Influence of Emotional Expression, Loudness,

and Gender on the Acoustic Parameters of Vibrato

in Classical Singers
*,†Marco A. Guzman, ‡Jayme Dowdall, *,§Adam D. Rubin, kAhmed Maki, {Samuel Levin, ‡Ross Mayerhoff,
and *,‡Maria Cristina Jackson-Menaldi, *St. Clair Shores, z{Detroit, xAnn Arbor, and kEast Lansing, Michigan and ySantiago, Chile

Summary: Objectives. Vibrato is considered one of the most essential characteristics of the classical singing voice.
Vibrato can be acoustically described by the rate, extent, onset, and regularity. The aim of this study was to determine the
influence of emotional expression on acoustic parameters of vibrato in classically trained singers.
Study Design. A prospective cohort study was performed.
Methods. Thirty healthy classical singers were recruited for this study, 29 singers met inclusion criteria. Inclusion cri-
teria for this study were as follow: 1) no history of vocal pathology in the past year, 2) to have at least 5 years of classical
singing training. Each subject was asked to sing the phrase ‘‘I Love You,’’ while expressing four different emotions (ten-
derness, anger, happiness, and sadness) and without emotion (neutral state). The musical tonality of the phrase was adapt-
ed to each singer’s vocal classification. Subjects were also recorded at three levels of loudness (pianissimo, mezzo forte,
and fortissimo), while expressing each emotion. Acoustical analysis was performed during the vowel /o/ of the word
‘‘Love’’ to determine rate of vibrato, the extent of vibrato, and vibrato jitter.
Results. Vibrato parameters did not vary significantly when different emotions were expressed. However, vibrato jit-
ter and extent did vary significantly between different levels of loudness. Significant differences were also noted in both
rate and extent of vibrato when compared between sexes.
Conclusions. Expression of emotions does not affect the acoustic parameters of vibrato, although some parameters
are affected by loudness and gender.
Key Words: Vibrato–Emotion–Singing voice–Acoustical analysis.

INTRODUCTION characteristics.4–10 These physiological effects are reflected in

The same phrase may sound different to the listener when it is
sung or spoken using different emotions. The information
coded in both singing and spoken language is not only coded
in the words themselves, but also in a nonverbal component ex-
pressing the real or feigned emotional state of the speaker or
singer. In the phonetic sciences, the vocal expression of emotion
has been studied for quite some time, and the vocal parameters
of emotions are now relatively well understood.1 Studies dem-
onstrate that listeners are able to reliably and accurately per-
ceive emotions in human voices.2 Pittam and Scherer,2 and
Tischer3 noted considerable consistency regarding the acoustic
parameters differentiating anger, happiness, sadness, and fear.
Specific associations between different emotions and measures
of pitch, voice intensity level, and speech rate have also been
elucidated.
Emotion produces involuntary physiological alterations in
the autonomic and somatic nervous systems that change vocal
Accepted for publication February 22, 2012.
Presented at the Voice Foundation’s 40th Annual Symposium, June 6, 2011; Philadel-
phia, PA.
From the *Lakeshore Professional Voice Center, Lakeshore Ear, Nose, and Throat Cen-
ter, St. Clair Shores, Michigan; ySchool of Communication Sciences, University of Chile,
Santiago, Chile; zDepartment of Otolaryngology, School of Medicine, Wayne State Uni-
versity, Detroit, Michigan; xDepartment of Otolaryngology-HNS, University of Michigan
Medical Center, Ann Arbor, Michigan; kPOH Regional Medical Center, Michigan State
University, East Lansing, Michigan; and the {School of Medicine, Wayne State University,
Detroit, Michigan.
