ARTICLE IN PRESS

Aerodynamic Characteristics of Growl Voice and

Reinforced Falsetto in Metal Singing
*Marco Guzman, †Karol Acevedo, ‡Fernando Leiva, §Vasti Ortiz, ¶Nicolas Hormazabal, and §Camilo Quezada,
*†‡§Santiago and ¶Chicureo, Chile

Summary: Objectives. The present study aimed to assess the aerodynamic characteristics of vocally healthy metal
singers when producing growl voice or reinforced falsetto.
Methods. Fifty-four participants (metal singers) were initially enrolled in this study, with 23 meeting the inclusion
criteria. Sixteen participants performed growl voice and seven performed reinforced falsetto as a voice resource during
metal singing. All participants were asked to undergo rigid laryngeal videostroboscopy to confirm the absence of la-
ryngeal pathology. Then, subjects were aerodynamically assessed while performing growl voice or reinforced falsetto.
Results. Higher glottal airflow rate, sound pressure level, and subglottic pressure (Psub) for growl voice samples com-
pared to vowel production without growl voice (keeping the same fundamental frequency [F0]) were found. Higher
Psub, sound pressure level, and glottal resistance for high-pitched reinforced falsetto compared to naïve falsetto (keeping
the same F0) were found. No differences for F0 were found for neither growl voice nor reinforced falsetto.
Conclusions. It seems that growl voice is produced by decreasing vocal folds adduction and increasing Psub, which
in turn, promotes an increased airflow rate. Reinforced falsetto is characterized by an increased vocal fold adduction
and an increased Psub. A proper resonance strategy in reinforced falsetto and a decreased glottal adduction in growl
voice might probably be the factors that contribute to prevent voice problems in singers who use these vocal re-
sources, classically labeled as vocal abuse.
Key Words: Subglottic pressure–Glottal airflow–Glottal resistance–Singing voice–Supraglottic compression.

INTRODUCTION Few previous studies have been addressed to explore the un-
The present study is a continuation of a line of research whose derling physiology of growl voice. In a laryngeal endoscopic
general objective is to explore supraglottic constriction and the study, Sakakibara et al6 showed that growl voice is produced
underlying physiology of different types of vocal sounds in through the simultaneous vibration of the vocal folds and la-
singing and speaking voice. ryngeal supraglottic structures. The vocal folds vibrate periodically
Certain contemporary commercial music (CCM) singing styles and the aryepiglottic folds generate subharmonics. The presence
commonly include vocal sounds that could be considered as of subharmonics during growl voice production has also been
harmful and detrimental for voice. Some of these vocal sounds found in other studies.7–9 Vibration of aryepiglottic folds was also
are produced by narrowing of epilaryngeal and/or other vocal reported in a radiologic study.6 Additionally, authors showed that
tract structures, causing a distorted and auditorily perceptual during growl voice production, the larynx rose to the fourth ver-
pressed sound. Growl voice is one of these vocal resources, being tebra and there was a large anterior-posterior (A-P) laryngeal
frequently used in jazz, blues, pop, gospel, and metal singing constriction.6 Recently, Caffier et al found that during growl voice
among others.1,2 Some ethnic singing from Brazil, Japan, and production, glottal closed phase is decreased because the vocal
South Africa also include similar voice productions.3 In metal folds vibrate decoupled due to reduced contact.5
singing (eg, dead metal, black metal, and metalcore), growl voice Other vocal resource that metal singers (especially in heavy
can be also called as death growl, death metal vocals, guttural metal) use as part of their vocal technique is the reinforced fal-
vocals, grunting, unclean vocals, and harsh vocals.4 Caffier et al5 setto. This type of falsetto has been described as a voice
stated that grunting singing is stylistically suited to reproduce production that uses a greater vocal folds adduction compared
gloomy or brutal lyrics. It has a devilish sound and can express to naïve falsetto (nontrained falsetto).10,11 The greater adduc-
anger and despair.5 tion during reinforced falsetto production results in loud and bright
sound. Acoustically, the perceptual differences between naïve
and reinforced falsetto are caused by an increment of harmon-
Accepted for publication April 30, 2018. ic energy in the high spectral region (2–5 kHz) in the latter.11
From the *Universidad de los Andes, Chile. Department of Communication Sciences
and Disorders. Avenida Monseñor Alvaro del Portillo 12455, Santiago, Chile; †Pontificia
Reinforced falsetto may be used in CCM musical styles such
Universidad Catolica de Chile, Department of Communication Sciences and Disorders. as blues, glam rock, epic metal, heavy metal, soul, etc.
Avenida Vicuña Mackena 4860, Santiago, Chile; ‡Universidad Pedro de Valdivia, Depart-
ment of Communication Sciences and Disorders. Avenida Vicuña Mackena 44, Santiago,
Guzman et al11 conducted a study aimed to determine the res-
Chile; §University of Chile, Department of Communication Sciences and Disorders. Avenida onance strategies of metal singers when singing high pitches with
Independencia 1027, Santiago, Chile; and the ¶Baume Academy. El alba 2, Parcela5, Chicureo,
Chile.
reinforced falsetto. Authors suggested that this type of voice pro-
Address correspondence and reprint requests to: Marco Guzman, Universidad de los duction is associated with a shortening of the vocal tract and a
Andes, Chile. Avenida Monseñor Álvaro del Portillo 12455, Santiago, Chile. E-mail:
guzmann.marcoa@gmail.com
very open mouth configuration. High first and second for-
Journal of Voice, Vol. ■■, No. ■■, pp. ■■-■■ mants values in all of the participants seem to reflect articulatory
0892-1997
© 2018 Published by Elsevier Inc. on behalf of The Voice Foundation.
characteristics, such as a wider jaw opening, a narrower pharynx,
https://doi.org/10.1016/j.jvoice.2018.04.022 spread lips, and high vertical laryngeal position.11
 ARTICLE IN PRESS
2 Journal of Voice, Vol. ■■, No. ■■, 2018

