Semi-Occluded Vocal Tract Exercises: Aerodynamic

and Electroglottographic Measurements in Singers

Troy Clifford Dargin and Jeff Searl, Kansas City, KS
Summary: Objective. The purpose of this study was to describe changes in aerodynamic and electroglottographic
(EGG) measures immediately after completing three semi-occluded vocal tract (SOVT) exercises.
Study Design. Prospective case series.
Methods. Aerodynamic and EGG measurements were obtained before and immediately after performing three
SOVTs (straw phonation, lip trill, and tongue trill) in four singers for prepost comparisons to evaluate laryngeal changes
persisting beyond the execution of SOVTs.
Results. Mean air flow, sound pressure level, and EGG closed quotient tended to increase after completing SOVTs.
The magnitude of change and consistency of change in measures across the SOVTs varied from subject-to-subject.
Conclusions. Aerodynamic and EGG changes did occur during and immediately after completing SOVTs. However,
there was marked variability within and across participants. Further investigation is needed to better understand which
SOVTs are likely to benefit a particular individual.
Key Words: Semi-occluded vocal tractAerodynamicElectroglottographySingers.

INTRODUCTION
Semi-occluded vocal tract (SOVT) exercises involve narrowing
the vocal tract, usually near the lips or tongue tip, while voicing.
These exercises have been used for many years by singers and
voice professionals as warm-ups and more recently have been
incorporated into therapeutic approaches by speech-language
pathologists (SLPs) for people with voice disorders. Although
there is a growing body of research about the physiological
impact of SOVTs, and growing clinical sentiment about the
therapeutic benefits, empirical data are lacking describing
expected laryngeal adjustments, identification of those for
whom the exercises may be of most benefit, dosing of the exercises for maximum gains, and other issues. This study contributes information about within subject changes in aerodynamic
and electroglottographic (EGG) measures of voice after
completing a brief trial of three different SOVTs.
Types of semi-occluded vocal tracts
SOVTs vary from high-to-low resistance to air flow in the vocal
tract. Increased resistance is created by narrowing or lengthening of the vocal tract; conversely, lower resistance is created
by opening or shortening the vocal tract. Examples of SOVTs
listed from higher to lower resistance are as follows: phonating
while holding a straw between the lips, humming, sustaining a
voiced labiodental fricative, voicing during lip or tongue trill,
sustaining voiced alveolar or velar nasal consonants, and
Accepted for publication May 23, 2014.
Portions of this work have been presented at The Voice Foundation, Philadelphia, PA,
June 2, 2012 (Speaker Differences in Laryngeal Activity, The Voice Foundation, Philadelphia, PA, June 2, 2012) and the National Center for Voice and Speech, Salt Lake City,
UT, July, 9, 2011 (Stroboscopic and Aerodynamic Changes Associated with SemiOccluded Vocal Tract Exercises-Preliminary Data,).
From the Hearing and Speech Department, University of Kansas Medical Center,
Kansas City, KS.
Address correspondence and reprint requests to Troy Clifford Dargin, 1000 Sunnyside
Avenue, 3017 Dole Human Development Center, Lawrence, KS 66045-7555. E-mail:
troydargin@yahoo.com
Journal of Voice, Vol. 29, No. 2, pp. 155-164
0892-1997/$36.00
2015 The Voice Foundation
http://dx.doi.org/10.1016/j.jvoice.2014.05.009

sustaining high tongue vowels.1 It is unclear whether the

respective SOVTs impact voice production in the same manner
or to the same magnitude within or across individuals. Two of
the more commonly used SOVTs that are included in the
present study are briefly reviewed in the following.
Straw phonation. Straw phonation involves holding a straw
between the lips while producing a sustained vowel. The length
and diameter of the tube can be altered to result in more or less
resistance to air flow. Phonating through a straw increases the
amount of intraoral air pressure that is generated during phonation. The increased intraoral pressure results in better impedance matching at the glottis.1 Titze2 found that phonating
with a high lung pressure and high pitch during straw phonation
could be completed without vocal fold trauma because the
vocal fold amplitude of vibration was relatively small due to
the aerodynamic changes induced by this SOVT. The acoustic
energy reflected back to the vocal folds from the straw and
the vocal tract aid in lowering the phonation threshold pressure
(PTP), which is the lowest amount of lung pressure required to
initiate vocal fold vibration.3
Lip and tongue trills and raspberries. Trills and raspberries have been incorporated into training of singers and
vocal warm-up routines with a variety of intended purposes.
Nix,4 based on observations as a teacher of singers, has suggested that these exercises are beneficial to singers who use
glottal onsets. Miller5 described these SOVT exercises as a
means of reducing tension in the tongue, jaw, and lips, and as
an assist in establishing legato lines. Some singing teachers
feel that trills and raspberries provide a sensation to the singer
of elevated pressure within the upper vocal tract which can
signal them to push less to generate the voice. During a lip
trill, the flow resistance at the lips is high and time-varying
(unlike straw phonation which produces only flow resistance),
resulting in low frequency oscillation of intraoral pressure
causing variation in transglottal pressure.2 A reduction in transglottal pressure results in less collision force between the vocal
folds. Bele6 reported that a lip trill reduced the amplitude of

