A fast measure of spatial separation for release from masking and its relation to intelligibility.

Abstract

This study validates a fast measure for spatial release from masking-minimum angular separation (MAS), the smallest spatial separation between a target and two-talker maskers to improve speech intelligibility by 20%. Three psychophysical methods to estimate MAS were compared, including the constant stimuli, adaptive staircase, and progressive tracking, which revealed no significant difference in the estimated threshold on the group level with bootstrapping. Results suggest that the MAS measurement can be expedited using the progressive tracking method without compromising robustness in the threshold estimation. The non-linear relationship between target-masker spatial separation, signal-to-noise ratio, and accuracy is explored.