The Low Mandible Maneuver and Its Resonential

Implications for Elite Singers

*,,Angelika Nair, *,Garyth Nair, and Gernot Reishofer, *yzNew Jersey, and xGraz, Austria

Summary: Many elite singers appear to frequently drop the posterior mandible while singing to optimize resonance
production. This study investigated the physiology of the Low Mandible Maneuver (LMM) through the use of magnetic
resonance imaging (MRI), ultrasound (US), and spectrographic analysis.
The study of elite singers has been hampered by the paucity of internal imagery. We have attempted to address this
problem by using portable US equipment that we could transport to the homes, studios, or dressing rooms of such
ranking singers. With the US and acoustic data gathered in fairly brief sessions, we were able to ascertain the resonance
gains garnered from the techniques use.
The study featured two phases: IMRI study of the maneuver and its physiological effect on surrounding structures
(in collaboration of the Medical University of Graz, Austria) and IIUS investigation that studied tongue shape during
the maneuver.
The LMM has significant ramifications for resonance production by enabling a concomitantly lowered larynx and
increased resonance space in the pharyngeal and oral cavities. Measurements of the LMM ranged between 0.7 and
3.1 cm and showed a boost in the first harmonics as well as an enhancement in the singers formant. Its use also has
a rather significant effect on the tongue shapes required for all sung phonemes. The advantage of using US for this study
was the ability to produce real-time videos of the singer in action and then, through the use of stop action, precisely study
both individual phoneme production and phoneme-to-phoneme transitions during the LMM.
Key Words: UltrasoundResonance productionTongue shapesBiofeedback.

INTRODUCTION for a remarkable unity in the techniques of elite singers, no mat-

Internal imagery of classical singers ter where or when they were trained.
The biggest challenge of voice research is the acquisition of in- Although there have been multiple studies on vocal reso-
ternal imagery without using an invasive method. This research nance17 in speech and singing, we suspect that there is a
study is designed to increase our knowledge of human vocal resonance-creation strategy practiced by elite singers that has
resonancespecifically our knowledge concerning the vocal received little or no study. Specifically, we are investigating a
production of those singers who inhabit the highest realms of resonance-enhancing effect produced by a use of the mandible
classical singingthe elite singers. But, it is also based on a phi- that is far from speech normsthe Low Mandible Maneuver
losophy and passion, shared by the authors of this articleto do (LMM; Figure 1). Sundberg8 reported that the singers
voice research with direct application into the voice studio. formant (Fs) is imperative for the singer to be heard above an
The authors would like to take a moment to define elite orchestra. Recent studies showed that the enhancement in the
singers. These singers are members of that rarefied confrater- range of 2.03.5 kHz was related to voice quality and resonant
nity occupied by those who sing in the major opera houses voice, respectively.9,10 With the LMM, we expect an overall
and concert halls on a truly international basis, for example, enhancement of the first harmonics and the Fs especially.
van Dam, Pavarotti, Kaufmann, Hvorostovsky, Sutherland, Because elite singers are difficult to get into a laboratory or
Horne, Baker, Netrebko, Fleming, Garan
ca, just to name a hospital for internal imagery, this maneuver has not been
few. Furthermore, our definition refers to those singers who explored. However, there is visual and aural evidence available
have sustained at least 10 years at the top rank. This second, du- for the existence of the LMM.
rational test of a singers rank automatically assumes that these The authors have spent countless hours studying videos of
singers possess a technique sufficient to produce the power and elite singers while paying close attention to the visual and aural
tone required by the Western classical operatic style while re- evidence they afford. As a result, we have compiled a listing of
maining vocally healthy throughout the duration of their ca- such external clues, including
reers. Although studies that involve low-rank professional
(1) far higher percentage of mouth opening during sung pas-
singers have produced valuable information, their techniques
sages than singers of lesser rank11;
are sometimes less convincing. From videos, we see evidence
(2) visible evidence of a lowered posterior mandible (the
angle of the ramus);
Accepted for publication March 18, 2015. (3) where visible, the laryngeal prominence (the protruding
From the *Pro Voce, Chatham, New Jersey; yDrew University, Music Department, Mad-
ison, New Jersey; zCollege of Saint Elizabeth, Music Department, Morristown, New Jer-
Adams apple) reveals a consistently low larynx; and
sey; and the xClinic for Radiology, MR Physics, Medical University of Graz, Graz, (4) less visible evidence of superficial tension in the lower
Austria.
Address correspondence and reprint requests to Angelika Nair, 420 River Road Apt. C-2,
neck region.
Chatham, NJ 07928. E-mail: [email protected]
Journal of Voice, Vol. 30, No. 1, pp. 128.e13-128.e32 In addition, both authors of this study have extensive profes-
0892-1997/$36.00
2016 The Voice Foundation
sional singing experience and have been trained in the use of the
http://dx.doi.org/10.1016/j.jvoice.2015.03.010 LMM technique.
Angelika Nair, et al The LMM and Its Resonential Implications for Elite Singers 128.e14

FIGURE 1. MRI images from phase I of our LMM study which illustrate the Low Mandible Maneuverleft image, speaking /i/ and the right
image, singing /i/ with full classical resonance. MRI courtesy of the Medical University of Graz, Austria.

Finally, in interviews with other singers and in looking for Problems with various image-collection modes. X-ray
clues in published interviews involving elite singers, the and CAT Scan Radiation Exposure. The most commonly used
LMM principle is mentioned although it does not have an offi- technologies for obtaining internal images involve the use of
cial name, as yet. radiation: x-ray and CAT scans. If we were to use x-ray with
a goal of assembling a catalog of images for all phonemes,
the subjects would have to endure radiation exposure far in
Current state of research in classical singing excess of allowable norms (morally, ethically, and legally).
Although there has been abundant scientific research on the With CAT scan, although techniques are available to drasti-
acoustic (aural) ramifications of classical singer resonance cally reduce the radiation exposure produced during x-ray
production, the anatomical and physiological elements use, the resulting images do not match the clarity of MRI.
have not been particularly well explored. To fully understand
the physical aspects of the art, we need top-grade internal imag- MRIThree Problems. Although MRI provides us with the
ery (magnetic resonance imaging [MRI], computerized axial sharpest view of the anatomical structure of the singers vocal
tomography [CAT], and ultrasound [US] images) demon- mechanisms, there are three drawbacks to this mode of imagery.
strating how a singers pharyngeal and oral spaces can be Static Imaging Often Requires the Subject To Freeze the
manipulated to produce the extraordinary sounds we enjoy in Vocal Mechanism. Standard MRI avoids the radiation expo-
the concert hall and the opera house. sure but has its own significant drawback because it often re-
However, we only rarely find such high-level internal imag- quires the subject to hold still for the entire time it takes the
ery in the scientific literature. This rarity becomes more pro- machine to capture an image. Because of this freezing of
nounced as one moves up through the professional ranks of anatomical structures during the utterance of a phoneme, the
classical singers. By the time we arrive at elite singers, we use of MRI can be highly susceptible to unwanted artifacts.
find absolutely no imagery on which to base our understanding Moving MRI (at up to 60 fps) is now possible and would
of how they produce their amazing sounds. Thus, we are left be an ideal method for studying the questions presented by
with an incomplete picture. this proposal. However, at the time of this study, such facil-
This research study is designed to acquire imagery from ities are rare and extremely costly.
those singers who have received the least scientific attention Cost. Normally, the cost of obtaining the number of images
because of their relative inaccessibility: elite singers. It is hoped needed for this study using standard MRI would be prohibitive.
that the collection of such imagery might add to our scientific However, thanks to the generosity of our colleagues at the Med-
knowledge of the physioacoustics of resonance creation of ical University of Graz (Austria), this drawback has been over-
the highest level and ultimately lead to the creation of a come during phase I of the study.
resource that may enlighten pedagogic approaches for all
singers, not just those of the highest rank. Gravity-Induced Artifacts. Singing is executed in an up-
right position but most MRI is done in a supine position.
This fact might induce a degree of tissue distortion because
Reasons for the dearth of internal imagery of gravitational effects. Although a recent study on tenors
Most of the voice research is being done on students or low-rank showed that a supine position has a rather small effect on
professionals. There are two principal reasons for the dearth of the vocal tract12 and that the tongue was more posterior on
internal imagery of classical singers vocal apparatus in action: the supine than upright position, it had no significant
128.e15 Journal of Voice, Vol. 30, No. 1, 2016