156
vocal fold movement with reduced vocal fold collision due to
lower PTP. In contrast, Cordeiro, Montagnoli, and Tsuji,7
concluded that the maximum amplitude of vocal fold vibration
is greater during lip and tongue trills than in nonoccluded voice.
Changes during semi-occluded vocal tracts
Aerodynamic changes. Laryngeal aerodynamics are
expected to change when doing SOVTs. According to Titze,3
manipulation of supraglottal inertance and supraglottal resistance could be used to achieve optimal PTP values. PTP
increases with supraglottal resistance and decreases with supraglottal inertance. Straw phonation results in both of these
changes within the supraglottal tract, effectively canceling out
the respective influence on PTP and allowing for PTP to remain
within a target range of approximately 00.5 kPa.3 It has been
shown that PTP is elevated after excessive vocal use8 and sometimes after warm-up exercises at high pitches.9 SOVTs may provide one means of helping reestablish a more favorable PTP
after excessive use or as a warm-up exercise because of the
beneficial impact on PTP. The aerodynamic changes that are
derived from SOVTs are expected to lessen vocal fold loading.2
Vocal fold movement and closure changes. The lessening of the vocal load resulting from SOVTs should be
apparent in other measures of phonation such as the closed quotient from EGG and vibration amplitude from stroboscopic
measures. Stroboscopic studies of laryngeal changes during
or after SOVTs are limited. Cordiero et al,7 performed a study
with 14 healthy subjects that had at least 3 years of professional
classical voice work and who used lip and tongue trills on a regular basis. The subjects produced a sustained nonoccluded /3/
vowel, lip trills, and tongue trills while videolaryngostroboscopy and EGG were performed. Stroboscopically, they found
that tongue and lip trills had similar amplitude of vocal fold
vibration (at high and low intensity voicing), which were both
larger amplitudes compared with those during a nonoccluded
/3/ vowel. Schwarz and Cielo10 also evaluated subjects stroboscopically before and after completing SOVTs. Twenty-four
women performed three series of 15 trills with a 30 second
interval of rest between each series. They found a statistically
significant increase in vibration amplitude after the SOVT exercises. In a group of singers and nonsingers, contact quotients
(CQs) as derived from EGG were reduced by about 50%
when phonating on a vowel /a/ as compared with a lip trill in
a study by Gaskill and Erickson.11 The reduction in CQ was
more prominent for the nonsingers. A study of the Mediterranean tongue trill, which is used during joyful celebration, also
showed a reduction in CQ.12 In contrast, Cordiero, Montagnoli,
and Tsuji7 reported an increase in the mean CQ at high intensity
but not low intensity. Additionally, they found that intraspeaker
standard deviation in CQ was significantly elevated during the
lip and tongue trills compared with the sustained vowel. Cordiero et al suggested that to complete the trill subjects must
increase pulmonary air flow to sustain vibration of both the
vocal folds and the lips or tongue. Greater pulmonary drive
can elevate subglottal pressure causing greater excursion of
the vocal fold edge during vibration.13

Journal of Voice, Vol. 29, No. 2, 2015

Acoustic changes. SOVTs are expected to change how the

voice is produced. These changes should be evident in the
acoustic signal. Titze14 wrote that resonant voice depends not
on concentrating on the sensation of feeling something in
the mask of the face, but rather on converting aerodynamic
energy into acoustic energy. He further explained that this is
accomplished, in part, by an inertive vocal tract wherein PTP
is lowered when the vocal tract inertance is increased. Tissue
vibration in the mask, is the result of high acoustic pressure
in the mouth.15
Using computational modeling, Titze and Laurkkanen1
reported that F1 is lowered from 300150 Hz when a resonance
tube was added to the vocal tract model. Adding tube length
increased the positive inertive reactance below F1. Vocal tract
inertive reactance was doubled when F0 was 100 Hz during
an SOVT, resulting in a lowering of PTP, increased maximum
flow declination rate (MFDR), and a reduction in the vocal
fold collision force.
Schwartz and Cielo10 performed an SOVT study using a
multidimensional voice profile (MDVP) analysis before and
after participants completed SOVTs. In addition to improvements in vocal quality as judged by an SLP, each of the
following acoustic and spectrographic measures were
increased: mean F0 on a sustained vowel, formant intensity,
and intensity across the spectrum. Additionally, trained judges
found an increase in formant definition and spectrum regularity,
as judged from wide- and narrow-band spectrograms, after
completing the set of trills.
Semi-occluded vocal tract applications
Vocal warm-up. SOVTs have a long tradition in the singing
voice studio, both as a means of warming up the voice before
performing and as a potential training exercise to improve the
voice. For example, Miller5 and Lewis16 argued that the use
of humming aids in breath-management, resonation, and
freeing laryngeal tension. Miller further stated the following
regarding humming: (1) hums could help expand registration;
and (2) the inability to produce a comfortable hum could be
an indication that the jaw or the pharyngeal wall has tension
or the larynx is in an undesirable position. There is no standardized set of warm-up activities that singers are taught or use.
Voice therapy. SOVTs have become relatively common
place among SLPs who specialize in laryngeal voice disorders
as an intervention technique. Titze17 discussed that SOVTs
should reduce collision impact between the vocal folds during
vibration and are economic vocal exercises. For those who
have voice disorders one primary goal of therapy typically is
to minimize trauma and hyperfunction of the larynx. SOVTs
may provide such benefits.18
In clinical voice practice today, many therapists use the
Lessac-Madsen Voice Therapy (LMVT) developed by Verdolini.19 LMVT is based largely on activities from earlier resonant
voice therapy approaches that promote vibratory sensations on
the alveolar ridge and other facial plates during phonation.20
In this therapy approach, clients focus on a forward placement
of the voice with resonance in the mask (bony part) of the face.

