Rendering of Source Spread for Arbitrary Playback Setups Based on Spatial Covariance Matching

This paper proposes an algorithm for rendering spread sound sources, which are mutually incoherent across their extents, over arbitrary playback formats. The approach involves first generating signals corresponding to the centre of the spread source for the intended playback setup, along with decorrelated variants, followed by defining a diffuse spatial covariance matrix for the confined target spreading area. The mixing matrices required to combine these signals, in a manner whereby the resulting output signals exhibit the target inter-channel relationships for an incoherently spread source, are computed based on an optimised solution which is constrained to preserve signal fidelity. The proposed solution is evaluated in the context of producing extended sound sources for binaural playback. Objective perceptual metrics are computed and shown to be comparable to those derived from an ideal incoherently spread reference. Signal distortion measures are also calculated for speech, musical, and ambience recordings, which indicate higher signal fidelity produced by the proposed constrained spatial covariance matching solution, compared to an unconstrained alternative. These improvements in signal fidelity are further demonstrated by the provided audio examples and open-source audio plug-in.