FIGURE 3. The spectrum of Western singing styles with those

closest to the norms of speech on the left and the nonspeech-like reso-
nance norms on the right. Also, to avoid a possible confusion of terms,
please note the difference between rehabilitated versus rehabituated.

Power. The classical singing style requires power sufficient to

fill large venues (routinely seating more than 1000 listeners)
without the aid of electrical amplification. Additionally, these
singers must often sing over the accompaniment of full
orchestra.
Tone. In addition to this requirement for vocal power, the
classical performing tradition demands a tonal richness that is
simply not found in the norms of everyday speech. To satisfy
this additional demand, singers must maximize the resonance
FIGURE 2. Phase I imagery from Grazoverlay of an ultrasound (internal space) available to their voices during every note,
image (light gray) tongue shape (/i/) and an MRI done in the prone po- loud or soft, high or low, all while clearly enunciating a variety
sition to account for the potential gravitational artifacts in the MRI im- of languages.
agery. Image from our collaboration with the Medical University of An interesting clue regarding just how far this style sits from
Graz, Austria. the norms of the singers own speech is the fact that it is a uni-
versal dictum that the most difficult language for anyone to sing
phoneme effect.13 In phase I of this study, by performing the well is his/her own native language. This is because it is so diffi-
same protocol using both MRI (subject supine) and US cult to break the fetters of ones own speech habits, a corpus of
(subject upright), we can correlate the two image modes to neuromuscular instructions that have been habituated through
ascertain the degree of gravity-induced artifacts in the MRI countless hours of talking during the singers lifetime.
imagery (Figure 2).
Range. Finally, the classical singer is expected to thrill audi-
Elite singers professional schedules. As mentioned ences with that power/tone combination from the very lowest to
above, previously published imagery has been obtained mostly the very highest notes that the singers anatomy and physiology
by using singers who have sufficient free time to both travel to can provide. The pitch range in which he or she is expected to
the laboratory and hospital as well as devote the time for the sing is extreme by most other stylistic standards.
acquisition of the images once on site. This virtually compels Combined, these demands can seem almost superhuman, and
the use of students and singers of local or regional rank. classical singers must accomplish this day after day, night after
Elite singers must endure hectic continent-spanning sched- night, without overstressing or injuring their vocal folds or
ules that rarely permit extended image sessions in either labo- attendant musculature.
ratories or hospitals. They also do not have the desire or The classical singer must significantly increase his/her reso-
interest because they know already how it works and what nance volume to achieve the tone and power needed for the style
they have to do. As a result, this rarefied population is not in which he or she performs. Aside from the genetic endowment
well studied. of the singer, many of the trainable technical advances in the
Had elite singers been available for internal imagery collec- singers technique arise from enhancing the volume of the reso-
tion, the LMM resonance-creation strategy at the heart of this nance areas of the voice in the oral cavity and the pharynx.
study may have been well documented by now. It is time to rem- This study investigates how the vocal resonance space is be-
edy that situation. ing manipulated in size and shape.

Elite singers use highly refined resonance creation

techniques Vocal resonance space
The higher one goes in the ranks of classical singers, the more When our vocal folds vibrate during speech and song, they pro-
you will find a refined use of resonance creation techniques that duce a very soft, dull sound (some have described it as a faint
are far more consistent, universal (in a phonemic sense), and buzz on pitch). The weak nature of this initial sound makes
highly habituated than the average classical singer. it difficult to identify it as recognizably humancertainly not
The classical singer inhabits a phonetic world far removed what we would call a voice.
from speech norms (Figure 3). Because of the demands of the It is only when that tiny, buzz-like laryngeal signal travels up
style, a massive retraining of the neuromuscular phonetic in- through the pharynx and oral cavity that a recognizable human
structions is required and involves mastery in three key areas: voice ensues. During its travel through the vocal tract
Angelika Nair, et al The LMM and Its Resonential Implications for Elite Singers 128.e16

FIGURE 4. The resonance spaces in the supraglottal vocal instru-

ment. (Note that the velopharyngeal port is shown closed because
most phonemes require the nasal space to be removed from the reso-
nance system.) Illustration by Garyth Nair.

FIGURE 5. An MRI overlay of the TMJ of the singer at rest (red)

(Figure 4), the laryngeal signal is greatly amplified and en-
and performing the LMM (green). The LMM condyle position of a
riched. All qualitative acoustic values that enable our enjoy-
sung [] is indicated by the arrow. (For interpretation of the references
ment of any voicespeaking or singingderive from this
to color in this figure legend, the reader is referred to the Web version
effect of resonance. The body structures that enclose our reso-
of this article.) Image from our collaboration with the Medical Univer-
nance airspace are remarkably plastic. Our ability to manipulate
sity of Graz, Austria.
these structures enables us to vary our vocal sounds to produce
all language, all tone quality and, in fact, even the emotional tion of poorly produced speech consonants. In consequence, the
content of our linguistic message. singer loses his or her beautiful sound.

Not all classical tone is created equal Definition

When one goes through the hallway of a conservatory, one When we are talking about LMM, we are referring to a maneu-
hears a singer warming up in a voice studiothe voice sounds ver on the basis of the first 10% of the yawn sequence in
glorious, rich, and resonant. However, if one walks past that which the entire mandible platform drops. It also creates signif-
same studio later when the singer is working on a song, most icant additions to
often one cannot believe that it is the same singer! Where did
his/her wonderful voice go? (1) oral space
(2) pharyngeal space (hyoid/larynx drops with the mandible
lengthening the pharynx)
HYPOTHESIS
Low mandible maneuver
We believe that the mandible use and its maintenance during
phonation and throughout a singers registers, respectively,
may be a principal factor in the tonal success of elite singers.
Because most warm-ups are vocalises involving only vowels,
the singers vocal production is unencumbered by the demands
of consonants. This frees the singer to concentrate on LMM
production of the warm-up vowels and, as a result, they sound
wonderful. However, while singing a song that must contain all
phonemes, the consonants are being added. Because the conso-
nants are produced at a speech speed, the brain turns to speech
instructions in the background processing. But, classical FIGURE 6. Diagram showing the anatomic components of the TMJ.
vowel resonance space cannot be maintained with the interjec- Illustration by Angelika Nair.
128.e17 Journal of Voice, Vol. 30, No. 1, 2016

FIGURE 7. Two jaw-opening paradigms. The two skulls at the top show the mandible at rest in its normal configuration (l) and the mouth opened
(r). The bottom skulls show the LMM in operation with the condyle dropped (l) and the resulting increase in mouth opening and oral resonance that
results from the maneuver. Illustration by Garyth Nair.