Because growl voice and reinforced falsetto seem to percep- Strobo 21 Led; ATMOS MedizinTechnik, Lenzkirch, Germany)
tually be a pressed and laryngeal hyperfunctional voice, it is not to confirm the absence of laryngeal pathology. Laryngoscopic
unreasonable to think that long-term use of these voice produc- examinations were performed by one experienced voice pathol-
tions may cause voice disorders. However, some metal singers ogist who is co-author of the present study. No topical anesthesia
and singing teachers maintain that long-term use of these tech- was used during endoscopic procedure.
niques will not cause harm to the voice if produced in a healthy
manner. In a study aimed to assess a group of metal singers who Equipment and data collection
engage in growl voice and/or reinforced falsetto, Guzman et al12 Aerodynamic and acoustic measures were captured for each par-
found that these techniques do not seem to contribute to laryn- ticipant during growl voice or reinforced falsetto productions.
geal disorders. Perceptual, acoustic, functional, and laryngoscopic All samples were recorded digitally at a sampling rate of 22.1
assessment showed no major alterations in most of the partici- KHz with 16 bits/sample quantization. Acoustic signal was re-
pants. Moreover, no significant differences in voice and laryngeal corded using the incorporated condenser microphone AKG CK
evaluation were found when comparing these metal singers with 77 (AKG Acoustics, Vienna, Austria) that Phonatory Aerody-
other CCM singers who do not use neither growl voice nor re- namic System provides (KayPENTAX, Lincoln Park, NJ). A
inforced falsetto. 12 Recently, in a study conducted to constant microphone-to-mouth distance of 20 cm was used.
multidimensionally investigate common vocal effects in expe- Acoustic samples were captured to obtain mean fundamental fre-
rienced professional CCM singers, laryngeal structures of all quency (F0) and mean sound pressure level (SPL). F0 was
singers were found to be healthy in spite of using vocal effects obtained to ensure that this variable was kept constant between
such as growl voice and vocal distortion over many years.5 To repetitions as required to participants.
the best of our knowledge, no more studies have been ad- Aerodynamic device was connected to a computer through
dressed to assess voice condition in metal singers. a Computerized Speech Lab, model 4500 (KayPENTAX, Lincoln
The present study aimed to assess aerodynamic characteris- Park, NJ). A Phonatory Aerodynamic System, model 4500
tics of vocally healthy metal singers during production of growl (KayPENTAX, Lincoln Park, NJ) was used to collect aerody-
voice or reinforced falsetto. We hypothesized that aerody- namic data. Calibration of the air pressure and airflow signals
namic characteristic during growl voice and reinforced falsetto was performed before data acquisition according to the manu-
productions would reflect a low phonatory effort even if these facturer’s instructions. A real-time aerodynamic analysis software,
techniques sound like vocally stressful behaviors. model 6600, version 3.4 (KayPENTAX, Lincoln Park, NJ) was
used to analyze all samples. From the middle section of each
METHODS sample, the most stable part was analyzed.
Participants
Phonatory tasks
Fifty-four participants (metal singers) were initially enrolled in
Phonatory tasks required to participants were not the same in
this study. Only 23 met the inclusion criteria. Sixteen of them
both groups. The following are the phonatory tasks for growl
performed growl voice and seven subjects performed rein-
voice group: (1) a sustained vowel [a:] producing growl voice
forced falsetto as a voice resource during metal singing. Mean
at a comfortable and usual pitch, (2) a sustained vowel [a:] without
age of growl voice group was 26 years, ranging from 21 to 32.
growl voice keeping the same F0 as task 1, (3) repetition of the
Mean age of the reinforced falsetto group was 27 years, ranging
syllable [pa:] producing growl voice at a comfortable and usual
from 22 to 35. Inclusion criteria for participants were: (1) at least
pitch, and (4) repetition of the syllable [pa:] without growl voice
1 year of experience singing growl voice or reinforced falsetto,
keeping the same F0 as task 3. The following are the phona-
(2) no current or past history of major voice problems during
tory tasks for reinforced falsetto group: (1) a sustained vowel
last year, (3) perceptually normal voice, (4) absence of any struc-
[a:] producing reinforced falsetto at a high and comfortable pitch,
tural or functional vocal fold disorders, (5) self-reported normal
(2) a sustained vowel [a:] producing naïve falsetto keeping the
hearing, and (6) technically appropriate singing voice tech-
same F0 as task 1, (3) repetition of the syllable [pa:] produc-
nique (growl voice or reinforced falsetto) assessed by a singing
ing reinforced falsetto at a high and comfortable pitch, and (4)
teacher who teaches these types of voice techniques for metal
repetition of the syllable [pa:] producing naïve falsetto keeping
singers (co-author of the present paper). Perceptual assessment
the same F0 as task 3. Repetition of the syllable [pa:] was asked
was conducted by one of the authors of this paper, using GRBAS
to obtain the subglottic pressure (Psub) from the oral pressure
(grade, roughness, breathiness, asthenia, strain) scale. Partici-
(Poral) during the occlusion of the consonant [p:]. A silicon tube
pants in both groups were native speakers of Spanish and reported
inserted into the mouth was used to acquire Poral. Participants
no hearing problems. The present study was approved by the
were asked not to touch the tube with their tongue or with any
institutional review board at University of Chile and all partici-
other oral structure so as to not block the airflow. Three repeti-
pants signed informed consent.
tions of all phonatory tasks were performed by each subject. F0
was required to be the same during all phonatory tasks.
Laryngoscopic assessment
Once subjects passed auditory perceptual assessment and singing Variables
voice technique evaluation, they were asked to undergo rigid la- From acoustic and aerodynamic samples acquired, the follow-
ryngeal videostroboscopy (digital videostroboscopy system; Atmos ing variables were obtained:
 ARTICLE IN PRESS
Marco Guzman, et al Growl Voice and Reinforced Falsetto 3

