Ambisonic decoding

Abstract: Ambisonics with height is a three-dimensional sound field reproduction technique for spherical loudspeaker arrangements surrounding the reproduction area. It employs spherical harmonics up to a given order to expand incident sound fields with a limited angular resolution. The expansion coefficients describe the spatial sound scene. For reproduction, these coefficients are decoded to a set of surrounding loudspeakers. Common decoding approaches either sample the spherical harmonic excitation at the loudspeaker positions or match the excitation modes of a continuous sound field to those of the loudspeakers. For well-designed spherical loudspeaker arrays, both decoding approaches achieve good perceptual localization of virtual sound sources. However, both approaches perform unsatisfactorily with non-uniformly arranged arrays. Sounds from directions with only sparse loudspeaker coverage appear with altered loudness levels. This distracting effect results from variations in the decoded energy. The present article demonstrates an improved decoding technique, which preserves the decoded energy. Using available objective estimators, the localization qualities of these energy-preserving decoders are shown to lie between both common decoding approaches.

Abstract: Accurate decoding and most notably mode-matching has always been a matter of concern within the Ambisonics community; it can be further expected to play a major role in future discussion and research. Specifically, ambisonic decoding by mode-matching attempts to perfectly reconstruct incident sound fields using a surrounding spherical arrangement of loudspeakers. This reproduction is valid locally within a bounded central area, the sweet-area. Surprisingly, many (experienced) listeners have reported good reproduction quality and fair localization accuracy of real-world Ambisonics systems even outside this sweet area. Hence in practice, mode-matching decoding performs better than expected from theory, but it frequently poses numerical problems for incomplete or non-uniform spherical loudspeaker arrangements. To cope with these issues, this article presents, discusses, and compares several alternative Ambisonics decoding strategies with and without mode-matching in terms of simple quality criteria.

Abstract: Ambisonics, a sound field synthesis and reproduction technique, has shown promising results in conveying three-dimensional spatialized sound. Ambisonic encodings directly describe the spatial properties of sound fields without reference to the reproduction system. Precise regeneration of a sound field requires a large number of loudspeakers arranged so as to adequately sample all directions; this is referred to as a regular layout. It is not difficult to find a decoding matrix to reproduce Ambisonic recordings using uniformly distributed loudspeakers. Nevertheless, evenly locating a large number of loudspeakers is not feasible in most scenarios. Irregular arrays, however, are known to lead to ill-conditioned and singular re-encoding matrices. We propose a method for accurate reproduction of high order Ambisonic recordings over irregular loudspeaker arrays. Our approach consists of three stages which aim to exploit any regularities of the array, while using asymmetrically located loudspeakers to expand the listening volume. We evaluated our proposal using an irregular 157-channel loudspeaker array. Comparisons with mainstream decoding methods were conducted. The proposed scheme results in an overall increase in the size of the listening volume.