Figure 5 shows two MRI images that have been overlaid. The When one speaks, the condyle rotates mostly in the
red line shows the temporomandibular joint (TMJ) of the singer mandibular fossa (hinge motion) with a relatively stable jaw
at rest, whereas the green line shows the singer performing the and minimal vertical tongue motion (see the top two skull illus-
LMM. The condyle position both at rest and with LMM can be trations in Figure 6). For loud speech, there occurs a minor trans-
observed by the yellow line (image from our collaboration with lation along the articular eminence (also, articular tubercle).
the Medical University of Graz, Austria). From this relatively high mandible position, there is not much
resonance space available in the oral cavity.
LMM physiology
The TMJ is one of the most complex joints in the body,
providing a hinging and gliding movement at the same time. SingingLMM adds an additional maneuver
Figure 6 shows how the articular disc separates the condyle LMM, however, consists of two interconnected, simultaneous
from the mandibular fossa. Between the condyle and the disc actions (see lower two skulls in Figure 7). Once the condyle rea-
is the inferior joint cavity. This is also where the first part of ches the end of the first part of the mouth opening (25 mm), it
mouth opening occurs via pure rotation of the condyle. now starts to translate (a movement of a body in the same direc-
tion and at the same rate) over the superior joint cavity (space
Speech between the disc and mandibular fossa) out of the fossa into
The action of the jaw during speech is envisioned as a lever the articular eminence (5060 mm).14
where the fulcrum point is at its connection to the skull, the TMJ. In summary the two actions are:
Angelika Nair, et al The LMM and Its Resonential Implications for Elite Singers 128.e18

it is no longer anchored in the TMJ. Even without any addi-

tional mouth opening, this maneuver yields more oral space
(Figure 8).

Thus, the first 10% of a yawn creates:

(1) Oral cavity resonance space potentialenabling optimal

mouth opening
(2) Pharyngeal resonance space potentialwe are
measuring between 1 and 2.5 cm of laryngeal drop, which
elongates the pharynx and may have a factor in Sundbergs
6:1 ratio for singers formant.1
(3) A possible wideningwe suspectof the top of the
oropharynx because of the forward mandible movement dur-
FIGURE 8. Close up of the downward and forward drift of the ing the LMM movements.
condyle to the bottom of the articular eminence.
Mandibular function also includes, but is not limited to, mus-
(1) A drop of the entire mandible, not just the anterior cles of mastication. During mastication, three muscles of masti-
portion as is experienced in most speech norms. Imagery cation (or musculi masticatorii) are responsible for adduction of
from phase I of our study clearly shows the TMJ changes the jaw, and two (the superior and inferior lateral pterygoid)
necessary for LMM (Figure 5the LMM condyle attitude help to abduct it. All five move the jaw laterally. Other muscles,
is denoted with the arrow). This condyle motion is similar such as the sternocleidomastoid and the posterior cervical mus-
to that which occurs during the first phase of a yawn. In peda- cles, play a major role in stabilizing the skull and allowing a
gogy, there is some debate concerning the use of the yawn as controlled movement of the mandible (Figure 9).
a means of inducing more resonance in a classical singers Previous studies investigated the effects of articulatory
technique. One benefit of our study will be commentary configuration in singers resonance. In 1997, Sundberg and
with backup imageryto support the use of the beginning Skoog15 measured jaw opening by means of mangnetometer
of the yawn maneuver as a pedagogical tool and refute those equipments and found that the jaw opening seems to be an
who inveigh against it. important, though not the only, tool for singers to avoid F0
(2) From that lower LMM posterior mandible position, the exceed F1. With the application of dynamic real-time MRI,
jaw is then used mostly as it is in speech, as a lever although recent studies tried to continue the examination of vocal tract

FIGURE 9. Head and neck muscles anatomy. Illustration by Angelika Nair.

128.e19 Journal of Voice, Vol. 30, No. 1, 2016

modifications in register transitions in male singer6 and the specifies the description of the sensation for the proper open
vocal tract in female registers.7 Among others, both studies throat and laryngeal position in the middle voice as the first
showed substantial changes in the jaw opening once F0 reached stage of a yawn, and that an expanded throat sensation must
the vicinity of F1. Thus, it appears that the use of jaw opening always be accentuated as the pitch is raised. Vennard19 ad-
was unanimously found to be applied once F0 reached the vicin- dresses the necessity of The Loose Jaw for better results in
ity of F1. Hence, preparing a register equalization in male and even more anatomical detail. A theory on the use of LMM,
female passaggi. Story et al16 investigated various voice quali- open mouth, relaxed ramus of the mandible, the lower larynx,
ties, such as yawny and found an increase in vocal tract vol- throughout all registers and the concomitant rehabituation of
ume through lengthening the vocal tract and widening the oral vowels and consonants, respectively was proposed in articles
cavity, confirming the yawn as a manipulation to achieve from Denver,11 Salzburg20 and the book, The Craft of Singing21
greater space and lower the larynx. by Garyth Nair. Last but not least, all the empirical visual/
The biomechanics of the TMJ is a finely tuned dynamic bal- audible evidence of its use by elite singers, whether live or on
ance among all the head and neck muscles. As a person yawns, video, exhibits a remarkable technical unanimity.
the head is brought back by contraction of the posterior cervical
muscles, which raises the maxillary teeth. This simple example
not only demonstrates that singing is solely orientated on the METHODS
first part of a yawn (avoiding a hyperextension of the skull) Our study is divided in 2 phases:
but also demonstrates that even normal functioning of the
masticatory system uses many more muscles than just those IPilot study with regionally ranked singers in which we
of mastication. With an understanding of this relationship, used both MRI and US, refined imaging methods and proto-
one can see that any effect on the function of the muscles of cols, and studied biomechanics of high-level resonance
mastication also has an effect on other head and neck muscles production.
(and vice versa) and ultimately the larynx. Further studies on IIInternal imagery of elite singers to investigate biome-
the complex orchestration of extrinsic neck and head muscles chanics of high-level resonance production. This phase of
and their influence on the larynx, tongue, and so forth may the study is still a work in progress and will attempt to ascer-
reveal more of the biomechanics and ultimately the techniques tain the prevalence of LMM in the technique of elite singers.
used in classical singing.
The potential benefits of the use of LMM are that it:

(1) produces the classical singers richest, most harmonic- Ad phase I

laden tone (clear enhancement of all upper partials and the Phase I of the pilot study was conducted in Graz, Austria, in
Fs, respectively). January 2013 and was constituted of five participants, all of
(2) indicates the singers rank because the tone and diction whom are professional singers with considerable performance
produced by LMM results in the consistency of all pho- experience as opera and concert soloists (Table 1).
nemes, not just vowels; and The same protocol was used for both modes of imagery:
(3) may be a possible factor in career longevity (impression-
istic and experiential assumption that would need to be (1) Subject at rest
analyzed with studies of singer careers; longitudinal (2) Swallow to locate the echo of the palate (primarily for
follow-ups or retrospective analyses would have to be US)
conducted). (3) [][i] toggle three times
 speaking
The evidence of LMM can be found in various pedagogical  singing (on a single pitch)
literatures. Garcia17 uses descriptions such as the natural fall (4) Italian phrase
of the jaw and jaw should be dropping loosely, respectively  speaking
as a general preparation for emitting the voice. Appelman18  singing (on a single pitch)

