Listening to degraded speech is associated with decreased intelligibility and increased effort. However, listeners are generally able to adapt to certain types of degradations. While intelligibility of degraded speech is modulated by talker acoustics, it is unclear whether talker acoustics also affect effort and adaptation. Moreover, it has been demonstrated that talker differences are preserved across spectral degradations, but it is not known whether this effect extends to temporal degradations and which acoustic-phonetic characteristics are responsible. In a listening experiment combined with pupillometry, participants were presented with speech in quiet as well as in masking noise, timecompressed, and noise-vocoded speech by 16 Southern British English speakers. Results showed that intelligibility, but not adaptation, was modulated by talker acoustics. Talkers who were more intelligible under noise-vocoding were also more intelligible under masking and time-compression. This effect was linked to acoustic-phonetic profiles with greater vowel space dispersion (VSD) and energy in mid-range frequencies, as well as slower speaking rate. While pupil dilation indicated increasing effort with decreasing intelligibility, this study also linked reduced effort in quiet to talkers with greater VSD. The results emphasize the relevance of talker acoustics for intelligibility and effort in degraded listening conditions. V

Individual differences in talker acoustics substantially affect intelligibility in adverse listening conditions. Spectral enhancement has been found to reliably boost intelligibility in noise while temporal enhancement remains less effective. A potentially mediating factor that has been ignored so far is listening effort, as objectively assessed by the pupil dilation response. In two perception experiments, we measured intelligibility (keyword recall scores) and listening effort (pupil dilation) for two talkers in two listening conditions and with varying degrees of temporal modification. Results suggest that while keyword recall scores are sensitive to individual talker differences across listening conditions, the pupil dilation response reflects the degree of temporal and spectral distortion introduced by the signal processing techniques.