Speech-Reception Threshold in a Fluctuating Background Sound and its Possible Relation to Temporal Auditory Resolution

Abstract

The speech-reception threshold (SRT) in noise is a frequently encountered limitation in everyday aural communication, especially for hearing-impaired listeners. The SRT for sentences can be described with a simple signal-to-noise ratio model (Plomp, 1978) containing two parameters related to hearing loss: ‘A’ for attenuation and ‘D’ for distortion. Hearing loss for speech in noise is represented by the parameter D and hearing loss for speech in quiet by the sum of A and D, both in decibels. In normal-hearing listeners the SRT for short meaningful sentences (50% correct) in noise is reached for about -5 dB signal-to-noise ratio. At threshold, hearing-impaired listeners need a better signal-to-noise ratio, up to 10 dB, depending on the kind of hearing loss (D < 10 dB). Although hearing loss for speech in noise may seem very moderate in terms of decibels, it should be noted that in critical conditions 1 dB increase in signal-to-noise ratio gives an 18% higher intelligibility score for sentences (Duquesnoy, 1983). As a result, in such conditions, speech that is intelligible to normal-hearing listeners may be completely unintelligible to hearing-impaired listeners.