TABLE 1.
Subjects Demographics
Number of Number as Number of
Subject No. Voice Age Years Study Professional/y Performances/y
1 Mezzo-sopr. 37 8 15 <20
2 Basso cant. 69 8 40 <10
3 Baritone 53 8 25 <40
4 Baritone 61 8 40 <50
5* Soprano 25 8 7 <50
* Unfortunately this participant got sick and could not participate in the ultrasound protocol.
Angelika Nair, et al The LMM and Its Resonential Implications for Elite Singers 128.e20

(5) Native language phrase (Native language was included

to see if there are differences induced by the partici-
pants speech template. However, because the goal of
this article was to first show the LMM in singing in gen-
eral, this aspect will form part of another article in which
the main focus will be the consonants in singing.)
 speaking
 singing (on a single pitch)

The following two modes of imagery have been used:

Ultrasound
The device used for this mode of imagery was a GE Logiq 7
FIGURE 11. EdgeTrak software capturing the edge of the sagittal
with 30 frame rates per second (misses most fast consonant
tongue profile (red line) as well as the hard palate (yellow line). (For
tongue actions), a GE 3S transducer, the HATS-ProVoce (stabi-
interpretation of the references to color in this figure legend, the reader
lizes distance from the transducer to palate), and the Standoff
is referred to the Web version of this article.)
(to allow mandible platform drop). Before the session, the sub-
jects were asked to choose a phrase in Italian and their Native
language. For the sung part of the vowels and phrases, the
Head and transducer stabilization. For utilization as a
singers were asked to choose a comfortable pitch.
research tool, the relativity between the US transducer and
Principals. An US relies on piezoelectric crystalsfound in the subjects head must be stabilized while still allowing for
the transducerthat emanate ultrahigh-frequency sound waves vertical mandible movement above the US transducer through
and produce an image by using their reflective properties. Both the use of a device called a Head and transducer stabilization
the tongue surface and the hard palate can be imaged in real (HATS). In the lead-up to the image collection in phase I, we
time using US22 (Figure 10). were able to test and refine our HATS system for use during
EdgeTrak. The software program EdgeTrak23 uses a sophisti- US data collection (Figure 13).
cated algorithm that searches through the noise (reflections of In-laboratory HATS systems are generally quite large, heavy
the sound waves) of the US and determines where the edges and very nonportable. However, because this project calls for
of the sagittal tongue profile are located (Figure 11). collecting US imagery on-location and not in the lab, we proto-
typed a HATS-ProVoce unit that had to be:
Mapping and fixing the position of the hard palate.
Before beginning the US data collection protocol, the subject
 light weight;
is asked to drink a bolus of water. On the US image, the US
 unitizeda system that can be broken down for transpor-
signal reflected by the thin layer of water between the top of
tation to remote locationssingers homes, studios, dres-
the tongue and the hard palate reveals its precise location.
sing rooms, and so forth; and
Once that stable hard palate position has been obtained, it can
 stablemaintains the spatial relationship between the
be superimposed on all subsequent images for that subject un-
subjects head and the transducer while allowing
der study and allows accurate measurement of the jaw elevation
mandible movement.
during the production of all phonemes (Figure 12).

FIGURE 10. Sagittal view of a US image. Illustration by Angelika FIGURE 12. US image showing the water bolus starting from the
Nair. alveolar ridge to obtain a stable position of the hard palate.
128.e21 Journal of Voice, Vol. 30, No. 1, 2016

FIGURE 13. The HATS rig built out of a professional microphone

FIGURE 14. Detail of the HATS in action showing the transducer
boom stand (vertical) and two Manfrotto universal photo arms.
and the intercessory fluid reservoir that permits mandible movement
during image acquisition. Also, note the reference points at the fore-
Vertical stability. Vertical stability is accomplished with the head and under the nose that stabilizes the skull.
use of a robust professional microphone boom stand as the
frame to which all other equipment is attached. To add addi-
Standoff to permit jaw movement. After the subjects
tional vertical stability as well as an aid in maintaining spatial
head is stabilized in the HATS, the transducer arm is rotated
reference between the subject and the system, a special Man-
into position below the chin (Figure 14). Because we are
frotto heavy-duty clamp solidly connects the vertical stand to
investigating the drop of the posterior mandible (LMM), we
a horizontal platform that is fixed to the seat bottom of the chair
supporting the subject. With this support in place, we enjoy two
controls on the positional accuracy of the vertical unit: the fully
extended tripod legs on the floor and the horizontal connection
to the chair seat.
Horizontal stability. To that vertical unit, we attach Man-
frotto photography arms that have the advantage of strong con-
struction and almost universal adjustment possibilities. These
enable us to precisely tune the rig to each subject. Using these
arms, we can fix the horizontal stabilizers (forehead, nose, and
transducer) and maintain the spatial relationship of our subject
to our equipment (Figure 11).
The remaining components of the HATS equipment are:
 Transducer stabilizer/holder with a universal articulated
fitting so the transducer can be adjusted on all planes;
 Forehead rest with an additional velcro strap to fit around
the subjects head to keep it in contact with the forehead
rest;
 Under-the-nose reference arm to help assure the vertical
stability of the subjects head and prevent head rotation;
B Before the HATS will be used during phase II, a course-

adjustment mechanism from an old microscope will be

added between the transducer holder and the transducer
to permit us to make very fine vertical adjustments to the
transducer. This ability will be vital to obtaining the best
tongue-surface imagery as the transducer must read up
through the fluid reservoir. FIGURE 15. US image embedded in a sagittal profile. The posterior
 Microphone for recording samples for acoustic analysis; tongue appears on the left. Note the angular shadow on the right side of
and the image. This is an acoustic shadow cast by the front of the mandible
 Separate side-view video camera that records throughout that unfortunately prevents us from seeing the extreme anterior tip of
the protocol, on its own tripod. the tongue. Illustration by Garyth Nair.
Angelika Nair, et al The LMM and Its Resonential Implications for Elite Singers 128.e22

FIGURE 16. Oral cavity producing a spoken // and /i/ vowel (left) and performing an LMM // and /i/ vowel (right). Notice the size and shape of
the fluid reservoir at the bottom of the wedge. It indicates that the mandible has dropped approximately 1 cm. (For interpretation of the references to
color in this figure legend, the reader is referred to the Web version of this article.) Illustration by Angelika Nair.

must make allowance for vertical jaw movement and still main- MRI
tain acoustic contact between the surface of the transducer and This part of the pilot study was conducted in collaboration with
the flesh on the bottom of the subjects jaw. the Medical University of Graz, Austria, Franz Ebner, head of
This continuity is accomplished through the use of a standoff radiology, and Gernot Reishofer, medical physics expert,
(water-filled, acoustically transparent reservoir) between the respectively. Measurements were performed on a Siemens Mag-
transducer head and the bottom of the jaw (Figure 12). By netom Tim TRIO using a 12-channel head coil combined with a
providing a generous amount of US gel between the (1) trans- four-channel neck coil. The imaging of the spoken and sung pas-
ducer head and the reservoir and the top of the reservoir and sages was performed with a temporal resolution of 0.8725 sec-
(2) the flesh on the bottom of the subjects jaw, acoustic conti- onds, frame rate 1.15/s. The sequence parameters for the 2D
nuity can be maintained so that the US signal remains viable no turbo flash sequence were as follows: echo time (TE) 1.49 ms,
matter what elevation the mandible attains during the produc- repetition time (TR) 868 ms, flip angle (FA) 10 , matrix 192
tion of each phoneme under study (Figure 15). 3 199, field of vision (FoV) 183 3 199 m2, 8 mm slice thickness.

