Thursday, March 4, 2010

Kashu-do (歌手道): Squillo, a symbol for the mythology of vocal pedagogy

Few words are thrown about in vocal pedagogy as much as squillo. Like all the Italian catchwords that make up the large lexicon of terminology attributed to the Bel Canto Tradition, this word too is equivocal and only symbolizes concepts that are much more profound than its literal meaning.

Squillo: Suono acuto e brilliante. [A high-pitched and brilliant sound] (From the Italian Dictionary Online).

I have felt inspired to address this issue due to a very thought-provoking thread on NFCS (The New Forum for Classical Singers), mainly because the discussion reveals the level to which this word, and by extension vocal pedagogy as a whole, is unclear to us, the very people who practice it. Perhaps anyone profoundly invested in any discipline would say that the more one learns, the more one realizes how little s/he knows.  The difference in vocal pedagogy is that nothing can be taken as a given other than that ten pedagogues may have different opinions on a vocal issue, each passionately defending his/her view and rejecting the others, without any empirical basis. Even among the performing and fine arts, singing is singular in that regard.

When we consider Ballet or Figure Skating or even playing the cello or piano, enough research has been done on the biomechanics of skeletal musculature that a committed pedagogue could point to specific research that supports his/her approach. Indeed there are articles on carpal tunnel syndrome that speak to anatomy of the hand and its action relative to playing the piano or cello.  In singing, the intrinsic musculature of the larynx that drive vocal fold posturing and consequently vocal fold oscillation have not been studied, cannot be studied during the act of singing.  In fact this is the missing piece that scientists should work to find answers about.  

The great majority of vocal science is done relative to the study of speech and vocal disorders, in short observing and studying the vocal mechanism during average and sub-par function.  Operatic singing in the traditional sense requires a near-perfect balance between the intrinsic muscles that sets up the vocal folds into vibration in a specific morphology that yields a vibration pattern that addresses the most sensitive acoustic range of the human ear. In my opinion, this is specifically the acoustic patterns we see in the utterances of newborn babies before their expressions become educated. This specific acoustic pattern produces an intense sympathetic vibration in the aural cavities that the Old Italian masters of vocal pedagogy refer to as Squillo. The Italian masters realized that this “ringing” in the ear is produced by a voice that is operatically viable. It is this ringing that makes the voice more audible to the listener than other sounds in the immediate environment. This realization however does not tell us what makes the ringing or whether it is intrinsic to some voices and not others. This mystery is but one among many mysteries that drives the dark art of operatic singing. Dark, because many practitioners within our field are content to have it exist in the shadows of mythology rather than in the inexact world of partial knowledge that is its true nature!

The questions are the following:

  1. Do some voices produce squillo and others not? The answer seems obvious.  Yes! That part at least is obvious. No is the more complete answer.  The superficial answer is what we can observe. We can certifiably observe that certain voices produce the ringing in the voice that is synonymous with the kind of acoustic pattern that the human ear is most sensitive to, and that many other voices do not.  This could lead one to believe that squillo is a special occurrence in voices gifted by Providence to sing opera. Yet every crying baby produces the sound we experience as the ring in the voice. Babies crying can be heard over operatic divas and full orchestras. The logical answer is that all of us had squillo and some of us lost it as we grow up. Not every singer that has squillo in the voice has the necessary other components to become an opera singer (e.g. musical sensitivity, emotional expressivity and linguistic facility at least).  Operatic talent requires so many different skills. It is a complex talent. Those who are lazy about how to select operatic talent might take the easy way out by saying that those without squillo cannot develop it and therefore should not sing opera. That is a lie. Squillo is a prerequisite for singing opera viably and experience has shown me that anyone can develop it (or better said, recover it).

    Those with some acoustic science information will call it The Singer’s Formant. But that too is inadequate.  The Singer’s Formant is only the clustering of the upper three vowel formants and lies between 2500-3200 Hz. But this does not mean that the bandwidth of the singer’s formant is 700Hz.  It is much narrower than that. It is simply that the exact frequency of the SF varies between singers and indeed between voice types. In the high female voice (higher than D5), the SF falls too far between the harmonics to even have an effect on them. Yet the high soprano voice often exhibits squillo despite the fact that the SF does not apply.

  2. Is squillo the same as full glottal closure? Certainly not! But glottal closure is one component of the necessary fold posture that results in squillo. There must be enough fold-mass for the given tone to produce a tone rich in overtones.  Strength in the 2000-3000 Hz area is not enough for squillo.  A voice that has strength in this frequency range will be heard, but the kind of sound pressure that squillo provides includes strength in vowel formant areas as well. In other words, a voice may have strong high harmonics and still not have squillo. It can be heard but may sound tight and lacking in vibrancy.  The presence of strong lower harmonics in addition to the high ones guarantee that the glottal tone is produced in a way that allows for strong breath release during each glottal cycle. The high frequencies alone only show discouragement of the lower frequencies in favor of high ones. This can be accomplished with pressed voice. Therefore glottal closure alone is not enough to produce the kind of presence associated with a voice that is squillante.

