Understanding effects of hearing loss on multitalker speech intelligibility in terms of glimpsing

Abstract

In multitalker mixtures, listeners with hearing loss demonstrate reduced spatial release from masking compared to listeners with normal hearing. However, it is not clear whether this problem reflects an inability to use spatial cues to segregate sounds, or a degraded representation of the target speech itself. In this work, a simple monaural glimpsing model was used to isolate the target information that is potentially available at each ear in spatialized speech mixtures, and intelligibility of these glimpsed stimuli was then measured directly. Performance in the glimpsed condition was strongly correlated with performance in the natural spatial condition, suggesting a common limit in both cases. Similar results were found for different kinds of speech mixtures in which the target and maskers were distinguished by cues other than spatial location (talker sex, or time-reversal). The results suggest that the primary detrimental effect of hearing loss might be on the representation of target glimpses, rather ...