Effects of reverberation and binaural sensitivity on spatial release from masking

The impacts of reverberation and binaural sensitivity on spatial release from masking (SRM) were assessed in an adaptive speech-in-speech recognition task with open-set sentence targets presented in continuous narrative speech maskers. Forty young adults with normal hearing (ages 19-55 years old) completed the speech task. Under anechoic conditions, target speech and two masker streams were presented from a single frontal loudspeaker (co-located) or three loudspeakers separated by 45° (separated). In simulated reverberant conditions, reflections were calculated via a modified source image model (10 m × 10 m; 0.4 s reverberation time) and presented throughout 360° of azimuth. Interaural time difference (ITD) and interaural level difference (ILD) discrimination thresholds were separately measured over headphones using a speech-based two-alternative forced choice task. Reverberation elevated speech reception thresholds (SRTs) and reduced SRM by 5.4 dB on average. Individual ITD thresholds were significantly associated with separated SRT but not with SRM-likely as a result of intercorrelation between co-located and separated scores. The overall results corroborate prior reports of roughly 5 dB SRM reduction in reverberation, and also provide evidence for individual ITD sensitivity as a predictor of spatial speech perception in noise, which may be obscured in analyses that focus only on difference scores such as SRM.