ISCA Archive VOQUAL 2003
ISCA Archive VOQUAL 2003

Exemplar-based voice quality analysis and control using a high quality auditory morphing procedure based on STRAIGHT

Hideki Kawahara

This paper tries to introduce a new strategy and tools for voice quality research that complements conventional approaches. A very high-quality speech analysis, modification and synthesis procedure STRAIGHT, which is basically a channel VOCODER based on a pitch-synchronous analysis synthesis framework, was extended to implement auditory morphing in terms of spectral, pitch and voice quality parameters. This extension enables voice quality modification by parametric transformation using STRAIGHT. It also enables an exemplar-based research strategy for perceptual aspects of voice quality analysis and control. In other words, manipulated synthetic voice having virtually equivalent naturalness to natural voice introduces a mean to perform a unique research strategy called systematic downgrading, that is suitable especially for para- and nonlinguistic aspects of human vocalization. In addition to morphing procedure, a set of visualization techniques were introduced based on fixed-point analyses in the time and the frequency domain for assisting exploratory data analysis that is indispensable in voice quality research.


Cite as: Kawahara, H. (2003) Exemplar-based voice quality analysis and control using a high quality auditory morphing procedure based on STRAIGHT. Proc. Voice Quality: Functions, Analysis and Synthesis (VOQUAL 2003), 109-114

@inproceedings{kawahara03_voqual,
  author={Hideki Kawahara},
  title={{Exemplar-based voice quality analysis and control using a high quality auditory morphing procedure based on STRAIGHT}},
  year=2003,
  booktitle={Proc. Voice Quality: Functions, Analysis and Synthesis (VOQUAL 2003)},
  pages={109--114}
}