  3. Can a voice be heard without squillo? Yes. It depends on the acoustic environment.  If the orchestra has a lot of doubling of the singer’s part, as is common in the music of Puccini and Wagner, then the singer needs a coordination that is close to what produces squillo in the voice. Even if the ring is not perceived, the sound pressure may be close enough for the singer to be heard, but that does not mean that any voice without squillo can be heard.  In fact my personal point of view is that the singer must become aware what a squillo feels like in the voices of other singers and should seek it in his/her voice. That should be the goal. And if the singer gets close most of the time, then the voice will have no problem being heard in the house. A well-produced opera voice should exhibit squillo throughout the voice most of the time, for that is what makes it possible to sing without a microphone when accompanied by challenging accompaniment that doubles the voice part.

  4. Is squillo tantamount to pressed voiced? Certainly not! Squillo cannot be produced in a pressed voice. But when squillo is equated with brightness or fold closure then certainly the perception of squillo could be that of pressing. Unfortunately this is the view of many teachers who promote loose phonation to avoid the pressing they associate with a voice with squillo. Those who have not been able to distinguish between the ringing in their ears and sheer loudness will fail to understand the subtleties.

In short squillo is the confirmation that the voice is at or near peak balance in terms of phonation, resonance and breath pressure. Quite often a singer has the correct laryngeal set up but the squillo is missing because the vocal tract shape (vowel) is incompatible with the glottal tone and suppresses certain harmonics that would contribute to squillo. Vowel modification is important as a refinement tool to achieve squillo. However, tracking formant resonances through vowel modification alone will not produce a squillo. The source tone must produce a rich spectrum of harmonics. This cannot happen when the fold mass is too little or when glottal closure is not efficient.

The tendency to oversimplify is what keeps vocal pedagogy in relative darkness and the complexity that lies below the simple Italian definition above is true of every Italian catch word that has come down as singer’s jargon. Voice teachers are a very opinionated bunch and to do the job well they have to be.  The problem lies only in stating an opinion as fact and shooting down anyone who disagrees for not understanding the principles of Bel Canto.  The term Bel Canto itself is thrown about as a shield against any suggestion to have a logical discussion about what we know and what we don’t know.

The voice teacher is in a very preacarious situation. S/he must make educated guesses because the vocal map is not complete. Scientists can only help us by giving us more information, but they are bound by the scientific process and thus in the absence of complete information they cannot make determinations. Singers cannot be so limited.  Despite incomplete information, we must deliver a product.  Yet the singer/voice teacher cannot fly totally blind.  If we use all the information the scientists provide, we can viably fill in the blanks.  Without using the information we do know we are likely to take the student on wild goose chase without any results to show for it.

The exercise that has kept applied vocal pedagogy in the dark is the following (often true in academia): two teachers with little empirical information will defend their diverging opinions with great passion, considering the other a quack. To keep the peace, they will agree in public that there is more than one way to skin a cat (i.e. that both approaches can work).  Often, neither approach is complete enough to yield consistent results.  It would have been prudent to consider a synthesis of the two approaches, which might indeed have produced a more complete regimen.  Unfortunately, on the surface, the two ideas seem opposite and neither teacher could conceive that an approach that seems so divergent from his/her own could actually work. 
When considering words like squillo and Bel Canto and appoggio, etc, it is important to realize that they are symbols for principles that touch on all three main functions (breath management, phonation and resonance).  The discussions that are sparked by these words provide much more complete information than any one opinion. A good teacher should always consider that what s/he dislikes most about a colleague’s approach might be the missing link that might make his/her approach more complete.

© 03/04/2010


Martin Berggren said...

I would like to add a few comments regarding the physics of the “singer’s formant”. As we all know, the amplification of harmonics around 3 kHz is a defining quality of operatic singing, particularly for the male voice. This amplification makes it possible for the voice to be heard through a loud orchestra. It is reasonable, I believe, to call this quality “squillo”, even though there are other aspects to this vague term, for instance for the high soprano voice.