Address correspondence and reprint requests to Marco A. Guzman, Faculty of Medicine,
School of Communication Sciences, University of Chile, Avenida Independencia 1027,
Santiago, Chile. E-mail: guzmanvoz@gmail.com
Journal of Voice, Vol. 26, No. 5, pp. 675.e5-675.e11
0892-1997/$36.00
Ó 2012 The Voice Foundation
doi:10.1016/j.jvoice.2012.02.006
muscle tone of the respiratory system, vocal tract, and vocal
folds leading to perceived changes in vocal quality. According
to Scherer’s theory, stimuli are evaluated according to function-
ally defined criteria, such as ‘‘novelty,’’ ‘‘need,’’ and ‘‘coping
potential.’’ The net result of the outcomes of all evaluation
checks affects the nervous system and, in turn, the physiological
consequences of the nervous system’s response define the
changes in voice characteristics that carry the emotional infor-
mation. For example, unpleasant stimuli cause ‘‘faucal and pha-
ryngeal constriction and tensing as well as shortening of the
vocal tract,’’ leading to stronger high-frequency resonances,
arise in the first formant, a fall in the second formant, narrow
formant bandwidths, etc.10
To date, the vast majority of studies on emotional influences
on human voice have focused on spoken word. Few studies ad-
dress the singing voice; some of which have reported the effect
of different emotions in acoustic characteristics of singing
voice quality,11–16 whereas other studies note changes in
breathing pattern changes during singing.17–20
The listening audience may be impressed by the beauty of
a singer’s voice, but may also be impressed by the manner in
which the singer conveys the emotional content of the song
lyrics or opera aria. Different emotional states are expressed
through changes in several acoustical parameters of voice,
such as fundamental frequency (F0), intensity level, formant
frequencies, and spectrum. Based on the assumption that vi-
brato is a modulation of some of these vocal parameters, one
may speculate that the singer’s vibrato could be affected by
emotional expression.
Professional opera singers in the Western classical tradition
usually sing with vibrato. There is general agreement in the
675.e6
FIGURE 1. Phrase ‘‘I love you.’’
singing community that vibrato occurs naturally when the voice
is produced with freedom and good technique.21–23 Profes-
sional singers appear to ‘‘acquire’’ vibrato without actively
striving to obtain vibrato. Vibrato seems to develop as voice
training proceeds successfully.24
The study of vocal vibrato among classical singers has re-
ceived attention from researchers in several fields, including
physiology, acoustics, voice pedagogy, and speech-language
pathology. In 1932, Seashore23 provided one of the earliest
comprehensive descriptions of vocal vibrato. His systematic re-
search is cited to this day, as is his definition: ‘‘A good vibrato is
a pulsation of pitch, usually accompanied with synchronous
pulsations of loudness and timbre, of such extent and rate as
to give a pleasing flexibility, tenderness, and richness to the
tones.’’
The acoustical properties of vibrato are characterized primar-
ily by the rate, the extent, onset, and regularity of the modula-
tions in the frequency and amplitude of the voice. All these
characteristics can be assumed to be perceptually relevant and
are generally considered to be indicative of a singer’s vocal
skills.25
The rate of vibrato describes how rapidly the modulations oc-
cur, and is expressed in hertz. According to Sundberg,25 typical
values range from 5 to 7 Hz. A rate slower than 5 undulations/s
tends to sounds unacceptably slow, and vibrato rates exceeding
8 undulations/s tend to sound nervous.25 The rate of vibrato for
a given individual singer seems to depend on a number of fac-
tors, such as gender and age.25 It is also suggested to depend on
the emotional involvement of the singer.26 Female singers tend
FIGURE 2. Spectrogram showing the highest harmonic available that shows a clear and strong vibrato trace without any overlap from adjacent
harmonics.
Journal of Voice, Vol. 26, No. 5, 2012
to have a slightly faster mean vibrato rate than male singers.
Shipp et al26 reported that the average range for vibrato was
5.4 undulations/s for male singers and 5.9 undulations/s for fe-
male singers during sustained vowels.
The extent of frequency modulation is measured in two ways:
the semitone (ST), equal to a musical half note, and the cent,
with 100 cents being equal to one ST (100 cents ¼ 1 ST). The
extent of frequency modulation describes how large or small
the modulations are relative to average frequency or amplitude
during the vibrato cycle.25 The accepted range for extent of vi-
brato is a third of a tone to a ST,19 although Vennard27 in 1967
stated that it may be up to a tone. Seashore28 considered optimal
extent to fall within the range of 30–70 cents, with 50 cents (50
cents ¼ 0.25 tone or 0.50 ST) being ideal. Extent of vibrato
among classically trained singers in another study averaged
1.20 ST.29 Current optimal ranges have been reported as
0.50–2.00 ST30 or 50–150 cents depending on pitch and
loudness.2,31
Frequency modulation of F0 characterized by its rate and
extent is the more basic feature of vibrato production. This
modulation, in turn, causes amplitude modulations of the in-
dividual spectrum partials,30,32 which result in a modulation
of the overall amplitude. This amplitude modulation may be
both in phase and out of phase with the original frequency
modulation.
