Second International Workshop on Models and Analysis of Vocal Emissions for Biomedical Applications (MAVEBA 2001)

Florence, Italy
September 13-15, 2001

Towards a Systematic Quantification of the Speech Speed

Rubén Wainschenker, Jorge Doorn, Marcela Castro

INTIA - Facultad de Ciencias Exactas, Universidad Nacional del Centro de la Provincia de Buenos Aires, Paraje Arroyo Seco, Campus Universitario, Tandil, Argentina

Although it is widely used in every day activities and in several professions such as locution, shorthand and stenography among others, the magnitude speed of the speech does not have a precise way of measurement. Such notion is fundamental in the context of the study of the typical behavior of a main set of language allophones as the basic components for a robust speech synthesis. No precise information was found for Spanish and less for the variant spoken in Uruguay and center and south of Argentina usually called Riverplatean Spanish.
   This article presents a quantitative characterization of the intuitive notions of fast, slow and normal speech speed. The article reports a research with a strong experimental foundation since its conclusions were obtained from 120 texts uttered by different speakers at different speeds in a context free of any sort of conditioning. More than a half of the texts were collected from public sources and from people who never knew they were recorded for this purpose.

Index Terms. speech speed, allophones, voice synthesis, voice parameters

Full Paper

Bibliographic reference.  Wainschenker, Rubén / Doorn, Jorge / Castro, Marcela (2001): "Towards a systematic quantification of the speech speed", In MAVEBA-2001, 108-112.