Friday, March 5, 2010

Kashu-do (歌手道): Epilarynx and Pulse Skewing in reference to Squillo: Reply to Martin Berggren's comment

In the last post, I dealt with neither the Aryepiglottic Fold (epilarynx) nor inertial loading that causes the pulse skewing that Martin Berggren referred to in his excellent comment.

For simplification, the ratio of the diameter of the aryepiglottic fold (sometimes called the collar of the larynx) and the vocal tract must be approximately 1:6 in order for the Singer's Formant to provide enough of an energy boost in the 3kHz region.  The scientists agree that this is an independent frequency that is not based on the the vowel formants, although it may cluster with the upper vowel formants (cluster because they occupy the same acoustic space).  It also says that the 1:6 ratio occurs when the aryepiglottic fold narrows and the vocal tract widens by lowering the larynx.

This is one of the instances where I think the scientists make a crucial error.  The quality of a vowel and the frequency of its formants is strongly affected by laryngeal depth. So to say that the resonance of the epilarynx is totally independent of vowels is at best inaccurate.  Scientists must define their terms very narrowly. So I suppose that vowel alteration constitutes a perceivable difference on how the vowel is recognized rather than the alteration of the vocal tract. But in singing when we talk about vowel modification, we are indeed talking about minor alterations that may not affect vowel recognition at all but simply renders the sound production more efficient.

Efficiency also touches upon pulse skewing as related to supra-glottal inertia.  The way I look at all of this is that it is all interdependent.  The lower larynx that creates the conditions for inertial loading and the resonance of the epylarynx that yield the Singer's Formant cannot be separated from fold posture.  The proper mechanics of phonation lead to the kind of laryngeal depth that produces the 1:6 ratio that produces the SF. These mechanics also influence inertial loading.

The central issue I believe is whether the laryngeal musculature can maintain the configuration necessary for the 1:6 ratio.  I am certain that adequate fold depth sets up a balanced fold posture that has a direct influence on laryngeal depth. If the larynx is depressed without balanced fold posture, I believe the results will not yield the beneficial acoustic results we speak of here.

Vocal tract adjustments, including vowel modification will have an effect on the laryngeal oscillation and consequently on laryngeal depth.  From the pedagogical point of view, the question is "how do we achieve the 1:6 ratio?" I have been able to ascertain that vowel modification alone does not alter the formants significantly enough to alter resonance strategy unless fold posture is adequate relative to the specific voice.

© 03/05/2010


Martin Berggren said...

You give an excellent summary of the mechanism behind the formation of the singer’s formant in the second paragraph. The only thing to add, perhaps, is that the 1:6 ratio, as I understand it, is nothing magical. The widening of the pharynx just has to be large enough for the resonance to occur; 1:7, 1:8 and so on (if physically possible) would be equally fine. If the ratio is less then 1:6, the singer’s formant will be weaker.

I cannot, however, understand what you are trying to say in the third paragraph. For instance: The quality of a vowel and the frequency of its formants is strongly affected by laryngeal depth. . Sure, by lowering the larynx you get a longer instrument and slightly lower formants, in addition to the widening of the pharynx that sets up for the singer’s formant. So the general quality of the sound will be much altered towards an operatic sound. So to say that the resonance of the epilarynx is totally independent of vowels is at best inaccurate. I cannot follow you here. The resonance of the epilarynx tube is the main source of the singer’s formant. The epilarynx tube is short and deep down. Formants 1 and 2 that decide the main vowel characteristics is formed in the rest of the vocal tract above the epilarynx tube.

In the fourth paragraph you say: The lower larynx that creates the conditions for inertial loading .... I don’t think this is correct. The lower larynx creates the condition of a widening pharynx as needed to form the singer’s formant. The inertial loading and associated pulse skewing (as argued by Titze) does not depend on the 1:6 widening; it is caused by the narrow epilarynx tube itself. Titze give an example (Titze: Acoustic Interpretation of Resonant Voice, Journal of Voice 15:4, 2001) of a setup with an epilarynx tube and a narrow pharynx caused by a high larynx. This causes even higher inertial load, more pulse skewing, and more brightness (or “whiteness”) compared to the low-larynx case. Titze use the word “twang” for this voice character. Country singing?