Regularity refers to the steadiness or repeatability of the
modulations.25 According to Robison et al,33 regularity is
considered to best reflect the skill of the singer; the smoother
the vibrato, the more beautiful the singer’s voice is perceived
to be. It was reported that regularity of vibrato is the principal
element influencing the perception of the quality of vibrato.34
The authors found there is a high negative correlation between
vibrato jitter (measurement of regularity) and the vibrato
quality.
The aim of this study was to determine the influence of
emotional expression, loudness, and gender in acoustic pa-
rameters of vibrato in classically trained singers. We hypoth-
esize that emotion will influence the rate, extent, and
regularity.
Marco A. Guzman, et al Influence of Emotional Expression on Acoustic Parameters of Vibrato
FIGURE 3. Isolated highest harmonic available to get values of rate, extent, and vibrato jitter.
METHODS
Participants
This study was reviewed and approved by the St. John Hospital
Institutional Review Board. Informed consent was obtained
from 29 adult classical singers (17 women and 12 men). The av-
erage age of this subject set was 38 years, with a range of 23–62
years old. The average age for female singers was 36 years and
41 years for male singers. Regarding voice classification, nine
singers were sopranos, 11 mezzo-sopranos, one alto, three
tenors, three baritones, and two basses. Inclusion criteria for
this study included 1) no history of vocal pathology in the past
year and 2) at least 5 years of classical singing training. Although
30 singers were recruited, one singer did not meet the inclusion
criteria. The average length of voice training of our subjects was
9.2 years, with a range of 5–20 years. Participants were recruited
from three different professional choirs and all of them were
asked to undergo both flexible and rigid laryngoscopy videostro-
boscopy to corroborate absence of voice pathology.
Voice recordings
Subjects were allowed to warm up the voice for 10 minutes be-
fore voice assessment. Warm-up exercises were freely chosen by
each participant. One of the authors (MAG) provided individual
demonstrations and verbal descriptions of the required singing
tasks. Each complete recording session was accomplished in ap-
proximately 18 minutes with the following protocol. Each sub-
ject was asked to sing and interpret the phrase ‘‘I Love you’’
(Figure 1) expressing five different basic emotions: tenderness,
FIGURE 4. Vibrato rate (mean ± standard error) for the five emo-
tional states was unchanged by emotion (ANOVA, P ¼ 0.941; n ¼ 86
per emotion).
675.e7
anger, happiness, sadness, and neutral, or no emotion (neutral).
The musical tonality of the phrase was adapted to each singer’s
vocal classification (baritones and basses F3 (174 Hz.); tenors
A3 (220 Hz.); alto and mezzo-sopranos F4 (349 Hz.); and sopra-
nos A4 (440 Hz.). Musical tonalities for each vocal classifica-
tion were arbitrarily chosen. The criteria were chosen so that
the highest and longest note of the musical phrase (vowel /o/)
would be at the beginning of the upper third of the tessitura.
This note was C for low voices and E for high voices. Partici-
pants sang in three different loudness levels for each emotion:
pianissimo, mezzo forte, and fortissimo. Although loudness
level was subjectively controlled by singers and one of the au-
thors (MAG), they were strongly encouraged to make a clear dif-
ference between the three different loudness levels. Objective
control of this parameter was not used because this might inter-
fere negatively in the expression of emotion during singing.
Acoustic output was measured at a constant microphone-to-
mouth distance of 10 cm. using a professional condenser mi-
crophone (Perception 120) connected to the DAT recorder
(Marantz PMD 671) in IAC sound suite. Samples were re-
corded digitally at a sampling rate of 48.000 Hz with 16 bits
per sample quantization.