FIGURE 17. Three MRI images showing the LMM at rest (left), back vowel (middle), and front consonant (right). The different lines showing the
SAR (turquoise), the positions of the mandible (green), and the larynx (blue) and their elevations (M, L). (For interpretation of the references to color
in this figure legend, the reader is referred to the Web version of this article.)
128.e23 Journal of Voice, Vol. 30, No. 1, 2016

One image at the time of each spoken and sung vowel was per-
formed using a turbo-spin-echo sequence with the following pa-
rameters: TE 17 ms, TR 2000, FA 122 , matrix 224 3 320, FoV
220 3 220, 2 mm slice thickness.
The subject was introduced to the same protocol as in the US
imaging and chose a comfortable pitch for the sung tasks.
Image analyses. The measurements from the US images
were obtained by using the onscreen measurement tool Screen
Calipers (Version 4.0). By calibrating the numbers of pixels to
the integrated axial length unit of the US machine measurements
for the elevation of the mandible were taken from the standoff
(Figure 16).
The acquired MRI imagery was displayed by means of the
syngo fastView (Siemens). The softwares integrated measuring
tool was used to obtain the mandible drop capacity and the
laryngeal elevation. For the mandible and laryngeal drop, a pro-
tocol with the following landmarks was applied (Figure 17). A
horizontal line was drawn from a posterior stable anatomical
reference (SAR), right around the atlas and axis area. From FIGURE 18. Increase in vocal resonance during LMM. The dark
there, in a 90 angle, the distance to the bottom midpoint of herring-boned area denotes the resonance space available during a
the mental protuberance (prominence of the chin at the anterior non-LMM /i/ vowel. The white areas are the spaces that are added dur-
part of the mandible bone) was taken to measure the jaw open- ing LMM. Image from our collaboration with the Medical University
ing and the laryngeal drop. of Graz, Austria.
To obtain the values of LMM from both US and MRI, mea-
surements were taken from the spoken/sung [] and [i] vowel as
well as the sung Italian phrase. relaxation of the singers posterior mandible that can produce
The values for both the mandible and the laryngeal move- enormous resonance gains through enlargement of space in
ment were normalized by calculating the resting mandible/lar- the oral cavity as well as concomitant resonance pharyngeal
ynx to the SAR (ie, D0 Rest  Rest). Subsequently, the area gains because of a significant drop in laryngeal elevation
mandible/laryngeal drop of any phoneme was subtracted from (Figure 18). These gains in resonance capacity for all areas of
its resting position (ie, D[] Rest  max). the instrument acoustically interact and are a major component
The formula for the efficiency value (%) of the laryngeal in the type of rich tone we associate with classical singing.
drop derived from the mean of range of drops divided by the (Phase II of this study will attempt to ascertain the prevalence
maximum drops (Rest  max). of LMM in the technique of elite singers.) When skillfully
used by the classical singer, LMM appears to be applied to
%Effcy mean drop=max: drop the production of all a singers vowels and consonants.
Figure 19 showing a spoken/sung comparison of the vowels
Acoustic considerations. Audio recordings of each subject [] and [i] from subjects 1 to 4 in the US and subjects 1 to 5 in
were made during the US and the MRI scanning procedure. For the MRI procedure. Overall, all four subjects showed a clear
the US, a headset microphone (Known-Brainer) was used and employment of LMM when singing. The average of the vertical
run through an external USB sound card (Andrea PureAudio drop ranged between 0.7 and 2.09 cm. Within the sung vowels,
USB) for optimal performance. Acoustical signals during the subjects 1 and 2 showed an increased drop of the mandible on
MRI scanning procedure were acquired through a commercial the vowel [i] versus subjects 3 and 4 who showed an increased
microphone and recorded with Windows sound recorder. drop on the vowel [].
Acoustical documentation was measured by means of the Pratt Also of note is subject 3 whose spoken vowels showed an
software (University of Amsterdam, NL) and VoceVista 3 (D. almost sung LMM use, indicating an impressionistically
Miller). Because a normal within-subject variability can be ex- sung spoken vowel. With the employment of LMM an ex-
pected in the toggle between spoken and sung [] and [i] vowel, pected increase in intensity can be observed (Table 2). Howev-
a single repetition may not be an ideal exemplar of the task. er, compared with the other subjects, the increased LMM on the
Thus, actual measurements were taken for the second repetition spoken vowels of subject 3 did not show a concomitant increase
and midpoint of the vowel. However, because of the loudness of in intensity but rather equal (0.5 dB) to subject 4 with less
the MRI noise, few viable audio samples could be used for LMM.
actual measurement. Measurements from the MRI procedure confirmed the drop
of the mandible. However, in three of the four US subjects, a
RESULTS general increase of LMM on all tasks was observed
The LMM is a specialized resonance creation technique that (Figure 19, middle). In particular, the sung vowels showed
seems to be used mostly by elite singers. It features a downward double (subjects 1 and 4) to quadruple (subject 2) the increase.
Angelika Nair, et al The LMM and Its Resonential Implications for Elite Singers 128.e24

TABLE 2.
Intensity Values for the Sung and Spoken Vowels From
the US Procedure
Intensity (dB SPL) From US

Task Singer Vowel Mean Max Min

Spoken Subject 1 [] 46.7 48.3 40
[i] 38.6 41 32.2
Subject 2 [] 51.8 55 45.6
[i] 47 51 39.8
Subject 3 [] 55.5 57.7 48.9
[i] 42.2 44.4 37.2
Subject 4 [] 56 57.7 50.9
[i] 43 44.4 40
Sung Subject 1 [] 65.1 67.6 63.1
[i] 61 65.4 57.3
Subject 2 [] 65.3 68.7 63.2
[i] 57.4 60.4 54.9
Subject 3 [] 69.3 71.7 61.9
[i] 54.6 56.3 52.2
Subject 4 [] 71.3 76.6 55.4
[i] 61.9 63.6 54.7

spoken and sung vowels [i], respectively, subject 3 has little

mandible and even less laryngeal drop. On the sung vowel
[i], subject 2 showed a 29.6-mm mandible versus a 9.7-mm
laryngeal drop.
Also of note is subject 4 who showed more laryngeal than
mandible drop on the spoken vowel [] and sung vowel [i] as
well as equal drop on the sung vowel [] of both the larynx
and mandible. Subject 5 on the other hand showed little
laryngeal drop on the spoken vowels and the sung vowel []
but more on the sung vowel [i] than the mandible drop.
A major difference in the efficiency of LMM on a sung Ital-
ian phrase can be observed between US and MRI (Figure 22).
Subjects 3 and 4 in particular showed a 40% higher efficiency,
FIGURE 19. Spoken and sung LMM comparison of the vowels [] whereas the LMM efficiency of subject 1 was almost identical
and [i] from the US procedure (top). LMM comparison of the vowels (+2% in MRI). In the laryngeal elevation, subject 1 showed the
[] and [i] spoken (spk) and sung (sng) of each subject from the MRI highest laryngeal elevation with 92.3% followed by subject 4
procedure (middle). Comparison of the laryngeal elevation on the with 86.9%, subject 5 with 84.2%, subject 3 with 63.9%, and
spoken (spk) and sung (sng) vowels [] and [i] from the MRI procedure subject 2 with 46.7%.
(bottom; light gray and white columns alternating between subjects).