(1) F0 (Hz) from acoustic signal. During sustained vowel [a:] (Figure 1), significant differences
(2) SPL (dB) from acoustic signal. were found for mean SPL (CI 5.176, 11.635, P < 0.001) and
(3) Psub (cm H2O) estimated from the maximum peak of glottal airflow (CI 0.585, 1.205, P < 0.001) when comparing growl
the Poral during the occlusion of the consonant [p:] in voice production and vowel production without growling. No
the syllable [pa:]. differences were detected for F0 (−28.226, 23.036, P = 0.52).
(4) Mean glottal airflow (L/seg) from aerodynamic signal. During repetition of the syllable [pa:] (Figure 2), significant
(5) Glottal resistance (cm H2O/L/seg) from aerodynamic differences were found for mean SPL (CI 5.961, 10.618,
signal, defined as Psub divided by glottal airflow. P < 0.001), glottal airflow (CI 0.666, 1.258, P < 0.001), Psub (CI
14.811, 35.654, P < 0.001), and glottal resistance (CI −187.033,
Statistical analysis −15.233, P < 0.001). No differences were detected for F0
Data were processed with R software (version 3.4.4, The R Foun- (−21.605, 39.678, P = 0.52).
dation for Statistical Computing, Vienna, Austria).13,14 Because Figures 3 and 4 display results from phonatory tasks (sus-
of limited sample size, differences between singing styles were tained vowel [a:] and repetition of the syllable [pa:]) performed
explored by means of a nonparametric test (paired-sample with reinforced falsetto and naïve falsetto. During sustained vowel
Wilcoxon test). Moreover, because small sample sizes are known [a:] (Figure 3), significant differences were found only for mean
to impact both P-values and effect size, a more robust measure SPL (CI −12.591, −3.853, P = 0.01). No differences were evi-
of differences was obtained by calculating confidence intervals denced neither for F0 (CI −16.308, 1.961, P = 0.1) or for glottal
(CIs) for each contrast, based on the median of the differences airflow (CI −0.0100, 0.2033, P = 0.15). During repetition of the
between the two measures. Thus, only CIs not including zero syllable [pa:] (Figure 4), significant differences for Psub (CI
were considered as significant and are here reported.15 −23.903, -9.441, P = 0.02) were observed. No differences were
detected for the rest of dependent variables (F0 −12.74, 5.16,
RESULTS P = 0.93; glottal airflow −0.0916, 0.138, P = 0.27; glottal resis-
Figures 1 and 2 show results from phonatory tasks (sustained tance −228.248, 12.916, P = 0.07).
vowel [a:] and repetition of the syllable [pa:]) performed during The lack of significant differences in F0 during all phona-
growl voice production and voice production without growl voice. tory tasks are expected and desired because all participants were