Samples were edited with the software Goldwave, version
V5.57 (GoldWave Inc., St. John’s, Newfoundland, Canada)
and acoustical analysis of vibrato parameters was performed
using Voce Vista software (developed by Miller et al35 from
the Voice Research Lab at the University of Groningen,
EGGs for Singers, Groningen, The Netherlands).
Computer analysis was performed during the vowel /o/ of the
word ‘‘Love’’ because this is the longest note in the selected
FIGURE 5. Vibrato jitter for the five emotional states also was un-
changed by emotion (P ¼ 0.267).
675.e8
FIGURE 6. Vibrato extent was unchanged during these five emo-
tional states (P ¼ 0.591).
musical phrase. The duration of the vowel /o/ was 4 seconds,
which is enough time to perform the analysis of acoustical pa-
rameters of vibrato (rate, extent, and regularity). Regularity was
obtained using vibrato jitter measurement.
An area of 3 seconds was selected from the middle part of the
edited vowel /o/ samples. The highest harmonic available that
shows a clear and strong vibrato trace without any overlap
from adjacent harmonics was chosen to analyze (Figure 2).
Twelve vibrato cycles from each sample were used to get values
of rate, extent, and vibrato jitter (Figure 3). According to Miller
et al35 (Voce Vista developers), a minimum of three or four vi-
brato cycles are required to measure vibrato rate and extent, and
at least 10 vibrato cycles are required to measure vibrato jitter
accurately.
Descriptive statistics were obtained for all variables. Data
were processed using SPSS version 12.0 for Windows (IBM
Corp., Armonk, NY). The strength of the association between
the five different emotions: tenderness, anger, happiness, sad-
ness, and neutral and their impact on the rate of vibrato, extent
of vibrato, and regularity of vibrato (vibrato jitter) were calcu-
lated. Neutral emotion was collected as a reference. The possible
effect of gender and intensity level on acoustic parameters of
vibrato was also examined. Although a multivariate analysis of
covariance was attempted, some variables were not normally dis-
tributed, invalidating the underlying assumptions of this statistic.
Analysis of variance (ANOVA) with post hoc analysis to deter-
mine differences between variables (such as loudness), or the
nonparametric Kruskal-Wallis tests were used, as appropriate.
RESULTS
The impact of emotions on vibrato
A 3-second snap shot was recorded during the sustained vowel
as described in the Methods section. This sustained vowel was
TABLE 1.
Vibrato Characteristics Did Not Vary by Emotion
Vibrato Characteristic Neutral Tenderness
Rate (Hz) 5.53 ± 0.85 5.57 ± 0.88
Jitter (%) 4.7 ± 3.9 4.3 ± 2.2
Extent (cents) 70.3 ± 28.0 73.2 ± 27.1
Notes: None of the rate (ANOVA, P ¼ 0.941), jitter (P ¼ 0.267), or extent (P ¼ 0.591) were altered by emotion. Values are expressed as mean ± SD.
Journal of Voice, Vol. 26, No. 5, 2012
compared with the expression of no emotion or ‘‘neutral’’ in
addition to the emotions of tenderness, anger, happiness, and
sadness. Each recording was examined for differences with re-
spect to the vibrato parameters. No significant differences were
noted with respect to emotion in any parameter of vibrato
(Figures 4–6). This is best depicted in Table 1.
The impact of loudness on vibrato
Three seconds of the same phrase were sung at three levels of
intensity to assess the impact of loudness on vibrato. Partici-
pants sang at three levels of loudness, which included pianis-
simo, mezzo forte, and fortissimo. A significant difference
(P ¼ 0.024) was noted in vibrato jitter and extent (P ¼ 0.019)
with respect to loudness. There was no significant difference
noted with regard to rate of vibrato with varied loudness
(Figures 7 and 8, Table 2).
The impact of gender of vibrato
Lastly, gender was noted to significantly impact the rate and ex-
tent of vibrato (Figures 9 and 10, Table 3), whereas no gender
differences were noted with regard to jitter.
DISCUSSION
The present study was designed to determine the influence of
emotional expression, loudness, and gender on acoustic param-
eters of vibrato in classically trained singers.