Also of note is subject 3 who showed a decrease of LMM on DISCUSSION

both spoken and sung vowels. In addition, with the applied The primary goal of this pilot study was to show that the LMM
LMM, an overall enhancement of the harmonics and the Fs is a resonance creating technique used by classical singers and
in particular can be observed (Figures 20 and 21). The its anatomical and physiological effects on the surrounding
increased LMM of subject 2 showed a substantial increase structures. Because only five subjects were examined, no statis-
in both Fs and the first harmonics (H2 and H4 in [] and tical analyses can be made. The use of US allowed for real-time
H1 in [i]). The Fs in subject 4 increased considerably and videos of the singers in action on individual phoneme produc-
enhanced H2 and H4 in the vowel [i]. Subject 3 on the tion and phoneme-to-phoneme transitions during the LMM.
other hand had a slight enhancement of Fs, H2 in the vowel The reason for the use of MRI and US was to correlate both im-
[] and H1 in the vowel [i]. aging modes, double check the US interpretation accuracy, and
Figure 22 shows the laryngeal elevation on the spoken and to study ancillary structures not visible in the US (ie, TMJ/
sung vowels. Overall, with the LMM, a concomitant laryngeal condyle environment, laryngeal elevation, etc).
drop can be observed in all subjects. The average from the ver- On the images shown from our study, one can easily see the
tical drop ranged from 0.14 to 2.73 cm. However, on the substantial difference in the actual tongue shape and position
128.e25 Journal of Voice, Vol. 30, No. 1, 2016

FIGURE 20. Graphics from the US procedure of the mean amplitude (dB) of the fundamental frequency (F0), the harmonics in the range of the
first formant (F1) for [] and [i], as well as in the range of 2.03.5 kHz. Various pitches and multiples of the fundamental frequency (nF0) are indi-
cated in the horizontal axis of each graph. Note, because of illness, subject 5 could not participate in this imaging procedure and is not included in this
illustration.
Angelika Nair, et al The LMM and Its Resonential Implications for Elite Singers 128.e26

FIGURE 21. Graphics from the MRI procedure of the mean amplitude (dB) of the fundamental frequency (F0), the harmonics in the range of the
first formant (F1) for [] and [i], as well as in the range of 2.03.5 kHz. Various pitches and multiples of the fundamental frequency (nF0) are indi-
cated in the horizontal axis of each graph. Note, the audio of subject 1 was not usable and could not be included in this illustration.

for the vowels. Both modes of imagery showed that LMM was sits much lower in the oral cavity. Hence, the tongue must
used while singing, with the concomitant lowering of the larynx work more intensely on the vertical plane to achieve the Point
and elongating the pharynx thus ultimately forcing the tongue of Articulation necessary for all phonemes (Figure 23). Howev-
to be far more active than in speech. The floor of the tongue er, this also means that the airspaces surrounding the tongue
128.e27 Journal of Voice, Vol. 30, No. 1, 2016

be validated. On the other hand, subject 3 showed a decrease

of LMM in the MRI. Because we observed it in both spoken
and sung vowels, a contrary speculation could be made that
for this singer, the standoff may have stimulated a more
conscious mandible movement, thus feeling the need for a
more prominent mandible drop. This may be echoed by the
acoustical measurements. A more active mandible elevation
of subject 3 could be observed although it did not seem to trans-
late into the acoustic outcome like in the other subjects. The
enhancement of the harmonics and the Fs, respectively showed
a minimal increase although the applied LMM was less in the
MRI than in the US. Perhaps, this may be correlated to the inter-
action of mandible and larynx. It could be speculated that the
employment of the LMM in subject 3 was rather more forceful
than naturally relaxed, hence not allowing for a concomitant
drop of the larynx and ultimately preventing an acoustic energy
increase.
Also, we observed a correlation between the laryngeal eleva-
tion and the LMM though not proportionate to the mandible
drop. Subject 4 showed more laryngeal drop on the spoken
vowel [] and sung vowel [i]. We believe this may be a probable
consequence of the complex interaction of the head and neck
muscles. In the case of subject 4, dropping the mandible with
a more pronounced relaxation of the masseter and an even
greater relaxation of the suprahyoids (joining the mandible to
the hyoid bone) and the infrahyoids (joining the hyoid bone
to the sternum and clavicle) created a seemingly posterior angle
of the mandible (Figure 24). Because the stability of the joint is
maintained by constant activity of the muscles that pull across
the joint, primarily the elevator, even in the resting state, these
muscles are in a mild state of contraction. However, when the
mouth opens, a controlled relaxation and lengthening in the
masseter can be observed. In addition, when the muscle groups
controlling the elevation of the larynx and hyoid bone are
FIGURE 22. Overall efficiency of LMM from US (top), MRI (mid- relaxed, like taking a breath, this will lower the larynx. Thus,
dle), and the laryngeal elevation (bottom) on an Italian sung phrase. with the measurements taken from the prominence of the chin
Note, because of illness, subject 5 is missing in the US. at the anterior part of the mandible bone, the actual drop of
the larynx seems to be correlated but in a nonproportional unit.
will produce far more resonance. Concomitant to the increase Additional aspects that could possibly be influential factors
of the LMM, measurements of the energy increase confirmed are the overall posture and the cervical spine. Sataloff24 points
a considerable enhancement of all harmonics and the Fs, respec- out that the extrinsic muscles are critical in maintaining a stable
tively (subjects 2 and 4 in particular). laryngeal skeleton that permits effective movement of the deli-
However, in three of the four subjects, we observed a cate intrinsic musculature. With the finely tuned balance of
significant increase of LMM in the MRI procedure. Possibly, head and neck muscles, hyperextension will certainly influence
this increase was associated with the standoff (water reservoir) the movement of the larynx and make it difficult to relax.
in the US that allows jaw movement but still maintains acoustic A critical question in classical singing resonance is the con-
contact between the surface of the transducer and the flesh on sistency of this maneuver, phoneme-to-phoneme. Although this
the bottom of the subjects jaw. Hence, the singer is feeling a will be a subject for another article, for the sake of complete-
slight resistance to the jaw and may subconsciously refrain ness, a few facets shall be included in this discussion. Unless
from relaxing the mandible to the maximum. Solutions for the singer has worked to perform all phonemesnot just
the future may be a training session in which the singer vowelsin the optimal acoustic environment of the LMM,
can get more acquainted with the sensation of the standoff his or her sound will lack consistency (Figure 25). For this
and its enlargement to the point before it loses acoustic continu- reason, we have measured the efficiency of LMM while singing
ity so the mandible has even more room for movement. Also, an Italian phrase and observed major differences in three of the
two of the four MRI subjects showed a double increase of the four subjects between US and MRI. Especially, subjects 3 and 4
mandible drop. However, a conclusion to double the measure- showed a 40% higher efficiency which could be attributed to the
ments taken from the US would need many more subjects to previously mentioned standoff resistance in the US procedure.
Angelika Nair, et al The LMM and Its Resonential Implications for Elite Singers 128.e28

FIGURE 23. Oral cavity producing a spoken // and /i/ vowel (left) and performing an LMM // and /i/ vowel (right). Notice the size and shape of
the fluid reservoir at the bottom of the wedge (green) as well as the advanced tongue root on the sung /i/ (circled and indicated by an arrow). The latter
indicating an increase in pharyngeal volume. (For interpretation of the references to color in this figure legend, the reader is referred to the Web
version of this article.) Illustration by Angelika Nair.