SPL (dB) F0 (Hz) Gottal airflow (Lt/sec)

1.5
100
200

75 150

1.0

50 100

0.5

25 50

0 0 0.0

Without growl Growl voice Without growl Growl voice Without growl Growl voice

FIGURE 1. Results from sustained vowel [a:] performed during growl voice production and voice production without growl voice. Error bars
represent 95% CI for the means of each variable for each condition. Nonoverlapping or slightly overlapping error bars are to be interpreted in
terms of significant differences between conditions. Widely overlapping error bars translate into nonsignificant differences.
 ARTICLE IN PRESS
4 Journal of Voice, Vol. ■■, No. ■■, 2018

SPL (dB) F0 (Hz) Psub (cm H2O)

200
100

40
150
75

30

50 100

20

25 50
10

0 0 0

Without growl Growl voice Without growl Growl voice Without growl Growl voice
Glottal airflow (Lt/sec) Glottal resistance (cm H2O/Lt/sec)
1.5

400

1.0 300

200

0.5
100

0
0.0

Without growl Growl voice Without growl Growl voice

FIGURE 2. Results from repetition of the syllable [pa:] performed during growl voice production and voice production without growl voice.
Error bars represent 95% CI for the means of each variable for each condition. Nonoverlapping or slightly overlapping error bars are to be inter-
preted in terms of significant differences between conditions. Widely overlapping error bars translate into nonsignificant differences.

