Above The Glottis: Vocal Tract
Above The Glottis: Vocal Tract
Above The Glottis: Vocal Tract
First of all I want to thank Professor Schutte not only for the invitation. I want to thank him
and the other scientists who really helped me to improve and correct (!) my voice teaching
with their deep insights. Opera singers made me what I did not want to be at first: a voice
teacher. I had some success in teaching exercises to reduce the vocal load for these stressed
professionals (helping to remove recent oedema). This means that I worked very much with
Today with the help of the voice science we can create different exercises of our own using
the vowel chart, choosing initial consonants to get a lower first formant for middle (mixed,
full head), a vocal tract nearly closed at the mouth, gradual reduction of the vibrating mass
of the vocal folds, correct coup de la glotte exercises after Garcia (1) connecting the vocal
folds, diphthongs which are suited to cross the vowel chart in search of harmonics.
In over 20 years of practice as an ENT-doctor and voice teacher I have seen damage of the
vocal folds primarily from too much breath pressure (hyperfunctional abuse of the voice)
due to lack of information on the part of the singer. This in turn causes imbalance of the
forces which make a voice acoustically strong, imbalance of the vibrato and pathology of
the vocal folds (e.g. oedema, nodules, varicosis). This is why a good voice teacher does not
Three possibilities which make a voice acoustically strong must be in balance (2):
2. At the glottis
More adductional contact
Vocal folds,
Higher closed quotient
Fig. 1 Balance
-3-
To be honest I have also taught things that I now regard as errors. The main reason for this
is the fact that some of the most important acoustic laws important for singing seem to be
counterintuitive. Let me address three of them, which often lead to dangerous clichés.
One cannot direct or project sound or place the voice forward (especially not with the help
of breath support). Air can be directed. After the passage of the larynx however, airflow is
transformed into acoustic energy and aerodynamic laws are no longer applicable. In the
vocal tract the acoustic laws are valid (omnidirectional propagation of sound)! The feeling
This means: placement and projection are by-products of correct tone production rather than
a means to it.
constricted regions because of the higher impedance of the air column(3). This is why a
Fig 2. Two tubes of inequal diameter: in an area of contraction the transmitted acoustic
pressure Pt is larger than in the expansion, indicated by the larger arrow. Titze 1994.
Vocal tract
Glottis Mouth
dog
Dogs can only bark (chest) or howl (falsetto).Pigs and men in contrast are able to create
powerful high tones in middle register depending on the dimension of their epilarynx tube.
The narrow epilarynx tube * provides feedback energy to the glottis through high inertance
(dashed lines). Nevertheless there is a weak point at 600-700 Hz in this computer model
(arrow in Fig.4 left side) indicating a register transition at one half of this frequency (pitch
frequency times two because high chest is a F1/H2 –first formant/second harmonic-
configuration)(4).
larynx * larynx
Mouth closed
With that occlusion at the mouth (Fig.4), we have an advantageous situation with a positive
inertance all the way through from 300 Hz (E4b) up to 1300 (E6) Hertz where most singers
have to earn their money (5). This is why exercises with that small mouth as lip trills, closed
vowels, hummings are so well suited to study the mix(ed) or middle and the shift to the
dominant second formant. For the female highs (700-1000 Hz) there is a better
solution with the F1/H1 configuration. We need that appropriate degree of inertia or
sluggishness of response (narrowing of some part of the vocal tract) in order to avoid an
middle, mix(ed) or full head register. The reader might object that tenors usually sing high
tones with a rather open mouth. Of course the mouth tends to be more open than in the
passaggio for the same vowel. But it is usually more rounded at the mouth than in high
chest.
DVD-example: at the end of “Ah si, ben mio” from Verdi’s “Il Trovatore” Pavarotti shows
an even more rounded and closed vowel at the B4b “in ciel” in mixed (middle) than in the
Chest/Abdomen and Larynx are unequal partners regarding muscular strength. Large
muscles of the body can furnish more than 200 watts. Our glottis can use only about 2 watts
(3).
With high abdominal pressures singers risk damage of the vocal folds!
Those large muscles should initially “learn” to deliver a steady flow. But they cannot even
achieve this, when the connection of the vocal folds is not perfect. The glottal valve has to
adjust the pressure with the help of the energy fed back from the vocal tract.
Hence the large abdominal and chest muscles are not suited to do the fine adjustment of the
subglottal pressure and the glottal adduction, because this 100 fold lever is too big. It‘s like
setting our watch with a pipe wrench. We rather destroy the watch. It is the other way
around: the glottal valve must have the leadership, and the chest and abdomen must learn to
adapt to it. 4 kPa of pressure (4 kilo Pascal or 40 cm of water) would be the weight of an
apple over the surface of the glottis and may be already enough for a forte tone.
One of my former voice teachers used to repeat that saying of Enrico Caruso: the high C of
a tenor affords the same power as a weight lifting of 100 kg. It is our duty as voice teachers
With today’s live DVD recordings we are all able to observe the great singers of our time.
None of them seems to “support”, but the abdomen seems to follow the needs of the glottis!!
