The effect of spatial separation on informational and energetic masking of speech in normal-hearing and hearing-impaired listeners

Listeners with sensorineural hearing loss often have inordinate difficulty understanding speech in a multi-source listening environment. The ability to perform well in such an environment may be related to the ability to perceptually separate the talkers (auditory objects). The interference produced by certain types of maskers may depend on the perceptual arrangement of auditory objects in space. The use of spatial separation by hearing-impaired listeners for improving speech recognition in the presence of an interfering sound was investigated. The task was closed-set speech recognition. Listeners were tested in two spatial conditions: 0° and 90° separation between signal and masker. Three maskers were tested separately: (1) the different-band sentence masker was designed to be primarily informational; (2) the different-band noise masker was a control for the different-band sentence; and (3) the same-band noise masker was designed to be primarily energetic. Performance was compared to that of normal-hearing listeners. In normal-hearing listeners, the benefit of spatial separation for the primarily informational masker was larger than for the control or the primarily energetic masker. The large advantage for the informational masker was attributed to the perceptual effects of spatial separation between signal and masker. The mean benefit of spatial separation in hearing-impaired listeners was the same as for normal-hearing listeners for the control and primarily energetic maskers, but was smaller for the primarily informational masker. The smaller benefit can be partially explained by the fact that stimuli were played at a reduced sensation level for some hearing-impaired listeners. However, the hearing-impaired listeners still obtained a larger spatial release from the primarily informational masker than from the primarily energetic masker, suggesting that these listeners can use the perceptual effect of spatial separation to improve speech recognition in the presence of an informational masker.