Subjective and Acoustical Data about Vocal Effectiveness in Western Opera: Preliminary Results

The notion of "vocal efficiency" is difficult to define. In order to contribute to this definition, several data were obtained: first the evaluation of glottal efficiency Eg using an acoustic analysis was obtained. Second, we run a study asking subjects to find synonyms of vocal effectiveness. Results showed the prevalence of the perceptual aspect over the communication and production aspects. This concept depends on the vocal use by persons. The subjects insist on the notion of carrying power without fatigue whatever the context. Some acoustic measurements of singing power Ratio, an acoustic correlate of this perceptual aspect, were conducted in opera pieces with and without pianistic accompaniement: the singing-formant, a well known perceptive correlate of vocal effectiveness, was always present whatever the condition. So it appears that the concept of vocal effectiveness has a polysemous definition regarding subject vocal perception (auditor), production (singer) and communication (the public). INTRODUCTION The term of vocal effectiveness is often taken as a synonym to glottal efficiency, which is measured traditionnally in purely physical terms [11, 12]. From the engineering viewpoint, it may be defined as the ratio of vocal intensity to the glottal power, i.e.. subglottal air pressure times glottal airflow. However, particularly in artistic performance, vocal effectiveness may also include a subjective dimension. Vocal efficiency is a complex matter: it was here studied especially from both a cognitive point of view and a subjective one. In this communication, emphasis was put on vocal effectiveness (optimization of effort) rather than vocal efficiency (energy conversion and output). After a short evaluation of glottal efficiency of our subjects in the first part of this communication, we evaluated the subjective definition of vocal effectiveness of a greater population through a questionnaire. Finally, in order to show the perceptive aspect of vocal effectiveness, weÕve conducted acoustic analyses. 1. EVALUATION OF GLOTTAL EFFICIENCY USING ACOUSTIC ANALYSIS 1.1. Method 17 professional opera singers (11 singers of the Bastille Opera House in Paris), were tested in this study , including 3 sopranos, 4 mezzo-sopranos and altos, 6 tenors and 4 basses. Ages ranged from 30 to 51 years (M=39±6,3). Number of years of vocal experience and training ranged from 6 to 35 years (M=16,7±8,2). There was 70% of French singers and three subjects were American. All received different kind of musical training. According to Shoji et al. [8], the "H-Index" represents the first proposed acoustical measure of glottal efficiency. "Hi" ([haj]) was chosen to evaluate glottal efficiency since it provides an excellent test of glottal transformation from voiceless ([h]) to voiced sound energy ([aj]). Indeed, Shoji found a high degree of correlation with the AC/DC ratio, which represents the transformation by the glottis of direct current airflow (DC) into alternative current airflow (AC). AC/DC ratio is a standard aerodynamic measure of glottal efficiency. Our subjects were recorded in a soundproof room with a Digital Audio Tape, a cardio•d microphone placed at 40 cm from the mouth. They were asked to prononce five spoken and five sung "hi" at whatever pitches. The power ratio of [h] to [aj] was calculated from power peak of each measure from the power envelope. 1.2. Results H-Index is higher in singing than in speech in our population; so, according to Shoji [8], singing quality is more efficient than speaking quality for our subjects (t17=-3,74; p=0,001; fig.1). The mean of our subjects in spoken [haj] is comparable with Shoji's results and is greater than in pathologic voice [8]. However, our standard deviations are greater than Shoji's one, probably because of the native language of our population: this measure depends on the way each one prononced the sequence "hai" in his native language. For our American subjects, the standard deviation is similar to Shoji's results; so supplementary measures should be done in speech and singing to compare our results with Shoji's ones. 1.3. Discussion 1.3.1. Glottal efficiency (Eg)values for singers. According to Schutte [6], if one considers energy, trained singers are no more efficient in their use of their larynxes than untrained persons. Moreover, Eg values between untrained and trained male and female subjects showed non significative differences in Sulter and Wit [10]. 1.3.2. Problems with Eg definitions and calculations. Difficulty of measuring subglottal air pressure makes this vocal efficiency impractical as a routine vocal function test [9]. Moreover, it's calculation doesn't include a cost-benefit ratio for vocal health and long term performance, an important parameter obtained in our second part. And in all studies, the efficiency is characterized by large standard deviations, indicating large variations in individual values. Beside that, it does not mean that the greater the efficiency, the better the glottal function in a physiologic sense: hyperfunctional voice such as ventricular phonation shows a very high value [9]. Finally, the efficiency of a trained singing voice is not reflected in the values of the subglottic pressure and airflow rate, but it will have its origin in the proper use of vocal tract tuning. [7] So vocal effectiveness in opera singing should not be approached only by acoustic measures such as the H-Index or aerodynamic data to determine the glottal efficiency. Other factors should be taken into consideration. page 2445 ICPhS99 San Francisco Context of emission H -I nd ex ( dB ) 0 5 10 15 20 25 30 35 40 45 50