required to keep the same pitch during voice production and rep- decreased vocal folds adduction and also a high Psub. A low glottal
etitions of growl voice and reinforced falsetto. closure (to allow a high airflow rate) clearly would decrease the
probability of phonotrauma during growling. Caffier et al,5 through
DISCUSSION endoscopic imaging and electroglottography, reported that during
Previous studies on vocal resources commonly used in metal production of some vocal resources commonly used in rock (eg,
singing (growl voice and reinforced falsetto) have explored un- grunting and rattle), the closed phase of the glottis is decreased
derlying physiology through endoscopic assessment, radiologic because the vocal folds vibrate decoupled due to reduced contact.
examination, and acoustic analysis.5–9 To the best of our knowl- Caffier et al mentioned that there is no real oscillation cycle but
edge, no research has been conducted to describe aerodynamic rather tonal noise.5 Although a high Psub has been associated to
behaviors of these techniques. Therefore, the purpose of the a high phonatory effort, data from the present study would allow
present study was to assess aerodynamic characteristics of vocally to speculate that high Psub was likely used as aerodynamic strat-
healthy metal singers during productions of growl voice and re- egy to produce vibration of supraglottic structures rather than
inforced falsetto. Based on our previous published data and earlier vocal folds oscillation. If vocal folds adduction would not be
studies by other researchers, we expected aerodynamic charac- low during growl voice production, a high glottal airflow would
teristics during growl voice and reinforced falsetto productions not be possible, making unlikely the vibration of supraglottic
to reflect a low phonatory effort, even though these techniques structures as described in earlier studies.6,7,16
sound like vocally stressful behaviors. Activity of supraglottic structures during growl voice pro-
Inspection of results revealed that some aerodynamic char- duction seems to be one of the most relevant aspects during this
acteristics of vowel production with growl voice might correlate technique. Results by Guzman et al12 revealed that growl voice
with nonphonotraumatic phonatory practice. Glottal airflow and is characterized by supraglottic compression, pharyngeal con-
Psub were higher during growl voice production as compared striction, and changes in vertical laryngeal position. Guzman et al
to voice production without growl. In addition, a lower glottal suggested that the rough and raspy perceptual quality of growl
resistance was observed during growl voice production. Because voice production is likely produced by the vibration of the su-
glottal airflow is defined as the ratio between Psub and glottal praglottic structures involved in the observed laryngeal
resistance, it seems natural to think that higher airflow found in compression.12 Similarly, Borch et al16 found significant supra-
the present study during growl voice production was caused by glottic activity including the aryepiglottic folds, anterior part of
 ARTICLE IN PRESS
Marco Guzman, et al Growl Voice and Reinforced Falsetto 5

SPL (dB) F0 (Hz) Gottal airflow (Lt/sec)

0.6
200
100

150
75

0.4

50 100

0.2

25 50

0 0 0.0

Reinforced falsetto Naive falsetto Reinforced falsetto Naive falsetto Reinforced falsetto Naive falsetto

FIGURE 3. Results from sustained vowel [a:] performed during production of reinforced falsetto and naïve falsetto. Error bars represent 95%
CI for the means of each variable for each condition. Nonoverlapping or slightly overlapping error bars are to be interpreted in terms of significant
differences between conditions. Widely overlapping error bars translate into nonsignificant differences.

