Vocal efficiency and carrying power in speech and singing using aerodynamic measurements and Magnetic Resonance Imaging

Programme de la manifestation et résumés des exposés disponibles sur: http://lpp.univ-paris3.fr/productions/conferences/2008/SpeechProductionWorkshop_2008/programme_workshop3.htmlThe aim of this poster is to present recent progresses of instrumentation techniques together with new interesting outcomes, especially in speech and western lyrical singing production. Here we studied two main characteristics of classical singing: vocal efficiency and carrying power ("vocal effectiveness") Vocal efficiency (VE) is the ratio of intensity (I) to subglottal pressure (Ps) times glottal airflow (q). In order to calculate VE in speech and singing, data were obtained for one male French subject: 1° recordings of tracheal and intra-oral Ps, q, I and fundamental frequency; 2° Calculation of VE on spoken and sung syllables, sentences, and glissandi. Results showed that 1° half of spoken values are similar to the published one; most of the sung values are smaller; 2° the intra-oral Ps (pio) is not consistent: so tracheal pressure are precious and rare data to study speech or singing production. In order to study vocal carrying power in speech and singing, Magnetic Resonance Imaging (MRI) slices of French vowels [a], [i] and [o] emitted by a professional bass singer were performed. They showed a great number of articulatory modifications from speech to singing, which interacted between each other. The transfer function estimated from physiological data (Maeda's vocal tract acoustic simulation program) showed several acoustic modifications. The most significative was the bringing together of F3 and F4. This phenomenon corresponds to the well known singing formant. But this model didn't take into account the role of piriform sinuses; according to the study of Dang and Honda (1997), the piriform fossa contributes spectral minima to speech spectra in a region of 4 to 5 kHz