Spectral correlates of carrying power in speech and western lyrical singing according to acoustic and phonetic factors
Résumé
In order to define the variability of carrying power (sometimes called “vocal effectiveness”) indexes in speech and singing, an acoustic analysis of vowels, sentences, singing exercises, and lyrical piece spoken and sung by 23 singers, was conducted. Two parameters were measured: (i) the difference in amplitude between the highest harmonic between 2 and 4 kHz and the one between 0 and 2 kHz ("Singing Power Ratio", Omori et al., 1996); and (ii) the difference between the singing formant (near 3000Hz) and a spectral minimum that follows it. Results showed that the bringing together of the third and fourth formants between spoken and sung vowels corresponded to the singing-formant, followed by the spectral minimum more significant in singing. The singing-formant was correlated with high values of Singing Power Ratio (-16dB) and high values of the difference in amplitude between the singing formant and the spectral minimum (32 dB). Only this last parameter differentiated untrained from trained subjects. High values of these two indexes were associated with great intensities, spoken [i], and singing in comparison with speech. The same values were obtained whatever the language. These two parameters varied depending on the task (sustained vowel versus singing exercises) and on vocal category.
Origine : Accord explicite pour ce dépôt