None of the vibrato parameters appear to be affected by the
expression of emotion, although some were affected by loud-
ness and gender.
Contrary to our data, other authors have proposed a relation-
ship between emotional expression and acoustic parameters of
vibrato. Shipp et al26 speculated that the emotional involvement
of the singer may increase the vibrato rate. Later, Sundberg36
found further support for this speculation by comparing sung
sequences of sustained vowels (neutral) to the singing of songs
(where the singer expresses emotion). The mean vibrato rate for
the sustained tones averaged to 5.4 Hz, whereas that from the
real performance of a song averaged to 6.2 Hz or about 15%
higher.
Vibrato extent has also been related to emotional expression.
Sundberg36 analyzed one professional singer singing a set of
music excerpts from the Lieder and opera repertoires in two de-
liberately contrasting fashions, as in a concert situation, and in
an emotionally neutral way, as if the singer did not engage emo-
tionally in the poem. The author found that in agitated excerpts,
Emotion (n ¼ 86)
Happiness Sadness Anger
5.62 ± 0.93 5.61 ± 0.90 5.64 ± 0.93
4.7 ± 2.9 5.0 ± 3.4 4.9 ± 2.8
79.2 ± 31.1 72.0 ± 30.5 77.0 ± 32.1
Marco A. Guzman, et al Influence of Emotional Expression on Acoustic Parameters of Vibrato
FIGURE 7. Pianissimo is represented by the lightest bar, mezzo
forte in the moderately shaded bar, and fortissimo in the darkest bar.
Vibrato jitter (mean ± standard error), shown in the left panel, differed
across the loudness levels (ANOVA, *P ¼ 0.024; n ¼ 145). In post hoc
testing, this difference was because of the comparison of pianissimo
and fortissimo (P ¼ 0.024). Right panel: Vibrato rate was unchanged
across these three levels of loudness (P ¼ 0.959). Values obtained at
mezzo forte were not different from pianissimo or fortissimo for either
jitter or rate.
the extent of the vibrato undulations was mostly considerably
greater in the expressive than in the neutral versions, whereas
in the nonagitated excerpts the two versions were more similar.
This could suggest that the vibrato extent is used for expressive
purposes in singing. Scherer37 notes that anger and joy are as-
sociated with a high F0 variability, hypothesizing that F0 vari-
ability in speech is translated into vibrato extent in singing.
In our study, regularity (vibrato jitter) and extent of vibrato
were affected by changes in loudness. The rate of vibrato re-
mained the same. Specifically, we found that extent increased
and vibrato jitter decreased (more regularity) with increased
loudness. Others have demonstrated a similar effect of loudness
on extent. According to Winckel38 the extent of the vibrato un-
dulations varies with loudness of phonation. This author pub-
lished data from a soprano who increased her vibrato extent
from 1 to 1.5 STs when she raised her vocal loudness by
15 dB, keeping the pitch constant. Corroborating this result,
Bretos and Sundberg39 used recordings of performances by op-
era singers in a section of the aria with a crescendo in two sus-
tained pitches. They found that the extent of vibrato varied
systematically and tended to increase over time in both of the
tones that were analyzed. Thus, the vibrato extent was wider
with the increase in the sound level. The increase varied be-
TABLE 2.
The Three Loudness Levels Did Not Affect Vibrato Characteristics Equally
Vibrato Characteristic Pianissimo
Rate (Hz) 5.61 ± 0.90
Jitter* (%) 5.2 ± 3.1
Extenty (Cents) 71.0 ± 25.5
Note: Values are expressed as mean ± SD.
* In post hoc testing, pianissimo differed from fortissimo (P ¼ 0.024).
y
In post hoc testing, pianissimo differed from fortissimo (P ¼ 0.036).