Now, how does this amplification around 3 kHz occur? What is the mechanism behind it? Well, there seem to be a clear consensus among voice scientists about the mechanism, starting with Sundberg’s 1974 article and confirmed many times since then. This peak can be noticed, but only weakly, in the spoken voice. It is also typically weak in nonoperatic singing, but the peak is very strong in good-quality operatic singing. This peak is not a quality of certain vowels; it is a constant feature of the voice, independent of the vowel. In fact, the consensus is that the source of the peak is a resonance of what Sundberg calls the “larynx tube” and Titze the “epilarynx tube” (I will use the latter term). The epilarynx tube is a narrowing of the vocal tract just above the vocal folds; some dimensions mentioned in the literature are a length of about 2-3 cm and a cross-sectional area of about 0.5 cm2 (which corresponds to a diameter of about 8 mm). If the epilarynx tube ends in a wide enough pharynx (a ratio of at least 1:6 of the cross sectional areas are often mentioned), then this setup acts as a resonator with a resonance frequency of around 3 kHz. This resonance may interact with neighboring formants of the vocal tract and form a “cluster”, the singer’s formant. Note that both an epilarynx narrowing (“focus” perhaps?) and a wide pharynx is needed. The wide pharynx is obtained by a low-larynx position (assisted maybe by the “released jaw” that you advocate?).

There is another, quite different consequence of the epilarynx tube that has been promoted lately by Titze, namely that it provides a “back pressure” (or “inertial load”) on the vocal folds, helping them to vibrate more easily (similar to the function of the mouthpiece of a brass instrument, and similar to what happens with a mouth occlusion). Titze claims that the epilarynx tube generates a pulse skewing, that is, it generates harmonics H2, H3, ... This could be another aspect of “squillo”, different than the resonance phenomenon mentioned above. In fact, this effect happens also without the widening of the pharynx. Thus, a pop singer could benefit from the “back pressure” from the epilarynx tube to generate harmonics and get a bright sound, even if the singer’s formant is weak due to insufficient pharynx widening. Also, in order for this “back pressure” to have good effect on the phonation, the vibration mode, the setup of the vocal folds, have to be right. It may be that the concept of adequate fold depth that you promote is related to this.

Sorry for the long comment. Congratulations if you made it this far...

請吃飯 said...


KG said...

While I generally agree and like where you're going with this, I'm not a big fan of the "baby's cry" line of reasoning. Baby vocal tracts and palates are smaller than those of any adult male and also formed somewhat differently (to allow the baby to feed and breathe at the same time) and therefore more than being "educated" out of this mode of producing squillo I would say we literally grow out of it.
To me a voice that is truly squillante seems to exhibit three different types of resonance which I am as yet at a loss for how to describe empirically--I hope to figure it out by trial and error soon. In terms of sensation I would describe the three components as low (below the larynx and in the upper chest), back (kind of in the back of the head) and forward (behind and slightly above the nose.) I stress here that these are just personal sensations I associate with certain sounds I can hear in my own voice and those of my students, not necessarily what's biologically happening. I do feel that they all three play a part in producing squillo. I hope to be able to discuss specific frequency ranges and relative strengths with more investigation.
Klaus Georg

Jean-Ronald LaFond said...

I appreciate the commentary Klaus.

The undeveloped baby larynx is worth considering. In the comparison of the baby cry and Pavarotti's own acoustic signal on Bb5 the only difference was that the baby's cry was comfortably F1 dominant.

I teach a husband and wife who bring their newborn to lessons quite routinely. It fascinates me that the 6 month child, aping his heldentenor father and dramatic coloratura mother makes sounds that are audibly more efficient than both, and they sing very efficiently.

We will have to diverge on this one. I do indeed think that most of us unlearn the primal expression mode that creates squillo. I find it also remarkable that cultures that are more emotive in daily life (e.g. Italians, Koreans, Latin Americans, African Americans)commonly produce speakers and singers with operatic qualities even before training.

I am a fan of the vibrations in the torso and in the mask. I believe they are related to fold mass/closure variations. I find that the back of the head has a distorting effect on the vocal tract resonance.

Any findings of yours would be welcome.

lucabonvini said...

Hello Jean-Ronald thank you for.... just everything! Very interesting.

I had an intuition this week that all the terminology used here in Europe about "fare girare la voce", "tourner la voix" and similar 'symbolic' expressions may all be 'lost in translation' from the english 'Ring' in the voice, (which is the 'squillo'(?)).

This could have happened somewhere earlier with the italian, american and italo-american singers and translation in their teaching exchanges.
what do you think? There is any literary source that mention 'girare la voce' dating before the mixing of italian and american school of singing? These 'lost in translation' happened in the last century in many international musical teaching fields, for example brass playing.

I'm going to lesson this week and will try to sing focusing on producing the 'ring' (squillo?) (that I think I can hear and even visualize with software at home). I want to see if my teacher, that never speaks of formants but say that my voice should 'turn', 'tourner', 'girare' more will find that it 'gira' (rings?)

Lb (beginner operatic singer)

Jean-Ronald LaFond said...

