The Influence of Male- and Female-Spoken Vowel Acoustics on Envelope-Following Responses

Abstract

The influence of male and female vowel characteristics on the envelope-following responses (EFRs) is not well understood. This study explored the role of vowel characteristics on the EFR at the fundamental frequency (f0) in response to the vowel /ε/ (as in “head”). Vowel tokens were spoken by five males and five females and EFRs were measured in 25 young adults (21 females). An auditory model was used to estimate changes in auditory processing that might account for talker effects on EFR amplitude. There were several differences between male and female vowels in relation to the EFR. For male talkers, EFR amplitudes were correlated with the bandwidth and harmonic count of the first formant, and the amplitude of the trough below the second formant. For female talkers, EFR amplitudes were correlated with the range of f0 frequencies and the amplitude of the trough above the second formant. The model suggested that the f0 EFR reflects a wide distribution of energy in speech, with primary contributions from high-frequency harmonics mediated from cochlear regions basal to the peaks of the first and second formants, not from low-frequency harmonics with energy near f0. Vowels produced by female talkers tend to produce lower-amplitude EFR, likely because they depend on higher-frequency harmonics where speech sound levels tend to be lower. This work advances auditory electrophysiology by showing how the EFR evoked by speech relates to the acoustics of speech, for both male and female voices.

Fundamental frequency predominantly drives talker differences in auditory brainstem responses to continuous speech

“…This study replicated the male/female talker effects on continuous speech ABRs (Figure 1b) but also demonstrated that f0 accounted for a large portion of these ABR differences (Figure 3a). While important to confirm, this result is not surprising, given the well-known rate effects for ABRs to brief stimuli (e.g., Burkard et al, 1990;Burkard and Hecox, 1983;Don et al, 1977;Jiang et al, 2009;Polonenko and Maddox, 2022), and the lower amplitudes for talkers with higher f0s in other auditory responses using different methods with continuous speech (Canneyt et al, 2021;Easwar et al, 2021Easwar et al, , 2022Maddox and Lee, 2018;Saiz-Alía et al, 2019;Saiz-Alia and Reichenbach, 2020). What this study adds is the systematic confirmation of f0 driving much of the ABR differences by keeping the same talkers while only adjusting the f0.…”

“…Only recently have researchers been able to use such naturally dynamic speech to investigate auditory neural processing through responses derived from electroencephalography (EEG) or magnetoencephalography. With several of these different new methods, however, the responses to female talkers seem to be smaller than to male talkers (Canneyt et al, 2021;Commuri et al, 2023;Easwar et al, 2021Easwar et al, , 2022Maddox and Lee, 2018;Polonenko and Maddox, 2021;Saiz-Alía et al, 2019;Saiz-Alia and Reichenbach, 2020), which can have implications/consequences for experiments requiring multiple talkers to more closely resemble real-life scenarios.…”

“…Therefore, it's reasonable to expect that differing f0, rather than gender itself, may be driving the talker differences in auditory responses. Recent studies do suggest that f0 contributes to the male/female differences in envelopefollowing responses, brainstem responses, and cortical temporal response functions (Canneyt et al, 2021;Easwar et al, 2021Easwar et al, , 2022Saiz-Alia and Reichenbach, 2020), but this effect was studied using talkers that also differ in characteristics besides f0.…”