However, comparing the LMM and laryngeal drop efficiency in But, let us have a look at the internationally ranked singer
the MRI, we observed (in that order) that subjects 1, 4, and 5 Thomas Hampson singing a single la from Rossinis Largo
maintained the highest laryngeal elevation while using an al factotum (Figure 27). In these two images, one can clearly
LMM strategy with an entire jaw drop (Figure 26C) such as see that the entire jaw platform is down and that the tongue is
subjects 1 and 5, and a posterior drop of the mandible doing all the work of the language. This use of the LMM engen-
(Figure 26D) such as subject 4. For subject 2, however, the ders optimal resonance in the posterior oral cavity and allows
LMM only reflects in a 47% laryngeal drop and subject 3, the shift from /l/ to // with a minimum of movement and a
63.9%. Unfortunately, with the loudness of the MRI noise, an consistently large resonance space. The resulting sound is a
acoustical analyses of the entire phrase were not possible. How- Hampson hallmark and is one of the factors that place him in
ever, some extracts show similar harmonic enhancement as in the highest rank of todays singers.
the sung vowels.
To summarize the results, we found that all five subjects: CONCLUSIONS
(1) utilized the LMM with greater or lesser success; The images from both US and MRI show that the LMM is a
(2) exhibited the mandible strategies shown in Figure 26; technique used in classical singing. The strategies of the
and LMM varied between the entire drop of the jaw platform and
(3) radiated signal correlated with the results shown in the the posterior drop of the mandible, but both resulted in a
imagery. concomitant drop of the larynx. However, the enhancement of
the first harmonics and the Fs as well as the increase of the in-
Also, there is visual and aural evidence available for the ex-
tensity was greater with the drop of the entire mandible. Clear
istence of the LMM: modifications in the tongue shape for all phonemes were
(4) far higher percentage of mouth opening during sung observed when the LMM was used to maintain the integrity
passages; of the phoneme while increasing the sound output. In a column
(5) visible evidence of a lowered posterior mandible (the of the Journal of Singing, Titze25 addressed the question why
angle of the ramus); classical singers widen their airways in the pharynx and the
(6) where visible, the laryngeal prominence (the protruding back of the oral cavity. Acoustically, the answer is that a large
Adams apple) reveals a consistently low larynx; and mouth opening is needed to radiate lots of sound to the
(7) less visible evidence of superficial tension in the lower listenersuch as brass and woodwinds instrument that have a
neck region. bell to get more acoustic power into free space. However, an
128.e29 Journal of Voice, Vol. 30, No. 1, 2016

FIGURE 24. Top right image, spoken /i/ and the left image, sung /i/ with full classical resonance. Note the drop of the mandible and the larynx.
Note that, in the top right-hand static image, the singers velopharyngeal port was not closedin a moving MRI, the port would have been closed and
would appear identical to the corollary region in the top left side image. Bottom images showing spoken // (right) and sung // (left) with full clas-
sical resonance. Also note the different angles following the LMM (yellow line). (For interpretation of the references to color in this figure legend, the
reader is referred to the Web version of this article.) Images from our collaboration with the Medical University of Graz, Austria.

inverted megaphone mouth shape is also needed to produce the thereby allowing the singer to sense the precise position
mixed registration with vowel modification. Hence, there is a of the tongue and how it is shaped); The front portion of
need to expand the throat. the tongue has many proprioceptors (those nerves that
The comparison of US and MRI imagery showed similar out- give the brain the awareness, often subconsciously, of
comes in shape and direction of the tongue and mandible weight, posture, movement, position in space in relation-
although the drop of the mandible measured in the US does ship to the body; based on sensory input from nerve ter-
not represent the actual drop. However, the deformation of minals in joints and muscles . .26 However, as we
the standoff does show clear directions as well as the angle of progress back along the tongue body, the density of pro-
the mandible drop. Also, with a frame rate of 30 fps, more prioceptors drops precipitously. Thus, the singer has a
tongue shapes of all phonemes can be observed. difficult time sensing where the tongue is and how it is
We are already working to quantify the useful factors of the shaped.
biodynamic system for use in the voice studio. With the use of  Faster correction of a singers phonemic problems;
US imagery in the voice studio (supported by MRI), the authors  Teaching the LMM; and
of this study have already experienced a rapid acceleration in  Redefining vowel shapes as well as consonants resonance.
their students understanding of these complex vocal strategies.
However, with the acquisition of a portable US machine (GE Most classical singers in training unconsciously attempt to
Logiq E) and its use in the voice studio, we are seeing even use LMM in their vowels as a way of attempting to duplicate
greater benefits to all classical singers in training: the classical vowel tone they hear around them. However,
within these vowel attempts, one rarely encounters singers or
 biofeedback for phoneme rehabituation (overcoming the pedagogues discussing the resonance requirements of conso-
problem of relative lack of tongue proprioceptors, nants. It is hoped that the collection of such imagery might
Angelika Nair, et al The LMM and Its Resonential Implications for Elite Singers 128.e30

 a better understanding of the previous verbally described

area and/or task;
 the images and shapes of the tongue did not match what
they had envisioned themselves;
 better control of the tongue because of the visual feed-
back; and
 sensitization to ones resonance creation, particularly
pharyngeal.

Similar eye-body biofeedback is already in use in the form of

virtual real-time spectrography. This technology is now in
extensive use in many voice studios worldwide, including
ours at Drew University and the College of St. Elizabeth. Stu-
dents progress is greatly accelerated once we present them
with a graphic of the acoustics of their singing. This innovative
pedagogical aid and the practical integration of new knowledge
into musical training were pioneered by Prof. Garyth Nair in his
book, VoiceTradition and Technology.27
As with any new imagery, some guidance is necessary to be
able to read the US imagery. However, the authors already inte-
grated the US in their own voice studio and use it in workshops
and master classes. We have found that after a few minutes of
FIGURE 25. Comparison of spoken (left) and sung (right) conso- introduction, the student very quickly familiarizes him/herself
nants. Note again the size and shape of the fluid reservoir on the bottom with the tongue contour and starts to learn to control its
wedge. EdgeTrak software capturing the edge of the sagittal tongue movement.
profile (red line) as well as the hard palate (yellow line). (For interpre- It would be financially unreasonable to expect voice peda-
tation of the references to color in this figure legend, the reader is gogues to buy their own portable US machine. The solution to
referred to the Web version of this article.) this problem lies in the development of USB US transducers
which can be used directly with any Windows computer (laptop,
add to our scientific knowledge of the physioacoustics of upper- tablet, or desktop). Unfortunately, the current state of USB trans-
level resonance creation and ultimately lead to the creation of a ducers does not meet the specifications necessary, especially for
resource that may enlighten pedagogical approaches for all sounds such as stops, clicks, and flaps. To obtain all the sounds,
singers, not just those of the highest rank. A new book The Se- one needs a minimum of 30 fps; and USB transducers at the
cret Life of Consonants: why dont my songs sound as good as moment only reach a maximum of 15 fps (though enough for
my warm-ups? is already in the making by Angelika Nair, using just vowels). It is the authors hope that, with the development
the US and MRI material and results acquired from this of USB probes, this will change and find its way into the voice
research project. studio, similar to the virtual real-time spectrography.
Some of the preliminary remarks from the students them- Beside the different types of equipment we are currently work-
selves are: ing with (portable US, laptop, converter, cables, etc), we also