the mucosa covering the arytenoids, and ventricular folds (medial previously assessed through acoustical analysis in a group of metal
compression). Additionally, high-speed digital imaging (fiber- singers.11 Results revealed high values of the two first vocal tract
scope inserted into the nose cavity) revealed that growl voice formant frequencies (F1 and F2) in all participants. Data from
production presented simultaneous vibration of laryngeal su- the present study showed higher values for total SPL during re-
praglottic structures (likely caused by high airflow rate).7 Recently, inforced falsetto production compared to naïve falsetto
Caffier et al5 reported that during production of rattle (a vocal production.11 An adequate resonance strategy might also be the
resource used in rock) the aryepiglottic folds play an impor- cause for this increase. In fact, in the previously cited reso-
tant role in producing crackling sound, the supraglottic nance strategy study,11 a clear increase of the spectral energy in
anteroposterior constriction is clearly visible, and the vocal folds one or two harmonics due to the coincidence or proximity with
vibrate relatively relaxed. Even if growl voice possesses a quality one or two first formants was shown for all singers. These out-
that may sound pressed or constricted, the major work may not comes suggest that the total SPL during reinforced falsetto
be performed by the vocal folds, but by the increased activity production may not be mainly supported by increasing vocal effort
of laryngeal supraglottic structures and aerodynamic behavior (high glottal resistance and high Psub), but also by an ade-
(high glottal airflow and Psub). quate vocal tract configuration strategy.
Results from falsetto analysis indicated that significant higher As for acoustic analysis during high-pitched voice produc-
values of both Psub and glottal resistance were found for rein- tion, similar results were reported by Borch et al in a group of
forced falsetto when compared to naïve falsetto’s. At the first rock singers.13 They claimed that high values of formant fre-
glance, it seems that reinforced falsetto may involve a more quencies reflected a relatively high larynx position and open
phonotraumatic voice production than naïve falsetto, as both high mouth configuration. Guzman et al11 found that all assessed metal
Psub and gottal resistance are associated to phonatory effort. singers spontaneously used the vowel /a/ (open vocal tract con-
However, in a previous study performed with metal singers who figuration) during the production of reinforced falsetto in high
regularly use this technique, no major vocal folds disorders were pitches. Earlier studies have shown that a similar vocal tract con-
reported. The use of proper resonance strategies (vocal tract shape) figuration is classically displayed by sopranos during production
to prevent excessive vocal folds muscle efforts during high- of high pitches to avoid laryngeal muscle effort.17–19
pitched reinforced falsetto might explain lack of voice damage. High activity and compression of supraglottic structures pre-
Resonance strategies during reinforced falsetto production were viously reported in growl voice7,12,16 and reinforced falsetto
 ARTICLE IN PRESS
6 Journal of Voice, Vol. ■■, No. ■■, 2018

SPL (dB) F0 (Hz) Psub (cm H2O)

200
40
90

150
30

60
100
20

30
50 10

0 0 0

Reinforced falsetto Naive falsetto Reinforced falsetto Naive falsetto Reinforced falsetto Naive falsetto
Glottal airflow (Lt/sec) Glottal resistance (cm H2O/Lt/sec)

300

0.2

200

0.1
100

0.0 0

Reinforced falsetto Naive falsetto Reinforced falsetto Naive falsetto

FIGURE 4. Results from repetition of the syllable [pa:] performed during production of reinforced falsetto and naïve falsetto. Error bars repre-
sent 95% CI for the means of each variable for each condition. Nonoverlapping or slightly overlapping error bars are to be interpreted in terms of
significant differences between conditions. Widely overlapping error bars translate into nonsignificant differences.