Troy Clifford Dargin and Jeff Searl

157

SOVT Exercises in Singers

To help elicit these sensations, various SOVTs, such as humming, are completed. Several outcome studies support the
effectiveness of LMVT.2025
The present study compared aerodynamic and EGG
measures acquired immediately after completing each of three
commonly used SOVTs in four classically trained singers. The
intent was to evaluate each subject as their own case study given
the expectation that there might be substantial across subject
variability in the effect of each SOVT.
METHODS
Participants
A convenience sample of three male tenors and one female
soprano from one Midwestern and one East coast graduate
music program participated. Demographic and voice training
history for each are in Table 1. All participants were recruited
in compliance with the guidelines of the Institutional Review
Board at the University of Kansas Medical Center, and written
informed consent was obtained from each participant. None had
known vocal pathology at the time of scoping. None of the subjects reported a history of laryngeal disease or significant voice
trouble in their past. Tenor 3 reported a history with reflux but
was not currently on medication. The other three subjects
reported no reflux history or medication history of importance
to the study.
Instrumentation
Aerodynamic recordings were obtained with the Phonatory
Aerodynamic System (PAS Model 6600, KayPENTAX, Montvale, NJ). This system allowed for recording of air flow, air
pressure, and the acoustic signal. EGG was obtained using
the 6103 KayPENTAX (KayPENTAX, Montvale, NJ) system.
Procedures
Baseline. PreSOVT aerodynamic and EGG measures were
taken with the subject seated in a quite clinic room. The PAS
facemask-pneumotach system was held firmly on the face by
the subject. Research personnel confirmed the mask was seated
flush against the skin at the start of and throughout data
recording. The system was calibrated per the manufacturer
instructions before data collection. The two EGG electrodes
were place on the neck at the level of the thyroid cartilage.
The EGG signal was acquired simultaneously with the aerodynamic signals and was displayed in the PAS software
environment.

The Comfortable Sustained Phonation (CSP) subroutine of

PAS was opened. This program allowed collection of mean
expiratory air flow (L/s) and % closed quotient on a sustained
vowel. The participant was instructed to sustain the vowel /a/
at a comfortable pitch and loudness on one breath for about
5 seconds. Each participant completed three trials with a brief
rest between trials. Then, the voicing efficiency (VE) subroutine was engaged to gather sound pressure level (SPL), aerodynamic resistance (cm H2OL/s), mean air flow during voicing
(L/s), and mean peak air pressure (cm H2O) during a /pa/ syllable train. For this data collection, an oral air pressure tube was
inserted through the port on the face mask. The tip of the tube
rested approximately 1 cm inside the mouth behind the central
incisors and above the tongue. The participant was instructed to
produce /pa/ five times in one breath using a comfortable pitch
and loudness. Three repetitions with a brief pause between trials were recorded. Study personnel monitored the productions
perceptually for rate, pitch, and loudness parameters and
requested an additional production if participants varied from
the instructions.
Post semi-occluded vocal tract. For each subject, a
random order of the three SOVTs (straw phonation, lip trill,
and tongue trill) was determined. A subject engaged in 2 minutes of one of the SOVTs without the aerodynamic or EGG
equipment in place. Immediately after completing 2 minutes
of the SOVT aerodynamic and EGG measures were obtained
as previously mentioned for the baseline testing. The PAS
and EGG equipment was removed and the subject completed
2 minutes of the next randomly ordered SOVT followed by
acquisition of another round of aerodynamic and EGG recordings. Equipment was removed and the subjected completed
2 minutes of the final SOVT before obtaining another set of
aerodynamic and EGG recordings. Finally, the subject sat
quietly for 2 minutes before a final set of recordings were
obtained. This last set of recordings after sitting silently was
intended to represent a potential carryover time period after
completion of all SOVTs.
Measures and analysis
Mean expiratory air flow (L/s) and % closed quotient measures
were obtained from the CSP subroutine on a sustained vowel.
For each sustained /ah/, cursors were placed at the start and
end of the vowel and the CSP subroutine automatically calculated values using default settings. For each subject, an average
value for each measure was calculated from three trials of the
/ah/ at each recording interval (baseline, after each SOVT,

TABLE 1.
Demographic and Voice History Training of Participants
Subject
Tenor 1
Tenor 2
Tenor 3
Soprano 1

Gender

Age

Years of Classical
Voice Training

Graduate School Status

Regular Use of SOVTs

M
M
M
F

28
32
28
25

10+
10+
10+
10+

MMopera performance
DMAvocal performance
MMvocal performance
MMopera performance

Yes
Yes
No
No

158

Journal of Voice, Vol. 29, No. 2, 2015

and after the silence interval). SPL values, aerodynamic resistance (cm H2Oliters/sec), mean air flow during voice (L/s),
and mean peak air pressure (cm H2O) measures were obtained
from the VE subroutine using default settings. Given the small
subject numbers and the preliminary nature of this study, group
analyses were avoided in favor of a within subject descriptive
approach. Raw air flow data for T3 were exceptionally low
during baseline and after some SOVT trials suggesting either
equipment or procedureal errors. For that reason no air flow
or resistance values (ie, dependent on air flow measure for the
calculation) are reported for that subject.

