By Donald Miller
The death of a great singer always brings to mind a circumstance that I have found frustrating from the point of view of vocal pedagogy: most of the “knowledge” on which an extraordinary singing technique is based — and Pavarotti’s was indeed extraordinary — is contained in the body of the singer: its particular complex of cavities, muscles, and cultivated reflexes. These are no longer functioning, and that small part of the complex that the singer has talked about gives us only vague hints of how it all worked.
As anyone knows who has read Jerome Hines’s Great Singers on Great Singing, singers are often not very good at describing just what they do to achieve their beautiful sounds. When Hines asks him about his enviable execution of the passaggio, Pavarotti talks about how, over a long period of study, he learned to make the voice more “squeezed” on those particular notes. Like many other less-than-explicit terms used in voice instruction, this description may well be adequate — between teacher and pupil — to refer to an action one has already learned to perform, but for the outsider the term remains ambiguous.
Rather than puzzling over the exact nature of the squeezed voice and how the singer produces it, I would turn to the sounds themselves, as they are preserved in a large number of recordings, using spectrum analysis as the basic tool for investigation. In doing this, it is important to keep in mind that the expert human ear is both the most exquisite computer and the ultimate judge of the singing voice. However, when it comes to identifying the separate frequency components of a complex sound, the spectrum analyzer is far more reliable. And the reason why the spectrum analyzer is important in analyzing the phenomenon Pavarotti is that he was a consistent master in the use of resonance. Spectrum analysis gives us clear and objective evidence of use of resonance.
First a bit of background information may be helpful. It is common among male opera singers to refer to a place in the (second) passaggio where the voice “turns over.” Although this turning over is commonly perceived, what actually happens is not always understood. As the voice ascends the scale, the highest note in the chest register is characterized by a dominant second harmonic (H2) in the spectrum, amplified by the first formant (F1, the lowest resonance of the vocal tract). The voice “turns over” when the second harmonic moves beyond the strong influence of the first formant, giving up its dominant position. At this point the more skillful singers start tuning the second formant to the third or fourth harmonics, respectively, for back and front vowels. In successful execution of this maneuver the loss of first-formant resonance is more than compensated by the gain of the second formant. This is illustrated by a classic Pavarotti example in the chromatic (nearly a capella) ascent from F4 up to B4-flat in the Rigoletto quartet (see Fig. 1). At the pitch F4 , still in chest register, he tunes F1 to H2. One tone higher he gives up that resonance, and F2 starts picking up the third harmonic, staying with it all the way to the climactic B- flat, a major third higher. On the B- flat the third harmonic dominates by more than 12 decibels, constituting the bulk of the power of his voice. The well-tuned F2 produces this acoustic power without excessive effort.
Figure 1. Power spectra of Pavarotti’s F4 and B4-flat, from Rigoletto quartet. Dominant harmonics are marked with the cursor, and their frequencies are given in the lower right-hand corner. Source: YouTube: Pavarotti, Rigoletto Quartet. Click here for the YouTube video. If you have VoceVista, you can use it to monitor the YouTube audio by launching VoceVista and clicking Run. If the display does not respond to the sound, go to Project > Volume Control and try the various options to find the one that does respond. With appropriate sizing, both video and VoceVista displays can be watched simultaneously.
While there is no evidence that Pavarotti was aware of this acoustic analysis of his passaggio, the fact remains that his F2-tuning in the upper extension of the voice was exemplary in its consistency. Most successful operatic tenors achieve at least some F2-tuning on G4 and A4 -flat, but continuing that resonance strategy further, past A4 , is less common. Still more unusual is the ability to track a harmonic with the second formant as pitch changes. Pavarotti could, and did, carry it up to “high C” (C5), where F2 needs to be above 1600 Hz in this configuration.