productions,11,12 have been also found in other singing styles and 2. Esling JH. Pharyngeal consonants and the aryepiglottic sphincter. J Int Phonet
in speaking voice. Medial and A-P laryngeal compression have Assoc. 1996;26:65–88.
3. Zemp H. Les Voix du Monde: Une Anthologie des Expressions Vocales. Paris:
commonly been described as endoscopic signs of vocal Musée de l’Homme; 1996.
hyperfunction.20,21 Previous studies, however, have shown that 4. Bang K. Your Head: The Rise and Fall of Heavy Metal. New York: Three
an increased supraglottic activity may also be a normal and even Rivers Press; 2002.
desirable vocal behavior in subjects diagnosed with normal 5. Caffier PP, Ibrahim Nasr A, Ropero Rendon MDM, et al. Common vocal
voice.22–27 Medial and A-P laryngeal supraglottic activity have effects and partial glottal vibration in professional nonclassical singers. J
Voice. 2018;32:340–346.
been observed in healthy opera singers,23,25 CCM singers,21,23,26 6. Sakakibara K, Fuks L, Imagawa H, et al.: Growl voice in ethnic and pop
and in professional theater actors.27 styles. Proc Int Symp on Musical Acoustics (ISMA 2004), Nara, 2004.
7. Fuks L, Hammarberg B, Sundberg J. A self-sustained vocal ventricular
phonation mode: acoustical, aerodynamic and glottographic evidences. KTH
CONCLUSION TMH-QPSR. 1998;3:49–59.
Growl voice seems to be produced by decreasing vocal folds ad- 8. Lindestad PA, Sodersten M, Merker B, Granqvist S. Voice source
duction and increasing Psub, which in turn promotes an increased characteristics in Mongolian “throat singing” studied with high-speed imaging
airflow rate. Reinforced falsetto is characterized by an in- technique, acoustic spectra, and inverse filtering. J Voice. 2001;15:78–85.
creased vocal fold adduction and an increased Psub. Likely, a 9. Sakakibara K, Konishi T, Kondo K, et al.: Vocal fold and false vocal fold
vibrations and synthesis of Khöömei. Proc ICMC 2001, Japan, 2001.
proper resonance strategy in reinforced falsetto and a de- 10. Miller DG, Svec JG, Schutte HK. Measurement of characteristic leap interval
creased glottal adduction in growl voice could be the factors that between chest and falsetto registers. J Voice. 2002;16:8–19.
contribute to the avoidance of voice problems in singers who 11. Guzman M, Barros M, Espinoza M, et al. Resonance strategies in rock
use these vocal resources that could be classically labeled as vocal singers. J Singing. 2014;71:183–192.
abuse. 12. Guzman M, Barros M, Espinoza F, et al. Laryngoscopic, acoustic, perceptual,
and functional assessment of voice in rock singers. Folia Phoniatr Logop.
2013;65:248–256.
REFERENCES 13. R Core Team. R: A Language and Environment for Statistical Computing.
1. Catford JC. Fundamental Problems in Phonetics. Edinburgh: Edinburgh Vienna, Austria: R Foundation for Statistical Computing; 2017. https://
University Press; 1977. www.R-project.org/.
 ARTICLE IN PRESS
Marco Guzman, et al Growl Voice and Reinforced Falsetto 7

14. Wickham H. ggplot2: Elegant Graphics for Data Analysis. New York: 21. Behrman A, Dahl L, Abramson A, et al. Anterior-posterior and medial
Springer-Verlag; 2009. compression of the supraglottis: signs of nonorganic dysphonia or normal
15. Fan X. Statistical significance and effect size in education research: two postures? J Voice. 2003;17:403–410.
sides of a coin. J Educ Res. 2001;94:275–282. 22. Lawrence V. Laryngological observations on belting. J Res Singing.
16. Borch DZ, Sundberg J, Lindestad PA, et al. Vocal fold vibration 1979;2:26–28.
and voice source aperiodicity in “dist” tones: a study of a timbral 23. Yanagisawa E, Estill J, Kmucha S, et al. The contribution of aryepiglottic
ornament in rock singing. Logoped Phoniatr Vocol. 2004;29:147– constriction to “ringing” voice quality—a videolaryngoscopic study with
153. acoustic analysis. J Voice. 1989;3:342–350.
17. Sundberg J, Skoog J. Dependence of jaw opening on pitch and vowel in 24. Pershall K, Boone S. Supraglottic contribution to voice quality. J Voice.
singers. J Voice. 1997;11:301–308. 1987;1:186–190.
18. Hertegard S, Gauffin J, Sundberg J. Open and covered fiberoptics, inverse 25. Mayerhoff RM, Guzman M, Jackson-Menaldi C, et al. Analysis of
singing as studied by means of filtering, and spectral analysis. J Voice. supraglottic activity during vocalization in healthy singers. Laryngoscope.
1990;4:220–230. 2014;124:504–509.
19. Sundberg J, Gramming P, LoVetri P. Comparisons of pharynx, source, 26. Guzman M, Lanas A, Olavarria C, et al. Laryngoscopic and spectral analysis
formant, and pressure characteristics in operatic and musical theatre singing. of laryngeal and pharyngeal configuration in non-classical singing styles.
J Voice. 1993;7:301–310. J Voice. 2015;29:130.e21–130.e28.
20. Stager S, Bielamowicz S, Regnell J, et al. Supraglottic activity: evidence 27. Guzman M, Ortega A, Olavarría C, et al. Comparison of supraglottic activity
of hyperfunction or laryngeal articulation? J Speech Hear Res. 2000;43:229– and spectral slope between theater actors and vocally untrained subjects. J
238. Voice. 2016;30:767.e1–767.e8.