phonation and 3 cm H2O (21%) for lip trills, but decreased by

0.5 cm H20 (4%) for tongue trills. T3 had a modest increase
on straw phonation (3 cm H2O37%) but was changed
marginally for lip trill (0.27 cm H2O3%) and tongue trill
(0.12 cm H2O1%) compared with preSOVT. Finally, S1
had a modest decrease (0.72 cm H20822%) after each
SOVT. Overall, there was no consistent pattern of mean air
pressure change within participants or across SOVTs. The
pressure values after completing the three SOVTs plus
2 minutes of vocal rest were decreased by 0.52 cm H2O for
three subjects (T24%, T310%, and S122%) and
increased by 1.5 cm H2O for one (T124%).

RESULTS
Aerodynamic data
Sound pressure level. The SPL values for the /pa/ syllable
train are in Table 2. Table 3 displays data by participant for SPL
(plus all other measures) denoting the direction of change from
baseline. To be marked as a change in the table a participant had
to demonstrate a change of 10% or greater from the preSOVT
sample. On average, SPL increased by approximately six to
seven SPL after completing each SOVT. However, there was
notable variation across subjects in the magnitude of the
increase. For tenors 1 and 2 the SPL increase was substantial
(greater than seven SPL) and occurred consistently across the
three SOVTs. For tenor 3, there was essentially no increase after
straw phonation or lip trill (approximately one SPL), and a
modest gain after tongue trill (three SPL). For Soprano 1 there
were modest gains after straw phonation and lip trill (three and
a half SPL), but no change after tongue trill. The SPL data gathered after 2 minutes of quiet after completion of all SOVTs indicated a substantial increase for two subjects (1 and 2), a modest
to small gain for another (tenor 3) and essentially no change for
the Soprano.

Mean air flow during voicing (L/s). Mean air flow data
during the /pa/ syllable train are in Tables 3 and 5. For T1
and S1, the three SOVTs resulted in at least a 1050%
increase in air flow over the precondition. For T2 a sizable
increase occurred for straw phonation (22%), a limited
increase for tongue trill (7%), but a decrease for lip trill (7%).
The three participants had an increase in mean air flow
immediately after completing the SOVTs plus 2 minutes of
rest (increases as a percent of the preSOVT recordings were
50%, 39%, and 13% for T1, T2, and S1, respectively).

Mean peak air pressure (cm H2O). Pressure data are presented in Tables 3 and 4. There was substantial variability in air
pressure changes across participants. Pressures for T1 increased
between 2.53 cm H2O (3946%) after each SOVT. T2s
pressures decreased by 3.5 cm H2O (24%) for straw

Mean expiratory air flow (L/s). Tables 3 and 7 displays

mean expiratory air flow data during sustained vowel. For
T1, air flow increased from 0.01 to 0.03 L/s, representing a
618% increase more than the baseline. T2 demonstrated
more substantial increases ranging from 0.18 on lip trill to

Aerodynamic resistance (cm H2O/L/s). These data are

presented in Tables 3 and 6. The magnitude and direction of
change varied within and across participants and across
specific SOVTs. Participant S1 showed the most consistent
changes with 2735% reductions in resistance for all three
SOVTs. T2 had at least a 10% decrease in resistance for two
of the three SOVTs (Straw and tongue trill). T1 had a 10%
increase after straw phonation, a 5% drop after lip trill, and a
2% increase after tongue trill. All three had >10% decrease in
resistance in the final recording after all SOVTs were done
plus 2 minutes of rest.

TABLE 2.
Sound Pressure Level (dB SPL) Values Before and After Each SOVT and After a Rest Period at the End of the Protocol
(PostSOVT)
Straw Phonation
Subject
Singer
T1
T2
T3
S1
Group
Mean
SD

Lip Trill

Tongue Trill

PostSOVT

PreSOVT

Mean

Versus Pre

Mean

Versus Pre

Mean

Versus Pre

Mean

Versus Pre

86.87
82.10
95.71
90.03

95.26
96.28
96.95
93.53

8.39
14.18
1.24
3.50

95.06
97.27
96.97
93.72

8.19
15.17
1.26
3.69

94.17
97.33
99.08
90.13

7.3
15.23
3.37
0.10

95.62
93.77
98.50
91.17

8.75
11.67
2.79
1.14

88.68
5.71

95.51
1.49

6.83
5.74

95.76
1.67

7.08
6.11

95.18
3.93

6.50
6.52

94.77
3.09

6.09
4.95

Abbreviations: T1, tenor 1; T2, tenor 2; T3, tenor 3; S1, soprano 1; SD, standard deviation.

