Figure 1.
Spectrograms and pitch tracks (dotted yellow lines) of the base sentences in Experiment 1 (a–c) and Experiments 2–5 (d–f). In order, the three rows of graphs represent utterances in normal, breathy, and pressed voices.
Table 1.
Parameters and their changes applied to the base sentences for the preparation of the stimuli.
Figure 2.
Judgments of voice attractiveness (a–c) and emotion (d–e), on a scale of 1–5, as a function of Voice quality, Pitch shift, Formant shift, Final F0 slope and Pitch range. Each row of the graphs (a–e) corresponds to Experiments 1–5 respectively. In each bar, the black figures represent mean rating score, while parameter values are in white. The error bars are standard errors.
Figure 3.
Band energy profiles of speech sounds. Each profile consists of fifteen signal energy values computed from overlapping spectral bands of 500-Hz bandwidth: 0–500, 250–750, 500–1000, … 3250–3750, 3500–4000. a, Mean band energy profiles of all 6 vowels in the three base sentences of Experiment 1, each with an intended voice quality. b, Band energy profiles of two sample files from Bruckert et al. (2010). c, Profiles of three synthetic sentences used in Experiment 2–5, each with a synthetic voice quality.
Table 2.
Measurements of voice quality.