The presentation concerns the evaluation of a synthesizer of disordered voices. The objective is the perceptual assessment of the ability of the synthesizer to simulate disordered voice timbres. Three perceptual experiments, based on a pairwise comparison paradigm, have been carried out. The first involved jitter, the second breathiness and the third a combination of both. Results of the first two experiments show that the perceptual ranking accords with the synthesis parameters as well as measured speech jitter, speech shimmer and harmonics-to-noise ratios. For the third experiment, which involved jitter as well as additive noise, a two-dimensional multidimensional scaling analysis shows that for lower levels of additive noise, increased jitter and additive noise are perceived as distinct disordered voice timbres.

Index Terms. synthesis of disordered voice timbres, perceptual evaluation