Spatial release from informational masking

Abstract

A new method for investigating spatial release from informational masking was developed and employed in two experiments. The new method is computer controlled and efficient. It employs the versatile coordinate response measure speech stimulus set [Bolia et al., J. Acoust. Soc. Am. 107, 1065 (2000)]. The experiments were conducted in an anechoic room, with a primary loudspeaker in front of the listener and a secondary loudspeaker at 60 deg to the right. Target messages were presented from the primary speaker only. For a standard, distractor messages, simultaneous with the target, were also presented from the primary speaker only. Spatial release was measured by presenting the distractors from both primary and secondary speakers with a temporal offset. Experiment 1 fixed the offset (secondary leading, +4 ms) and varied the number of distractors (1 to 3) and the target-to-distractor ratio (−12 to +4 dB). Masking release, sometimes as large as 10 dB, was found for all combinations of these variables. Experiment 2 varied the offset over a wide range of values. Substantial release from masking was found for both positive and negative offsets, but only in the range in which speech echoes are suppressed (<50 ms). [Work supported by NIDCD grant DC 00181.]