Regarding fold posture: I feel that it is easier to keep a low larynx when using proper fold depth. In fact, your writing on this has helped me a lot. I also feel less stress on the vocal folds when the vibrations are coming sort-off from a deeper place. This may have to do with the inertial loading effect (I have said this several times before -:)) It turns out that for the inertial loading to have maximal positive effect, the vocal folds have to open up slightly upwards (slightly bell like). When the phonation is to thin, I believe that a shallow portion of the top part of the vocal folds are squeezed together, which causes the acoustic loading of the vocal tract, for instance inertia, to have little positive feedback effect on the vibration.

Jean-Ronald LaFond said...

Great comments Martin,

Relative to the vowel formants. One of Titze's own rules is that the lowered larynx lengthens the tube and has a profound effect on the bandwidth of the vowel formants. If the lowered larynx also causes the 1:6 ratio that influences the strength of the SF, then it would seem we cannot really talk about the issues being totally separate.

Sorry about the error on inertial loading. I did mean the narrowing of the epylarynx as opposed to the widening of the pharynx.

The larger point that I hope to make is that although we can point to localized events in the acoustics of the vocal tract, I believe the events have more global connotations.


Martin Berggren said...

Yes, I agree that the lowered larynx has several effects on the voice quality. The overall affect on the formant structure due to the lengthening of the instrument is the scuro in Chiaroscuro, I believe. The chiaro , for male voices at least, is very much dependent on the epilarynx tube resonance, whose peak frequency (around 3 kHz) and strength depends on the dimensions of the epilarynx tube and its sudden expansion in the widening pharynx, caused by the lowered larynx (here the “released jaw” also helps, I think!). But the epilarynx tube resonance depends very little on the shape of the rest of the vocal tract, so the shape modifications that affects the first and second formants do not much affect the singer’s formant. This is how I understand the science. In fact, I sort-of think that we agree on this -:).

One thing I have noted is that most practitioners (singers and teachers) seem uncomfortable with the concept of the epilarynx tube. The idea that an opera singer should narrow something in the throat may sounds shocking for a singer, but the effect has been well established for like 35 years. Such a narrowing may appear contradictory to the idea of an open throat, although I believe this narrowing makes phonation easier and promotes a sensation of flow through the vocal folds. One of the few writings I’ve seen that seems in line with the physics of epilarynx tube resonance is a forum response on Grand Tenori from Gioacchino. He speaks of the “three layers” of the voice. The first layer is a “released dark sound”, which may relate to a low larynx / widened pharynx, and the second layer is, to quote: Directly around this layer is a sound of brilliance – the squillo. It is like a bell like quality of the voice. This sound emerges as we feel the cords "bite the air" or resist the air. As a consequence we feel a narrowing of the first layer... there is a resistance to the air movement. The brightness is felt in the larynx and directly around it. So the combination of these feelings of vibration, along with the sensation of the air column narrowing, conveys the feeling of the squillo compressing the air, or leaning on it. To me, this sounds like a perfect description of the epilarynx tube resonance, and also the inertial loading!

KGJames said...

If the lowered larynx also causes the 1:6 ratio that influences the strength of the SF, then it would seem we cannot really talk about the issues being totally separate.

On the other hand, knowing now that interaction is non-linear may ease the tendency to assume these events are linked. For instance, phonating through straws or singing through a very small /u/ vowel affects the ratio.

I myself have become quite fond of addressing the narrow epilarynx head on in my studio. The benefits have been measurable!
~ Kimberly James

Jean-Ronald LaFond said...

Thank you for your comment Kimberly and welcome. Glad to have another science-minded teacher on board.

Have you tried occlusion? And how do you measure your results?

Warmest regards,


Lauren Lucille said...

How do you actually 'narrow' the epilarynx??? I understand why it works but HOW do you actually get that right??

Kashu-Do said...

Dear Lauren Lucille, I believe that certain vowels, [i], [e], in particular, have a narrowing effect on the epilarynx. But narrowing the epilarynx alone is not the point. It is narrowing it while maintaining the pharynx wide. This requires a mode of phonation that is at once strong enough to produce the required high partials and also flexible enough to allow good transglottal flow!

Try the following exercise:

d4 d5 c5 b4 a4 g4

The first note (d4) should have strong chest content; d5 on [hwi] will tend to cluster the upper formants. Combining the two sensations gives excellent and consistent results.