Sensitivity to spectrotemporal modulations accounts for a significant portion of the variance in speech intelligibility not explained by audibility measures such as the audiogram or the Speech Intelligibility Index. This suggests that the ability of the auditory system to accurately encode the patterns of modulations present in speech is crucial for successful speech recognition. Word recognition worsens in background noise conditions, but the negative effect of noise diminishes when speech is delayed by a few hundred milliseconds from the noise onset—a phenomenon known as adaptation to noise. This improvement in speech recognition in noise over time may result from enhanced encoding of temporal, spectral, and spectrotemporal modulations inherent to speech. This study investigated whether adaptation to noise occurs in the detection of temporal, spectral, and spectrotemporal modulations, as well as in vocoded and natural word recognition, in normal-hearing participants. Vocoded-word recognition and sensitivity to temporal and spectral modulations in noise improved when the word or modulated signal was delayed from the noise onset. In contrast, no such improvement was observed in natural-word recognition or in spectrotemporal modulation detection. These findings suggest that adaptation to noise in speech recognition is unlikely to be mediated by improved encoding of spectrotemporal modulations.
López-Ramos et al. (Wed,) studied this question.