Trying to teach someone to do this without being aware of the acoustic facts and without feedback from spectrum analysis is a challenging task, and it is interesting to observe Pavarotti’s attempts with a fine young tenor, available on YouTube as Pavarotti’s Masterclass 001. The duke’s aria from Rigoletto, “Questa o quella,” has a showy melisma on the vowel [e] at “se mi punge” with the sequence of notes A4 -flat, B4 -flat, A4 -flat, G4 . Pavarotti’s own recordings of the aria (see Fig. 2) clearly show F2 tracking the fourth harmonic (on the front vowel [e]) throughout the melisma, entailing moving F2 within a range of about 250 Hz. The A- flat and G fit relatively easily with this tuning, but the B- flat is a stretch, and tenors like to avoid that difficulty by changing to the back vowel [a] on the B- flat, thus dropping F2 to the third harmonic. Pavarotti is quite specific in his instructions, insisting that the pupil not move his mouth when he ascends to the B- flat. In this he is clearly teaching his own strategy of tracking H4 with F2. What he does not say, and what seems to me essential for moving F2 that high (ca. 1900 Hz), is that the tongue should be fronted, reducing the size of the front cavity and thus raising the second formant. (Interestingly, Pavarotti seems unaware of the tongue in the conversation reported in Hines, and Franco Corelli, another master of F2-tuning and presumably tongue-fronting, tells Hines when questioned, “The tongue for me is a thing I would not care to think about.”)
Figure 2. Pavarotti’s A- flat (above) and B- Flat on the vowel [e] from the aria “Questa o quella” from Rigoletto. He tracks the fourth harmonic with the second formant, moving it up 200 Hz for the higher note. Source: YouTube, Pavarotti, Questa o quella (the Ponnelle DVD)
The pupil does not follow the master’s advice and drops his jaw to produce the vowel [a] on the B- flat, which is unstable in the lesson (see Fig. 3).
Figure 3. The pupil’s A4 -flat (above), which resembles Pavarotti’s in putting F2 on H4, and B4 -flat, where he drops F2 to the third harmonic. Source: YouTube : Pavarotti Masterclass
For anyone interested in following this argument in detail, I have indicated the YouTube addresses where the sounds can be found, as well as added figures that show the pertinent spectra. These are conveniently gathered on the VoceVista website: www.vocevista.com If you have VoceVista, or any other program for spectrum analysis, I recommend that you look for yourself at the spectra.
Of course Pavarotti’s singing gives us a great deal more than second-format tuning. But I would argue that his precise attention to technique was an essential ingredient in his success, resulting in such consistent sound and freeing him from having to battle the orchestra to be heard. A good part of the excitement generated by his singing came from the way he let the audience feel the challenge of the high notes without interrupting the flow of emotion, reflected in his face, that came from the dramatic situation. And here he made video an important factor, helping to bring intimacy to the high drama of opera, which not so long ago was typically limited to long-range visuals or lip-synching close-ups.
In conclusion, a word of caution to those who would learn from his example. First, it should be understood that the perfection of F2-tuning for high notes is a male phenomenon. What sopranos do an octave higher was pointed out three decades ago by Johan Sundberg in the famous Scientific American article (1), which gave a large impulse to applying voice science to the singing voice. In the female high voice the tracking of harmonics with formants is also of great importance, but there it is the first formant that follows the fundamental, H1. This may sound similar, but an important difference is that F1-H1 tracking is a natural phenomenon and relatively easy to implement. Tuning F2 to a higher harmonic is by contrast elusive. There is no easy groove to slip the formant into, as there is in a hoot that glides upward. One of the things that distinguished Pavarotti from most other tenors is that he could nonetheless find such a resonant groove in his whole upper extension.
A second cautionary note is based on a phenomenon that is not so easily demonstrated with the audio signal alone. It has to do with the closed quotient, the percentage of each glottal cycle in which the vocal folds are closed, preventing the passage of air. It is best monitored with the electroglottograph (EGG), a non-invasive device that can detect the opening and closing of the vocal folds. As far as I know, no one has seen an EGG waveform of Pavarotti’s high notes, but my own investigations lead to the speculation that an unusually high closed quotient is required to produce the very dominant H4 that one sees in Pavarotti’s B4 -flat on the vowel [e]. And such a closed quotient is seldom found in the same voice with the lightness of production that Pavarotti could also employ.
As we pay homage to a truly exceptional singer, let us try to emulate him where possible — for example, in his legato and the F2-tuning on the big (male) high notes — but let us also be wise enough to distinguish between those places where we may learn to follow his example and those where it is better simply to be thankful that at least someone was gifted and motivated enough to take the voice there.