Troy Clifford Dargin and Jeff Searl

159

SOVT Exercises in Singers

TABLE 3.
Predominant Direction of Change for Each Measure Across the Four Participants

Parameter
SPL
Mean expiratory
flow
Oral pressure
Mean flow during
voicing
Laryngeal
resistance
% Closed quotient

T1

T2

T3

S1

Outcome Versus
PreSOVT

Outcome Versus
PreSOVT

Outcome Versus
PreSOVT

Outcome Versus
PreSOVT

SP

LT

TT

CO

SP

LT

TT

CO

SP

LT

CO

SP

LT

TT

CO

>
>

>
>

>


>
>

>
>

>
>

>
>

>
>


>
No data

TT

>
>

>
>


>




>
>

>
>

>
>

>
>

<
>

<






>

>



No data

<

<
>


>

<
>

<
>

>

<

<

<

<

<

<

<

<

>

>

>

>

>

No data
>

>

>

>

Abbreviations: SP, straw phonation; LT, lip trill; TT, tongue trill; CO, carryover condition.
Notes: To be considered a change, the measure had to increase or decrease by  10% versus the preSOVT value.

0.26 L/s on tongue trill (78113%). S1 increased by 0.06

0.07 L/sec for all three SOVTs (1923% of preSOVT). After
completing all three SOVTs and resting for 2 minutes the
three participants increased in expiratory air flow by 6%
(S1), 25% (T1), and 35% (T2).
Electroglottography mean closed quotient (%CQ)
The closed quotient data are in Tables 3 and 8. In the postSOVT
tasks T1 increased 5% more than the preSOVT value. T2
demonstrated closed quotient increase of 12%. T3 increased
between 6% and 7% and S1 increased between 3% and 4%.
The closed quotient after completing all three SOVTs plus
2 minutes of vocal rest increased 5% (T1), 7% (T3), and 4%
(S1). T2 showed no change in closed quotient after
completing the three SOVTs plus the rest period.
Overall patterns of change by semi-occluded vocal
tract exercise
It was of interest to identify commonalities and differences
across the three SOVTs in terms of aerodynamic and EGG

changes that persist to the time period immediately after

completing each exercise. Table 9 provides a summary of
changes for the six measures obtained for each SOVT relative
to the values from the preSOVT phonation. A measure had to
increase or decrease by at least 10% to be considered changed
from the baseline. The 10% mark was arbitrarily chosen as a
working definition of change. It remains to be determined
what magnitude of change is meaningful in terms of function
or perception.
There were some consistent patterns of change that occurred
not only within measures for a given SOVT but also for specific measures across the three SOVTs. The SPL and mean
expiratory air flow during sustained vowel production
increased after each SOVT for most of the participants.
Mean air flow during /pa/ also increased after each SOVT
for a majority of participants. Laryngeal resistance was
decreased for a majority of participants for straw phonation
and tongue trill, but not lip trill.
Two of the six parameters, oral air pressure and %CQ, had no
consistent direction of change for any SOVT. Of note for the %

TABLE 4.
Mean Peak Air Pressure Values (cm H2O) Before and After Each SOVT and After a Rest Period at the End of the Protocol
(PostSOVT)
Straw Phonation
Subject
Singer
T1
T2
T3
S1
Group
Mean
SD

Lip Trill

Tongue Trill

PostSOVT

PreSOVT

Mean

Versus Pre

Mean

Versus Pre

Mean

Versus Pre

Mean

Versus Pre

6.46
14.52
8.88
9.00

9.50
11.00
12.17
6.88

3.04

3.52
3.29

2.12

9.18
11.46
9.15
8.31

2.72

3.06
0.27

0.69

8.88
13.85
8.56
8.00

2.42

0.67
0.32

1.00

8.13
14.00
7.58
7.23

1.67

0.52

1.30

1.77

9.72
3.41

9.89
2.28

0.17
3.50

9.53
1.35

0.19
2.39

10.00
2.71

0.29
1.57

9.25
3.20

0.47
1.48

160

Journal of Voice, Vol. 29, No. 2, 2015

TABLE 5.
Mean Air flow During Voicing (L/s)
Straw Phonation
Subject
Singer
T1
T2
T3
S1
Group
Mean
SD

Lip Trill

Tongue Trill

PostSOVT

PreSOVT

Mean

Versus Pre

Mean

Versus Pre

Mean

Versus Pre

Mean

Versus Pre

0.15
0.31

0.30

0.20
0.38

0.33

0.05
0.07

0.03

0.23
0.29

0.37

0.08

0.02

0.07

0.21
0.33

0.36

0.06
0.02

0.06

0.23
0.43

0.34

0.08
0.12

0.04

0.20
0.14

0.24
0.16

0.04
0.03

0.24
0.12

0.05
0.05

0.32
0.07

0.12
0.15

0.34
0.08

0.14
0.12

CQ was that two subjects (T3, S1) had a 10% or greater increase
after each SOVTs and the other two had no change on any of the
SOVTs.
Table 9 also includes the predominate direction of change for
each measure after completing all three SOVTs and 2 minutes
of vocal rest (ie, carryover). For the carryover condition, four
measures had a predominant direction of change. Mean expiratory flow on sustained vowel and mean air flow on /pa/ both
were increased, laryngeal resistance was decreased, and %CQ
increased for most of the participants in the carryover condition.
SPL, which was consistently increased immediately after each
SOVT, had no predominant direction of change in the carryover
condition.