675.e9
FIGURE 8. Vibrato extent differed for the loudness levels (ANOVA,
P ¼ 0.019). In post hoc testing, this difference again was because of
a difference at pianissimo and fortissimo loudness levels
(P ¼ 0.036). Mezzo forte was not different from either pianissimo or
fortissimo.
tween 17 and 80 cents and in many cases it could be approxi-
mated by a linear function of time with a quite high
correlation. In another study using dynamic changes, Michel
and Myers40 tried to verify this dependence of extent on loud-
ness, by measurements on singing students who sang crescen-
dos and decrescendos at low, medium, and high pitches. At
high pitches, most subjects exhibited a similar effect; during
a crescendo, the vibrato extent increased slightly with vocal
loudness, from 0.6 or 0.7 STs in pianissimo to about 1.0 ST
in fortissimo.
When reviewing the literature, we did not find any studies ad-
dressing the relationship between loudness and regularity of vi-
brato. The results of our study showed that there is a significant
decrease of vibrato jitter value with an increase in loudness. In
other words, the louder the voice the more regular the vibrato
modulations. This is not surprising as most singers would agree
it is more difficult to sing softly than loudly. It requires more
breath support and laryngeal muscle control. Regularity is con-
sidered to best reflect the skill of the singer; the more regular
and symmetric the vibrato, the more beautiful the singer’s voice
is perceived to be.33,34,41 Therefore, it is possible in subjects
with a higher degree of training (more skillful singers), the re-
lationship between regularity of vibrato and vocal loudness
may be different to reflect a higher degree of laryngeal control.
Orlikoff and Kahane,42 andJacob43 found that laryngeal jitter
tended to decrease with increasing vocal intensity. Although
Loudness (n ¼ 145)
Mezzo Forte Fortissimo P Value (ANOVA)
5.59 ± 0.91 5.58 ± 0.88 0.959
4.7 ± 3.5 4.3 ± 2.5 0.024
72.0 ± 29.5 80.0 ± 33.4 0.019
675.e10
FIGURE 9. Women are depicted in the dark bars (n ¼ 254) and men
in the light bars (n ¼ 180). As shown in the left panel, gender altered
vibrato rate (t test, P < 0.0005) but not jitter (right panel; P ¼ 0.865).
N refers to the number of measurements used for comparison, not
the number of vocalists.
these studies were performed with vocal jitter (not vibrato jit-
ter), the acoustic principle is the same and therefore both, vi-
brato jitter and vocal jitter could be compared. In addition, it
is important to note that in the present study none of the partic-
ipants produced a pressed voice during fortissimo or breathy
voice during pianissmo. Therefore, it could be possible to as-
sume that the regularity of the vibrato modulations is higher
(low vibrato jitter) when there is more vocal adduction without
resulting in a pressed voice and regularity is lower (high vibrato
jitter) when there is less vocal adduction without resulting in
a breathy voice.
Gender significantly affected vibrato rate and extent in this
study. Female singers had higher rate values and greater extent.
Shipp et al26 reported that female singers tend to have a slightly
faster mean vibrato rate than male singers. Five male and five
female professional opera singers were analyzed. All subjects
went through the same protocol singing sustained vowels.
The average rate of vibrato was 5.4 Hz for male singers and
5.9 Hz for female singers. In our study, the average rate of vi-
brato was 5.2 Hz. for male singers and 5.8 Hz for female singers
(slightly larger difference). Similar results were found in a study
with nonprofessional choir singers.44 Vibrato rate was higher
for women than men, suggesting that this difference between
genders is not necessarily related to the level of vocal training.
In the same study with choir singers, the author found that the
extent is greater for female singers than for male singers.44
These results are consistent with our results, where the mean
FIGURE 10. Gender also altered vibrato extent (P < 0.0005).
Journal of Voice, Vol. 26, No. 5, 2012
TABLE 3.
Gender Altered Rate and Extent, But Not Jitter
Gender
Vibrato Male Female P Value
Characteristic (n ¼ 180) (n ¼ 254) (t Test)
Rate (Hz) 5.20 ± 0.74 5.87 ± 0.89 <0.0005
Jitter (%) 4.8 ± 2.5 4.7 ± 3.5 0.865
Extent (cents) 67.3 ± 26.2 79.3 ± 31.3 <0.0005
of vibrato extent was 67 cents for male and 79 cents for female
singers. This difference was statistically significant.