Sundberg (6): is it the absence of such instances of increases of glottal resistance that
The often quoted saying: he who knows how to breathe well, knows how to sing well, is
less than a half-truth. It is totally misleading. (Compared with an organ: he who knows how
to tread the bellows, knows how to play the organ). Here too: effect is mistaken for the
cause. Breath support or correct breathing is a by-product of correct tone production rather
When the closure of the valve (vocal folds) is not sufficient, even the strongest external
intercostal muscles (arrows) cannot prevent running out of air. The glottis is the controlling
apparatus and the most vulnerable part at the same time. Singing never causes oedema,
Victor Maurel (Verdis first Jago and Falstaff) says: “All is won when we regard the vocal
folds directly as the breath regulators and relieve them of all overwork through the
Formant Tuning. Formants are the maximum resonances of the vocal tract. Harmonics in
the vicinity of these formants are boosted. The two lowest formants “form” (are!) the vowel.
When the fundamental is high (high chest, middle and upper register), the harmonics may
be so widely spaced, that not all the formants can be energized. The fundamental (pitch) and
all the other harmonics are prescribed by the composer, so the adjustment must take place in
the vowels. With a subtle vowel modification a formant can be tuned to one of these
harmonics and boost the vocal intensity dramatically(3). The singer might gain 30 dB in
extreme cases(6). With a doubling (!) of the lung pressure the gain is only 6 dB.
-8-
Barney and Peterson diagram In the mouth and pharynx it looks like that
2. Formant
i 2. Formant
u
a i
1. Formant
u a
1. Formant frequency
Fig.7. The vowels as a function of the two lowest formants (vocal tract resonances ; )corner vowels emphasized.
F2
F1
The corner vowels are not suited for the middle register, because they have a tendency to a
There is another advantage of this acoustical approach. The singer keeps a relatively low
larynx without the fixed conscious control of the outer neck muscles, but stays flexibly
consistent with an optimal formant tuning, because the position of the larynx is important
for the wavelength and the frequency of the vowel formants! Sundberg (6): vowels sung
a. lyric: F2/H2 . b. robust: F2/H3 for back vowels, higher CQ than in a.,
Female mixed (middle): more variations (middle range, less tension than in male middle)
Sound Qualities.
F 2 / H 2: lyric mixed
1. Audio example “un aura amorosa“ Luciano Pavarotti F2/H3 (lyric spinto), Leopold
2. Classical belt: end of “Goldfinger” sung by Shirley Bassey F1/H2 on D5b, “he loves
g[a]ld” instead of “g[ou]ld”, because for this frequency only F1 of [a] and [ae] is available
Fig11. Only F1 of open vowels at 1108 Hz (554 Hz x 2 because of the H2- tuning).
3. VoceVista: young female pop singer a capella, singing “somewhere over the rainbow”:
H2
where
H3 find
d
Fig.12. Female pop singer high chest (belt)F1/H2 on the word “where” above and
In the power spectrum (right part in VoceVista above) we see the line spectrum of the
harmonics(8). Scientists might argue that the continuous spectrum of the formants cannot be
seen as such, but mainly in its influence on the harmonics (higher amplitude = boost). But
voice teachers and singers do not need to measure formants. They want to train their ears
with the help of a visual feed back! They want to coordinate their auditory impression with
the vocal tract (F) /voice source (H) configurations mentioned above. They want a diagnosis
of the registers and they want to find suitable vowels or diphthongs in order to create their
own exercises, when problems emerge. Notice also the loss of vibrato in “where” in (figure
12). The first passaggio (bridge) to the female middle and male upper range needs a vowel
modification customized exactly for every singer. Thus figure 13 can only show the
principle of narrowing or centralizing of the vowels like in the following words: we to sit,
let to German Lehm as in the Irish faith, cat to let, father to mother,
Singers have different vocal tracts and vocal folds. So diphthongs may be the better choice
attention to any emerging boost without (!) increasing breath pressure of course (because
many singers mistake the boost created by increased lung pressure for an improved focus).
Especially the English language has many useful diphthongs for this purpose.
-12-
The first harmonic (arrow) is dominating 25 dB (!) above the second harmonic. The dotted
vertical lines are at 1000 Hz intervals, the dotted horizontal lines are at intervals of 10 dB.
1985 Miller and Schutte wrote a very important paper on the effect of F1/H1 tuning in
soprano high notes (4). At that time it was difficult to explain how the high female
singing voice, with its relatively short closed quotient, can sustain long phrases of high
intensity, instead of running out of breath prematurely. Schutte and Miller measured the
pressure variations just above and below the glottis and found that the airflow actually
reversed during a short part of the glottal cycle. It is the experience of many female singers
of high notes that a properly tuned vowel along with a correct muscular adjustment in the
vocal folds offer an increased resistance to breath pressure contributing to a feeling of ease.
The sound reflections of the powerful F1/H1 tuning from the vocal tract may be so strong
that they even override the subglottal pressure and the airstream from the lungs.
This unexpected finding again shows the overestimation of breath support in voice teaching.
Singing means to make music with an invisible instrument for the most part. But even
beginners can recognize the relatively large respiratory movements of chest and abdomen.
So the overestimation of breath support at the cost of glottis and vocal tract is a specific
mistake of the beginner (singer and teacher). This even involves danger.
-13-
Conclusion. Breath support, placement and vibrato are by- products of correct tone
But what can be seen as pre-condition for this correct tone production? Respiration: correct
posture for a regular breath supply (legato). Larynx: a correct connection of the vocal folds
in respect of pitch, intensity and register. Vocal tract: a correct vowel adjustment regarding
Today our singers need an up-to-date analysis of their problem, which leads to a diagnosis