DISCUSSION
The purpose of this study was to evaluate aerodynamic and
EGG closed quotient changes immediately after completion
of each of three SOVTs relative to preSOVT values. Singers
were chosen as participants because use of SOVTs is routinely
incorporated into their training. However, very little empirical
data on changes to voice measures extending beyond the time
of execution of the SOVT are available. Likewise, there is
limited data evaluating differences in laryngeal function within
the same speaker across different SOVTs. If SOVTs are to be
used as a training tool or therapeutic approach it is of importance to determine not only whether laryngeal function can

be changed while engaged in the SOVT, but also whether any

changes that are induced persist into phonation trials not specifically involving semiocclusion. The primary findings from this
study were: (1) aerodynamic and EGG measures of voice generally did change immediately after completing each of three
SOVTs, (2) there was variation in the magnitude, direction,
and consistency of change on a given variable across the four
singers.
Before describing the overall patterns of change as a function
of the SOVTs, a brief comment regarding the baseline values
for each measure is offered here to help place the findings in
context with what has been reported in the literature. Zraick,
Smith-Olinda, and Shottshave26 reported normative values for
aged 18 to 86 years using the same equipment and software
as what was used here. The Zraick et al values in their youngest
age group (1839 year-olds) were used for comparison. The
preSOVT SPL values of the four subjects in the present study
were notably higher (group mean SPL of 88) than the Zraick
et al normative values (SPL of 7677 depending on gender).
This is consistent with the investigators informal observation
that the singers tended to produce a stage voice during sustained vowel production and syllable series even with instructions for comfortable pitch and loudness. Consistent with
the stronger stimulus production was a higher mean peak air
pressure value in the four subjects (group mean of 9.7 cm
H2O vs normative values of 56 cm H2O). Likewise, mean
air flow during voicing for the /pa/ and mean air flow on

TABLE 6.
Laryngeal Resistance (cm H2O/L/s)
Straw Phonation
Subject
Singer
T1
T2
T3
S1
Group
Mean
SD

PreSOVT

Mean

41
46

31

45
28

20

39.33
7.64

31
12.77

Versus Pre
4

18

11

8.33
11.24

Lip Trill
Mean

Versus Pre

Tongue Trill
Mean

Versus Pre

PostSOVT
Mean

39
46

22

2
0

9

42
41

22

1

5

9

35
33

21

35.67
12.34

3.67
4.73

35
11.27

4.33
5.03

29.67
7.57

Versus Pre

6

13

10

9.67
3.51

Troy Clifford Dargin and Jeff Searl

161

SOVT Exercises in Singers

TABLE 7.
Mean Expiratory Air flow (L/s) During Sustained Vowel
Straw Phonation

Lip Trill

Tongue Trill

PostSOVT

Subject

PreSOVT

Mean

Versus Pre

Mean

Versus Pre

Mean

Versus Pre

Mean

Versus Pre

Singer
T1
T2
T3
S1
Mean
SD

0.16
0.23

0.31
0.19
0.11

0.19
0.43

0.37
0.32
0.10

0.03
0.20

0.06
0.14
0.11

0.18
0.41

0.38
0.34
0.11

0.02
0.18

0.07
0.15
0.14

0.17
0.49

0.37
0.36
0.14

0.01
0.26

0.06
0.18
0.17

0.20
0.31

0.33
0.31
0.09

0.04
0.08

0.02
0.13
0.16

sustained vowel productions were both at or above the normative values. Laryngeal resistance values in our four subjects
(group mean 39 cm H2O/L/s) were lower than those in the
normative sample (68 and 79 cm H2O/L/s for females and
males, respectively). Overall, compared with young adult nonsingers in the Zraick et al study, the four singers produced samples with higher SPL, oral air pressure, and air flow rates, and
lower laryngeal resistance. It is possible that the patterns of
change after SOVT execution described in the following are
unique to those with singing training. It should be considered
that in some respects the aerodynamic differences that our
singers demonstrated in the baseline condition compared with
the norms suggest a somewhat optimized phonation situation
with perhaps restricted room to show change on the measures
that were evaluated.

Changes in aerodynamic and electroglottographic

measures
A function of the specific semi-occluded vocal tract.
There were some consistent changes that occurred for all three
SOVT exercises. In particular, most of the participants had at
least a 10% increase detectable right after each SOVT exercise
completion for SPL, air flow on sustained vowel, and air flow
during /pa/. Although not explicitly stated, a common assumption of teachers of singing and SLPs seems to be that various
SOVTs are likely to induce similar changes in laryngeal function. In this study, it appeared that SPL, and air flow measures

did, in fact, tend to increase regardless of which SOVT was being done. For two other measures, oral pressure (as an estimate
of subglottal pressure) and %CQ, there was no predominant
pattern of change occurring across the three SOVTs. Guzman27 found that subglottal air pressure increased during straw
phonation and remained higher for a short time afterward. In
the present study, there was not a consistent direction or
magnitude of change in the air pressure measure. The %CQ deserves further mention, however. Although only half of the participants had a change for each of the SOVTs, it was always the
same two subjects who showed the increase whereas the other
two remained unchanged. In this case, it seems that the impact
of the SOVT is dependent more on the individual rather than
the specific SOVT. Cordiero et al,7 found significant differences with higher mean %CQ during the lip and tongue trill
as compared with the nonoccluded /3/ at high intensities only.
Overall, the results support a tentative conclusion that SPL
and air flow measures are likely to increase for singers immediately after completing any of the three SOVTs trialed here.
Additionally, some individuals consistently have an increase
in %CQ after doing any of the SOVTs, but others may consistently not show a change in this measure. None of these SOVTs
induced a consistent change in oral air pressure.
Recall that participants also were assessed at the end of the
full protocol after resting the voice for a brief period of time.
This final measurement was an attempt to see if changes in
aerodynamic or EGG measures persisted beyond the immediate
postexercise time (ie, did the effects carry over for a brief period

