Contribution of supra-threshold processing to speech masking release

Normal-hearing (NH) listeners can typically better understand speech in the presence of a fluctuating noise or a competing talker compared to a stationary noise interferer. However, for hearing-impaired (HI) listeners, this masking release (MR) is strongly reduced or completely absent. Traditionally, this has been attributed to the ability of NH listeners to utilize the speech in the low-amplitude periods of the masker, an ability that is supposed to be reduced for HI listeners due to reduced temporal and spectral resolution. However, [1] proposed that the reduced MR experienced by HI listeners is due to their higher speech reception threshold (SRT) in stationary noise. In the present study, this hypothesis was investigated by presenting noise-band vocoded as well as low-pass and high-pass filtered stimuli to the NH listeners. In this way, the SRTs of the NH listeners were similar to those of the HI listeners in stationary noise. It is shown that the relation between the MR and the SRT in stationary noise clearly depends on the type of processing used. The results therefore demonstrate that the SRT in stationary noise has only a minor effect on the amount of MR, which is in contrast to [1].