

“…2. The longest common delay, or modelling delay, d 0 is arbitrarily chosen to 10 ms. 10 Causally constrained designs with acoustic contrast greater than or equal to −10, +20 and +50 dB over the entire audible spectra are computed based on the system described above. The latter contrast is chosen in order to push the filters beyond what they can be expected to deliver for the investigated system.…”

Generation of Private Sound With a Circular Loudspeaker Array and the Weighted Pressure Matching Method

“…More generally, a performance trade-off exists between directivity performance, input energy required by the array to perform directional sound radiation, and accuracy of reproduction of the target field in the listening area [18], [12], [23], [24], hereafter succinctly referred to as "quality". For example, for a given level of the input energy, a given system may be able to provide high directivity performance at the expense of a reduced quality in the listening zone [9], [10], [12].…”

“…The input signals may be designed by means of beamforming methods, such as the Time-Reversal Mirror [3], sound focusing [18], the Minimum Variance Distortionless Response [2], as well as methods that aim at the control of the acoustic energy density in the control area [6], [7], [4], [8], hereafter also referred to as energy-based methods, and methods that aim at the synthesis of a target field defined at a set of control points, such as the Pressure Matching Method (PMM) [17], [8] and the Weighted Pressure Matching Method (WPMM) [13], [9]. In this work, we limit our analysis to the PMM and the WPMM, as they may provide better sound quality than that achieved with energy-based methods [8], [24], [9], [18].…”

“…In this case, we consider only the bright point. The input signals q MQS are calculated by means of a sound focusing technique [18], that is…”