Speech Prosody 2004
Single word stimuli from twelve female and twelve male natural speakers of various ages and from two synthetic voices were acoustically analyzed for duration and F0. Listening experiments were carried out to test if spectral features or F0 and duration provide the more dominant age cues and to test if listeners are equally good at estimating the age of female and male speakers. Results of the listening tests indicate that listeners are equally good at estimating the age of female and male voices and that spectral information is more important than F0 and duration in age perception of both male and female speakers. Strong correlations of duration with biological and perceived age were found for both female and male voices, indicating that duration is an important cue for agedness. When correlating F0 (mean, range, SD in Hertz and semitones) with biological and perceived age, the result was significant only for the female speakers (F0 range and F0SD in Hz), but no indication was found that listeners use different strategies when judging the age of female and male speakers. The acoustics and perception of speaker age will be studied further using a larger and more varied material.
Bibliographic reference. Schötz, Susanne (2004): "The role of f0 and duration in perception of female and male speaker age", In SP-2004, 379-382.