Spatial Audio Quality Perception (Part 1): Impact of Commonly Encountered Processes

Abstract

Spatial audio processes (SAPs) commonly encountered in consumer audio reproduction systems are known to generate a range of impairments to spatial quality. Two listening tests (involving two listening positions, six 5-channel audio recordings, and 48 SAPs) indicate that the degree of quality degradation is determined largely by the nature of the SAP but that the effect of a particular SAP can depend on program material and on listening position. Combining off-center listening with another SAP can reduce spatial quality significantly compared to auditioning that SAP centrally. These findings, and the associated listening test data, can guide the development of an artificial-listener-based spatial audio quality evaluation system.

Qualitative Evaluation of Media Device Orchestration for Immersive Spatial Audio Reproduction

“…For channel-based transmission, the reproduction format is predefined at the production stage; the channel feeds are transmitted with the expectation that they will be reproduced over the same or a very similar loudspeaker array. The quality of the listening experience is adversely affected when loudspeakers are placed away from the correct, standardized positions [11]. While it is possible to adapt the signals for alternative loudspeaker layouts, methods for doing this have limited flexibility (for example, matrix upmixing [12] or downmixing [13]), or involve significant complexity and a risk of audible artifacts (for example, using source separation or signal analysis and separation techniques [14]).…”

Spatial Audio Quality Perception (Part 2): A Linear Regression Model

“…A previous study [1] made the case for a new artificiallistener-based evaluation system capable of predicting the perceived quality degradations resulting from spatial audio processes (SAPs) commonly encountered in consumer audio multichannel loudspeaker reproduction systems (e.g., downmixing, multichannel coding, loudspeaker misplacement); it explained how such a system would be useful for quickly assessing overall spatial sound quality for research, product development, and quality control where assessment by a listening panel would be impractical or impossible. That study determined the degree of quality degradation resulting from a wide range of such SAPs and the influences of listening position and source material on that degradation.…”

“…A previous paper [1] established the perceived qualities of six spatial audio program items after being processed by up to 48 SAPs and auditioned at two listening positions. The mean (across listeners) quality rating for each combination of program item, SAP, and listening position potentially provided a unique data point for modeling.…”

“…The previous paper [1] suggested that this situation would not be ideal and so the effects of this simplification will be specifically explored in the evaluation section of this paper.…”

“…Genre have a weakness: as noted in the previous paper [1], some SAPs degrade quality more for some program items than for others, but the model has no input variable related to program type and so is incapable of making program-specific predictions. This insensitivity to program type contributes to the model's overall RMSE since its quality prediction for a particular SAP might be too high for one program item and too low for another.…”