7th International Workshop on Models and Analysis of Vocal Emissions for Biomedical Applications (MAVEBA 2011)

Florence, Italy
August 25-27, 2011

Spectral Analysis of Pathological Voices: Sustained Vowels Vs Running Speech

Rubén Fraile (1), Juan Ignacio Godino-Llorente (2), Nicolás Sáenz-Lechón (2), Juana M. Gutiérrez-Arriola (2), Víctor Osma-Ruiz (2)

(1) Universidad CEU Cardenal Herrera, Moncada, Valencia, Spain
(2) Department of Circuits & Systems Engineering, Universidad Politécnica de Madrid, Madrid, Spain

A short-time spectral analysis of normophonic and dysphonic voices is presented. This analysis has been performed on recordings of both sustained vowels and running speech for comparison purposes. The reported results indicate that pathological voices tend to have a higher concentration of energy in the lowest frequencies (20 to 300 Hz) and less energy in frequencies from 630 to 1,270 Hz. Additionally, pathological voices tend to experience quicker variations in spectral energy from 770 to 1,720 Hz.

Index Terms. Speech analysis, spectral analysis, correlation

Full Paper (reprinted with permission from Firenze University Press)

Bibliographic reference.  Fraile, Rubén / Godino-Llorente, Juan Ignacio / Sáenz-Lechón, Nicolás / Gutiérrez-Arriola, Juana M. / Osma-Ruiz, Víctor (2011): "Spectral analysis of pathological voices: sustained vowels vs running speech", In MAVEBA-2011, 67-70.