On the color of voices: Does good perception of vocal differences relate to better speech intelligibility in cocktail-party settings?

Abstract

Understanding speech in cocktail-party settings poses a challenge for cochlear implant (CI) users. Distortions of vocal cues as they are perceived through the implant have been hypothesized to partially contribute to this difficulty. Vocal cues can be defined along two orthogonal dimensions: the fundamental frequency, F0, related to the pitch, and the vocal tract length (VTL), correlated with speaker size. While deficiency in F0 perception in CI users has been long known, recent research has shown a large deficit in VTL perception. How this contributes to speech intelligibility in the presence of a competing talker (speech-on-speech; SOS) remains unknown. In this study, we investigated the relationship between SOS intelligibility and the sensitivity to F0 and VTL differences (JNDs) in 18 CI listeners using two separate speech tasks. Results indicate a strong correlation between voice cue sensitivity and SOS intelligibility: participants who are more sensitive to small differences in F0 and VTL tend to perform better on the two SOS tasks compared to those who are less sensitive. These results are consistent with the hypothesis that degraded voice cue sensitivity contributes to SOS perception deficits, but other potential explanations, like differences in cognitive processing and attention, cannot yet be ruled out. [Funding: The University Medical Center Groningen (UMCG), Advanced Bionics, Netherlands Organization for Scientific Research (NWO).]