TABLE 8.
Percent Closed Quotient (%CQ)
Straw Phonation
Subject
Singer
T1
T2
T3
S1
Group
Mean
SD

PreSOVT

Mean

Versus Pre

Lip Trill
Mean

Versus Pre

Tongue Trill
Mean

Versus Pre

PostSOVT
Mean

Versus Pre

48
40
41
27

50
42
47
31

2
2
6
4

51
42
48
30

3
2
7
3

52
41
47
31

4
1
6
4

53
40
48
31

5
0
7
4

39.0
8.76

42.5
8.35

3.5
1.91

42.75
9.29

3.75
2.22

42.75
9.03

3.75
2.06

43.0
9.63

4.0
2.94

162

TABLE 9.
Predominant Direction of Change for Each Measure Across the Three SOVT Conditions and the Carryover Condition
Straw Phonation
Data
Considered
From N
Participants

Parameter

Lip Trill

Tongue Trill

Carry-Over Post All SOVTs

Predominant
Outcome
Versus
PreSOVT

# Participants
Showing
Predominant
Outcome

Predominant
Outcome
Versus
PreSOVT

# Participants
Showing
Predominant
Outcome

Predominant
Outcome
Versus
PreSOVT

# Participants
Showing
Predominant
Outcome

Predominant
Outcome
Versus
PreSOVT

# Participants
Showing
Predominant
Outcome
2>
2
3/3

SPL mean

>

3/4

>

3/4

>

3/4

None

Expiratory
flow
Oral pressure

>

3/3

>

2/3

>

2/3

>

None

2>
2<

None

None

3/3

>

>

2
1>
1<
2/3

None

>

2
1>
1<
2/3

>

1
1>
2<
3/3

<

2/3

3/3

<

2/3

<

2/3

None

2
2>

None

2
2>

None

2
2>

>

3/4

3
4

Mean flow
during
voicing
Laryngeal
resistance
% Closed
quotient

Journal of Voice, Vol. 29, No. 2, 2015

Abbreviations: , less than a 10% change; None, there was no consistent outcome across subjects.
Notes: To be considered a change, the measure had to increase or decrease by  10% versus the preSOVT value.

Troy Clifford Dargin and Jeff Searl

SOVT Exercises in Singers

of time?). This carryover time period could also be viewed as a

means of potentially assessing the cumulative effect of
completing all three SOVTs. Using the 10% change criterion,
four of the six aerodynamic and EGG measures demonstrated
a change in the carryover condition relative to the preSOVT
data. Mean air flow during sustained vowel, mean air flow during /pa/, and laryngeal resistance had a predominant direction
of change in the carryover condition that matched the direction
of change seen immediately after completing most of the individual SOVTs. Both flow measures were increased and resistance was decreased for a majority of the participants in the
carryover condition.
The %CQ was increased in the carryover condition for three
of four participants. This is in contrast to %CQ not showing a
majority direction of change after completing any of the individual SOVTs. Perhaps there was a summative effect of
completing the three SOVTs in combination that ultimately
resulted in a change in the carryover condition for either T1
or T2 (recall that T3 and S1 consistently showed a %CQ change
in all conditions).
Also of note was that SPL in the carryover condition was
increased for two participants but unchanged for the other
two. This is in contrast to a consistent pattern of change in
SPL immediately after completing each of the individual
SOVTs. Oral air pressure did not show a consistent pattern of
change in the carryover condition matching what occurred in
the individual SOVT recordings.
A function of individual participant. If one considers
changes in the six measures within a given participant across
all three of the SOVTs it is clear that each participant demonstrated their own set of outcomes from the semiocclusion exercises. T1 had increases more than the preSOVT baseline for
SPL, oral pressure, and mean air flow during voicing on /pa/
after completing each of the three SOVTs. T2 demonstrated
the same direction of change after all three SOVTs for SPL
and mean expiratory air flow during the sustained vowel. S1
had four measures show the same direction of change after
each SOVT such as mean expiratory flow on sustained vowel
(>), mean air flow during voicing on /pa/ (>), laryngeal resistance (<), and %CQ (>). Results for T3 cannot be described
confidently in this manner because three of the six measures
are not reported because of suspected measurement error.
Based on these results a tentative conclusion is that SOVTs
may have a relatively consistent immediate impact on laryngeal function for an individual regardless of the SOVT that
is completed. This is congruent with Titze15 and Nix4 hypotheses that SOVTs help vocal economy. However, the SOVT
impact on aerodynamic and EGG measures is likely to vary
across individuals.

CONCLUSIONS AND LIMITATIONS

Overall, completion of the SOVTs did result in aerodynamic
and EGG differences immediately after the exercises although
there was significant variability in the specific parameters that
changed across participants.