Dear Luca,

Thank you for a very interesting comment. I believe you are correct. I have not researched the terms with specifically that objective in mind. However I have to say that my teachers in Italy did not use the term "girare" but I have heard the term a lot among Italians of the current generation. The American School became very well respected for technical achievement since the 70s and because many American teachers travel to Europe to teach, I imagine the back and forth has confused the language a lot.

I must say that turning the voice (i.e. girare, i.e. change from F1 to F2 dominance in the male voice) does not guarantee squillo. The formant adjustment as a refinement tool, can help achieve squillo if the fold posture is appropriate.

Welcome to the blog. Looking forward to more of your commentary.


Blue Yonder said...

Dear TS, I had a question about one of your posts on the NFCS squillo thread. What is the difference between "second formant strategy" and "singer's formant strategy"? What singers could I listen to, to help me understand the difference?

- said...

I too take issue with the comparison of squillo to a baby's cry. While the two phenomena resemble each other acoustically, extensive anatomical differences between the infant and adult larynx suggest that the respective modes of resonance that give rise to that sound are probably very different. To wit: Infants have a relatively larger tongue; a more anterior and _higher_ glottal opening (sic! the glottis is at spinal level C1 at birth, descends to C3-C4 by about age 7, and in adults sits at C4-C5); a larger, floppier, and more angled epiglottis (in fact, the newborn's epiglottis approximates with her posterior palate to help prevent aspiration during feeding; infants are obligate nose breathers for this reason). Finally, the infant larynx is narrowest at the cricoid cartilage, rather than at the glottis as in adults. In my mind, these distinctions essentially preclude any possibility of "relearning" an infantile mode of vocal production.

Indeed, if acculturation were the only factor in "loss" of squillo, we would expect that young children, with a little training, would be better able to produce a ringing sound, all other things being equal, than adults with similar experience. Is that what is seen by teachers of children's voices?


Jean-Ronald LaFond said...

Dear David,

The present of squillo does not signify an adult sound. The experience of squillo is simply a strong presence of the singer's formant region which coincides with the most sensitive acoustic area for humans and parenthetically with the acoustics of the inner ear.

Strength in this area is a given in baby's voices and we hear it in the voices of children at play. It is not something to be trained in children but rather one that is avoided in certain traditions of children's choral singing.

The yound Bjoerling, Carreras, Joselito (child star in Spain in 50s and 60s), the young Michael Jackson even and many children singers exhibit strong singer's formants.

The difference between the children and adults is less an issue of squillo but the one of greater fold mass and relative to opera a specific acoustic strategy.

Some years ago, a teacher in North Carolina brought her young students in the children's division at NATS auditions. Her 10 year olds consistently sang early Italian songs better than the college undergrads. While the young larynx cannot handle the kind of volume that an adult voice can, it does not mean that young voices have to be uncoordinated. Children cry and scream in joyful play often all day with no harm to their voices.

I do indeed believe that natural sounds are trained out of us and regaining a viable operatic sound depends on finding that coordination again. Good adduction is possible at any age. But developing dynamically with the changing voice is a major challenge particular for boys. The great Blanche Thebom was of the opinion that young girls could develop viable operatic voices. She was not wrong. In developing safety for young voices, the baby is thrown out with the bath-water.

Best to you,


Jean-Ronald LaFond said...

...More to the point, laryngeal anatomy has a great deal to do with the specific acoustic signals that are yielded. Although the changes in the size of the various pharyngeal tissues will change the acoustic nature of the individual voice, it should be obvious that squillo is not dependent on a specific laryngeal structure but rather upon the ability of the larynx to produce a strong source tone, which most children can do barring deformities and the ability of the phraryngeal structure to achieve the 1:6 ratio of aeryepiglottic fold circumference to pharyngeal circumference.

The single substantial acoustic difference still remains that the baby's laryngeal structure yields an F1 dominance strategy while Pavarotti exhibits an F2 strategy in the Bb4 in question. That can be explained through the nature of the primal vowel in the baby's cry. Tenors like Di Stefano, Kraus and Florez and most pop singers would yield signals similar to that of the baby.

Furthermore, that a leggiero tenor with a small laryngeal structure yields excellent squillo just as a dramatic Bass with a larger larynx is proof certainly that size of the laryngeal structure is not the reason for squillo.

Certainly the baby voice is not the same as the adult voice, but the process of sound making is no different. The nature of the phonation is what we should be paying attention with in a baby's voice, not so much the acoustics of the vocal tract because a bona fide singer's formant can be created with varying laryngeal adjustments as is amply discussed relative to F1 vs F2 strategies among the great tenors.

Still thank you for bringing this up, David.

I look forward to your further commentary as your comment suggests a strong scientific background.