There are limitations to our study. Most importantly, we can
only test acted or feigned emotions. We depend on the acting
ability of our subjects to express each emotion. Furthermore,
the participants are asked to express the emotion through a sim-
ple musical phrase with the words ‘‘I Love You.’’ Perhaps, lon-
ger passages or known operatic arias would demonstrate more
variability in the investigated parameters. Having each sing the
same words and musical line, however, likely is more appropri-
ate for data analysis. In any case, this is the first study involving
a large sample size with multiple vocal classifications to look at
the effect of emotion on vibrato.
CONCLUSIONS
The expression of emotions does not appear to impact the
acoustic parameters of vibrato. Loudness effects vibrato extent
and jitter, whereas vibrato rate and extent are affected by
gender.
Acknowledgments
Special thanks to Dr Ruth Savoy-Moore (St. John Hospital and
Medical Center, Detroit, Michigan); Suzanne Acton (MOT
Opera House and Rackham Symphony Choir, Michigan);
Dr B. Augustus Hill (Brazeal Dennard Choral, Detroit, Mich-
igan); Dr B. Emery Stephens (Wayne State University, Detroit,
Michigan); and Prof. John-Paul White (Oakland University,
Michigan).
REFERENCES
1. Murray IR, Arnott JL. Toward a simulation of emotion in synthetic speech:
a review of the literature on human vocal emotion. J Acoust Soc Am. 1993;
93:1097–1108.
2. Pittam J, Scherer KR. Vocal expression and communication of emotion. In:
Lewis M, Haviland JM, eds. Handbook of Emotions. New York, NY: Guil-
ford Press; 1993:185–197.
3. Tischer B. The Vocal Communication of Feelings. Weinheim, Germany:
Beltz Press; 1994.
4. Grant KW, Walden BE. Spectral distribution of prosodic information.
J Speech Hear Res. 1996;39:228–238.
5. Johnstone T, Scherer KR. Vocal communication of emotion. In: Lewis ML,
Haviland-Jones JM, eds. Handbook of Emotions. New York, NY: The Guil-
ford Press; 2003.
6. Protopapas A, Lieberman P. Fundamental frequency of phonation and per-
ceived emotional stress. J Acoust Soc Am. 1997;101:2267–2277.
7. Schirmer A, Kotz SA. Beyond the right hemisphere: brain mechanisms
mediating vocal emotional processing. Trends Cogn Sci. 2006;19:24–30.
8. Tartler VC, Braun D. Hearing smiles and frowns in normal and whisper reg-
isters. J Acoust Soc Am. 1994;96:2101–2107.
Marco A. Guzman, et al Influence of Emotional Expression on Acoustic Parameters of Vibrato
9. Wilson D, Wharton T. Relevance and prosody. J Pragmatics. 2006;36:
1559–1579.
10. Scherer KR. Vocal affect expression: a review and a model for future re-
search. Psychol Bull. 1986;99:143–165.
11. Kotlyar GM, Morosov VP. Acoustical correlates of the emotional content of
vocalized speech. Sov Phys Acoust. 1976;22:208–211.
12. Siegwarth H, Scherer K. Acoustic concomitants of emotional expression in
operatic singing: the case of Lucia in Ardi gli incensi. J Voice. 1995;9:
249–260.
13. Johnstone T, Scherer K. Spectral measurement of voice quality in opera
singers: the case of Gruberova. In: Elenius K, Branderud P, eds. Proceed-
ings of the XIIIth International Congress of Phonetic Sciences. Vol 1.
Stockholm; 1995.
14. Kotlyar G, Morozov V. Acoustical correlates of the emotional content of
vocalised speech. Sov Phys Acoust. 1976;22:208–221.
15. Scherer K, Oschinsky J. Cue utilization in emotion attribution from audi-
tory stimuli. Motiv Emot. 1977;1:331–346.
16. Gabrielsson A. Emotional expression in music performance. The Stock-
holm Music Acoustics Conference, Stockholm; July 1993; Sweden.
17. Pettersen V, Bjørkøy K. Consequences from emotional stimulus on breath-
ing for singing. J Voice. 2009;23:295–303.
18. Foulds-Elliott S, Thorp CW, Cala S, Davis P. Respiratory function in oper-
atic singing: effects of emotional connection. Logoped Phoniatr Vocol.