163

There are several limitations to this study. The small number of participants precluded generalization of results. The
types and magnitude of changes in the voice varied substantially after the SOVTs, indicating that very large groups of
participants will be needed to better define what changes
should be anticipated from these exercises. However, within
this small sample it is noted that significant individual variation in aerodynamic and EGG parameters was noted. The
period of SOVT execution was brief and may not reflect durations of exercise activity that an SLP or singing teacher may
request. However, even with just brief periods of SOVT
execution, aerodynamic and EGG differences did occur. It is
possible that longer periods of SOVT execution within the
setting or over an extended number of days or weeks may
result in more consistent, impressive, or different changes
than what were found here. Related to this issue is the fact
that persistent use of the SOVTs, as opposed to one-time
use in this study, may be necessary for adjustments to occur
or to stabilize. It would be of interest to compare baseline
to postSOVT exercises with participants completing a more
rigorous SOVT schedule over an extended time frame.
REFERENCES
1. Titze IR, Laukkanen A-M. Can vocal economy in phonation be increased
with an artificially lengthened vocal tract? A computer modeling study.
Logoped Phoniatr Vocol. 2007;32:147156.
2. Titze I, Finnegan E, Laukkanen A, Jaiswal S. Raising lung pressure and
pitch in vocal warm-ups: the use of flow-resistant straws. J Sing. 2002;
58:329338.
3. Titze IR. Phonation threshold pressure measurement with a semi-occluded
vocal tract. J Speech Lang Hear Res. 2009;52:10621072.
4. Nix J. Lip trills and raspberries: high spit factor: alternatives to the nasal
continuant consonants. J Sing. 1999;55:1519.
5. Miller R. Sotto voce: what does humming accomplish? J Sing. 1996;52:
4950.
6. Bele IV. Artificially lengthened and constricted vocal tract in vocal training
methods. Logoped Phoniatr Vocol. 2005;30:3440.
7. Cordeiro GF, Montagnoli AN, Nemr NK, Menezes MHM, Tsuji DH.
Comparative analysis of the closed quotient for lip and tongue rills in relation to the sustained vowel //. J Voice. 2012;26:e17e22.
8. Milbrath RL, Solomon NP. Do vocal warm-up exercises alleviate vocal fatigue? J Speech Lang Hear Res. 2003;46:422436.
9. Motel T, Fisher KV, Leydon C. Vocal warm-up increases phonation
threshold pressure in soprano singers at high pitch. J Voice. 2003;17:
160167.
10. Schwarz K, Cielo CA. Modificac~oes larngeas e vocais produzidas
pela tecnica de vibrac~ao sonorizada de lngua. Pro Fono. 2009;21:
161166.
11. Gaskill CS, Quinney DM. The effect of resonance tubes on glottal contact
quotient with and without task instruction: a comparison of trained and untrained voices. J Voice. 2012;26:e79e93.
12. Hamdan AL, Nassar J, Al Zaghal Z, El-Khoury E, Bsat M, Tabri D. Glottal
contact quotient in Mediterranean tongue trill. J Voice. 2012;26:669.
e11669.e15.
13. Garrel R, Scherer R, Nicollas R, Giovanni A, Ouaknine M. Using the relaxation oscillations principle for simple phonation modeling. J Voice. 2008;
22:385398.
14. Titze IR. Acoustic interpretation of resonant voice. J Voice. 2001;15:
519528.
15. Titze IR. Voice training and therapy with a semi-occluded vocal tract: rationale and scientific underpinnings. J Speech Lang Hear Res. 2006;49:448459.
16. Lewis C. Dr. Marissa Silverman Music Reference and Research E85. 2021
2007 The Dynamics of Voice Studios at New York University.

164
17. Titze IR. How to use the flow-resistant straws. J Sing. 2002;58:
429430.
18. Barrichelo-Lindstr
om V, Behlau M. Resonant voice in acting students:
perceptual and acoustic correlates of the trained Y-Buzz by Lessac. J Voice.
2009;23:603609.
19. Gartner-Schmidt J. Voice therapy for the treatment of voice disorders.
20. Verdolini-Marston K, Katherine Burke M, Lessac A, Glaze L, Caldwell E.
Preliminary study of two methods of treatment for laryngeal nodules. J
Voice. 1995;9:7485.
21. Schindler A, Bottero A, Capaccio P, Ginocchio D, Adorni F, Ottaviani F.
Vocal improvement after voice therapy in unilateral vocal fold paralysis.
J Voice. 2008;22:113118.
22. Chen SH, Hsiao T-Y, Hsiao L-C, Chung Y-M, Chiang S-C. Outcome of
resonant voice therapy for female teachers with voice disorders: perceptual,

Journal of Voice, Vol. 29, No. 2, 2015

23.
24.
25.
26.

27.

physiological, acoustic, aerodynamic, and functional measurements. J

Voice. 2007;21:415425.
Ramig LO, Verdolini K. Treatment efficacy: voice disorders. J Speech Lang
Hear Res. 1998;41:S101S116.
Verdolini K, Druker DG, Palmer PM, Samawi H. Laryngeal adduction in
resonant voice. J Voice. 1998;12:315327.
Yiu EM-L, Chen FC, Lo G, Pang G. Vibratory and perceptual measurement
of resonant voice. J Voice. 2012;26:675.e13675.e19.
Zraick RI, Smith-Olinde L, Shotts LL. Adult Normative Data for the KayPENTAX Phonatory Aerodynamic System Model 6600. J Voice. 2012;26:
164176.

Guzman M, Laukkanen A-M, Krupa P, Horacek J, Svec
JG, Geneid A.
Vocal tract and glottal function during and after vocal exercising with resonance tube and straw. J Voice. 2013;27:523.e19523.e34.