FIGURE 26. Different mandible strategies found in all five subjects. (A) mandible at rest. (B) Anterior drop of the mandible (speech template). (C)
Complete LMM, entire jaw platform is down. (D) Posterior drop of the mandible.
128.e31 Journal of Voice, Vol. 30, No. 1, 2016

FIGURE 27. The internationally ranked singer, Thomas Hampson, in two frames from a video of him singing the word la from Rossinis Largo
al factotum. The two images are just 1/30000 of a second apart in time. On the left, one sees the /l/, and on the right, the open //. The two guidelines
indicate the elevation of the posterior and anterior mandible.

have to combine both sets of images on the screen by capturing It is a very short leap from the use of US as feedback in the
the US video on the laptop through the Debut Video Capture soft- voice studio to its use by the speech-language pathologist to
ware in one Window and running the spectrogram (Gram Lite, help those with speech defects correct them faster. Thus, the
VoicePrint) software in another Window. The vision for the future gains for such use of US are not limited to research but may
with the USB transducer would be to combine the US with the have a broader application big enough to constitute a new mar-
real-time spectrum analyses within one software, similar to Voce- ket for US used in this way.
Vista which displays the signals in combination of spectrogram, Voice science has not pointed out any new techniques during
power spectrum, and high time-resolution waveforms. We hope the past 30+ years, but it has shown the efficacy of certain tech-
to have the interest of Donald G. Miller to consider the ability nical approaches and helped to illuminate and eliminate un-
to include US real-time imagery in VoceVista (Figure 28). healthy vocal practices. There are areas of the voice yet to be
Additionally, in a recently started collaboration with the studied just like this study of the LMM, that will benefit all
Interdisciplinary Speech Research Laboratory (ISRL) at Uni- singers and teachers of singing into the future.
versity of Columbia, Vancouver (research project: The Efficacy
of Ultrasound in the Voice Studio), we have also presented our
vision to the Electrical and Computer Engineering (ECE). The
ISRL and ECE developed ArtiSynth (www.artisynth.org), a
state-of-the-art biomechanical simulation platform focused on
modeling the human vocal tract. It is our hope for the future
that this 3D mechanical modeling system may be developing
into a real-time computer animation, showing the US signal
of the tongue in 3D.
All these developments may take us toward removing facets
of vocal technique from the realm of opinion and into a more
objective understanding of how the instrument works. Given
the nonfact-based technique that we routinely encounter in
singerstraining that cannot be understood in terms of anat-
omy or physicswe need more voice science that can clarify
how the instrument actually works and that can help move
pedagogy more toward the objective side (a movement, that FIGURE 28. Authors vision of US through a USB transducer, incor-
has already begun in the best voice teaching centers). porated in VoceVista and run on a tablet. Illustration by Angelika Nair.
Angelika Nair, et al The LMM and Its Resonential Implications for Elite Singers 128.e32

Acknowledgments 11. Nair G. Acoustics and physiology of singers consonantsa little explored
The authors would like to acknowledge Dr. Maureen Stone for area. Paper Presented at the 2nd Physiology and Acoustics of Singing Con-
ference, Denver, 2004.
her generous contributions in her consultations and lab equip- 12. Traser L, Burdumy M, Richter B, Vicari M, Echternach M. The effect of
ment; Dr. Wolfgang Klinger (Neurologist) for making his prac- supine and upright position on vocal tract configurations during
tice with its GE Logiq 7 ultrasound machine available to us; and singinga comparative study in professional tenors. J Voice. 2013;27:
Univ. Prof., Dr. med., Dr. phil., Franz Ebner (clinical expertise) 141148.
and team of the Medical University of Graz, Austria for 13. Stone M, STock G, Bunin K, Kumar K, Epstein M. Comparison of speech
production in upright and supine position. J Acoust Soc Am. 2007;122:
providing us >150 MRI scans. Last but not least, the authors 532541.
also thank Anne Jacobson for improving my English. 14. Okeson JP. Management of Temporomandibular Disorders and Occlusion.
6th ed. St. Louis, MO: Mosby; 2008.
15. Sundberg J, Skoog J. Dependance of jaw opening on pitch and vowel in
singers. J Voice. 1997;11:301306.
REFERENCES 16. Story BH, Titze IR, Hoffman EA. The relationship of vocal tract shape to
1. Sundberg J. Articulatory interpretation of the singing formant. J Acoust three voice qualities. J Acoust Soc Am. 2001;109:16511667.
Soc Am. 1974;55:838844. 17. Garcia M. Hints of Singing (B. Garcia, trans.). NY: Edward Schuberth Co;
2. Titze IR. Acoustic Interpretation of Resonant Voice. J Voice. 2001;15: 1894.
519528. 18. Appelman DR. The Science of Vocal Pedagogy, Theory and Application.
3. Raphael Bonnie N, Scherer R. Voice modifications of stage actors: acoustic Bloomington, IN: Indiana University Press; 1967.
analyses. J Voice. 1987;1:8387. 19. Vennard W. Singing. The Mechanics and the Technic. rev. and greatly
4. Sundberg J, L~a Filipa MB, Gill Brian P. Professional male singers formant enlarged ed. New York, NY: Carl Fischer; 1967.
tuning strategies for the vowel /a/. Logoped Phoniatr Vocol. 2011;36:156167. 20. Nair G. Consonant resonance and the highly skilled singera study in
5. Sundberg J. Articulatory configuration on pitch in classically trained so- physio acoustics. Paper Presented at the 7th International Voice Symposium
prano singers. J Voice. 2009;23:546551. on Care and Cure of the Professional Voice, Austria, Aug. 8th 2006.
6. Echternach M, Sundberg J, Arndt S, Breyer T, Markl M, Schumacher M, 21. Nair G. The Craft of Singing. San Diego, CA: Plural Publishing; 2007.
Richter B. Vocal tract and register changes analysed by real-time MRI in 22. Stone M. A guide to analyzing tongue motion from ultrasound images. Clin
male professional singersa pilot study. Logoped Phoniatr Vocol. 2008;33: Linguist Phon. 2005;19:455502.
6773. 23. Li M, Kambhamettu C, Stone M. Automatic contour tracking in ultrasound
7. Echternach M. Vocal tract in female registersa dynamic real-time MRI images. Clin Linguist Phon. 2005;19:545554.
study. J Voice. 2010;24:133139. 24. Sataloff RT. 3rd ed. In: Professional Voice. The Science and Art of Clinical
8. Sundberg J. The Science of the Singing Voice. DesKalb, Illinois: Nothern Care, Vo. 1. San Diego, CA, Oxford, UK: Plural Publishing; 2005.
Illinois University Press; 1988. 25. Titze IR. Why do classically trained singers widen their throat? J Sing.
9. Oliveira Barrichelo VM, Heuer RJ, Dean CM, Sataloff RT. Comparison of 2012;69:177178.
singers formant, speakers ring, and LTA spectrum among classical singers 26. Singh S, Kent R. Illustrated Dictionary of Speech-Language Pathology. San
and untrained normal speakers. J Voice. 2001;15:344350. Diego, CA: Singular Publishing; 2000.
10. Smith CG, Finnegan EM, Karnell MP. Resonant voice: spectral and nasen- 27. Nair G. VoiceTradition and Technology. San Diego, CA: Singular Pub-
doscopic analysis. J Voice. 2005;19:607622. lishing; 1999.