2000;25:151–168.
19. Miller R. The Structure of Singing. New York, NY: Shirmer Books; 1986.
20. Chapman J. Singing and Teaching Singing: A Holistic Approach to Classi-
cal Voice. San Diego, CA: Plural; 2006.
21. Howes P, Callaghan J, Davis P, Kenny D, Thorpe W. The relationship be-
tween measured vibrato characteristics and perception in Western operatic
singing. J Voice. 2004;18:216–230.
22. Prame E. Vibrato extent and intonation in professional Western lyric sing-
ing. J Acoust Soc Am. 1997;102:616–621.
23. Seashore CE, ed. The Vibrato. University of Iowa Studies in the Psychology
of Music, 1932;Vol 1. Iowa City, IA: University of Iowa; 1932.
24. Bjorklund A. Analyses of soprano voices. J Acoust Soc Am. 1961;33:575–582.
25. Sundberg J. Acoustic and psychoacoustic aspects of vocal vibrato. In:
Dejonckere P, Hirano M, Sundberg J, eds. Vibrato. San Diego, CA: Singular
Publishing Company; 1995.
26. Shipp T, Leanderson R, Sundberg J. Some acoustic characteristics of vocal
vibrato. J Res Singing. 1980;4:18–25.
675.e11
27. Vennard W. Singing: The Mechanism and the Technic. 5th ed. New York,
NY: Carl Fischer; 1967.
28. Seashore C. Psychology of Music. New York, NY: McGraw-Hill; 1967.
29. Arroabarren I, Zivanovic M, Bretos J, Ezcurra A, Carlosena A. Measure-
ment of vibrato in lyric singers. Proceedings of the IEEE Instrumentation
and Measurement Technology Conference. Hungary; 2001.
30. Horii Y. Acoustic analysis of vocal vibrato: a theoretical interpretation of
data. J Voice. 1989;3:36–43.
31. Sundberg J. Where does the sound come from?. In: Potter A, ed. The Cam-
bridge Companion to Singing. Cambridge, UK: Cambridge University
Press; 2000.
32. Sundberg J. The Science of the Singing Voice. De Kalb, IL: Northern Illinois
University Press; 1987.
33. Robison CW, Bounous B, Bailey R. Vocal beauty: a study proposing its
acoustical definition and relevant causes in classical baritones and female
belt singers. J Singing. 1994;51:19–30.
34. Cecconello L, Guzm
an M. Correlaci
on entre valoraci
on perceptual y
par
ametros ac
usticos del vibrato en tenores famosos. XIX ENT World
Congress-IFOS 2009 Brazil. S~ao Paulo, Brazil, June 2009. CD Room.
35. Miller D, Nair G. Scgutte H, Horne R. Voce Vista Manual (V. 3.2). [Com-
puter program]. Available at: http://www.vocevista.com/. University of
Groeningen, voice research laboratory, Groningen, 2008.
36. Sundberg J. Acoustic and psychoacoustic aspects of vocal vibrato. Q Prog
Status Rep. 1994;35:45–68. KTH Speech Transmission Laboratory.
37. Scherer KR. Expression of emotion in voice and music. J Voice. 1995;9:
235–248.
38. Winckel F. Physical criteria for objective voice assessment. Folia Phoniatr.
1953;5:231–252.
39. Bretos J, Sundberg J. Measurements of vibrato parameters in long sustained
crescendo notes as sung by ten sopranos. J Voice. 2003;17:343–352.
40. Michel J, Myers D. The effect on crescendo of vocal vibrato. J Voice. 1991;
5:292–298.
41. Diaz J, Rothman H. Acoustical comparison between samples of good and
poor vibrato in singers. J Voice. 2003;17:179–184.
42. Orlikoff RF, kahane JC. Influence of mean sound pressure level on jitter and
shimmer measures. J Voice. 1991;5:113–119.
43. Jacob L. A normative study of laryngeal jitter [unpublished master’s thesis].
Kansas: University of Kansas, 1968.
44. Cecconello C. Acoustic characteristic of vibrato in choral singers. Rev


Arete. 2010;10:17–25.