You are on page 1of 19

ARTICLE IN PRESS

Laryngeal Vibratory Mechanisms: The Notion of Vocal
Register Revisited

*Bernard Roubeau, †Nathalie Henrich, and ‡Miche`le Castellengo
*‡Paris, France and †Grenoble, France

Summary: This study, focused on the laryngeal source level, introduces the
concept of laryngeal vibratory mechanism. Human phonation is characterized
by the use of four laryngeal mechanisms, labeled M0–M3, as evidenced by
the electroglottographic (EGG) study of the transition phenomena between
mechanisms with a population of men and women, trained and untrained
singers. Macroscopic and local descriptions of the EGG signal are analyzed
during the production of glissandos and held notes with different mecha-
nisms. The transition from one mechanism to another of higher rank is char-
acterized by a jump in frequency, a reduction of EGG amplitude, and
a change in the shape of the derivative of the EGG (which may correspond
to a reduction of the vibratory mass). These characteristics are used to iden-
tify a transition between two mechanisms, in complement with acoustic spec-
trographic analyses. The pitches of transitions between the two main
mechanisms M1 and M2 and the range of the frequency-overlap region are
described in detail. The notion of vocal register is revisited in the light of
these concepts of laryngeal mechanism. The literature on vocal registers is
reviewed, and it is shown that the confusion often cited with respect to this
notion may be related to the heterogeneity of the approaches and methods
used to describe the phenomena and to the multiplicity of descriptors. There-
fore, the terminology of the registers is organized depending on their relation
to the four laryngeal vibratory mechanisms.
Key Words: Laryngeal mechanism—Electroglottography—Larynx—Sing-
ing voice—Voice range—Register.

INTRODUCTION apparatus, encompassing zones called registers.1
Human voice production over the whole frequency These registers are described by physiologists,2,3
range involves different adjustments of the vocal physicians,4,5 phoneticians and voice scientists,6–11
voice teachers,12,13 and singers. Obviously such

Accepted for publication October 17, 2007. Avenue Fe´lix Viallet, 38031 Grenoble Cedex 1, France.
From the *Service d’ORL et de Chirurgie Cervico-faciale, E-mail: nathalie.henrich@gipsa-lab.inpg.fr
Hoˆpital Tenon, Paris, France; †De´partement Parole et Cogni- Journal of Voice, Vol. -, No. -, pp. -
tion, GIPSA-lab, CNRS-INPG-UJF-UnivStendhal, Grenoble, 0892-1997/$34.00
France; and the ‡L.A.M.-IJLRA, CNRS-University Paris 6, Ó 2008 The Voice Foundation
Ministe`re de la Culture, Paris, France. doi:10.1016/j.jvoice.2007.10.014
Address correspondence and reprint requests to Nathalie
Henrich, De´partement Parole et Cognition, GIPSA-lab, 46,

1

The fundamental frequency was auto- a macroscopic and a microscopic study of the con.3. Then. MA) placed at 1 m Journal of Voice. geal vibratory mechanism is introduced to describe The subject’s pitch range was covered in semitones the different configurations that the laryngeal vibra. 10 amateur. In the following section. -. No. be they audible or not. matically extracted by an FIES-CNET melograph tact-area signal. Nineteen subjects participated in this study: 10 male This noninvasive observation technique of laryn. or vibratory dis- terests in the voice. and three untrained singers) and 21 female subjects neous audio and EGG recordings of singers and (five professional. The relationship between a laryngeal mecha- scopic.15 audi. oscillation itself. a voice range profile for each laryngeal mechanism. which can be geneous. and their approaches are hetero. room (LEM microphone. Database 2 (DB2) tions in amplitude and frequency. Denmark). continuities. for and 329 Hz—E4.14. Geneva. contractions or laryngeal vibrations. nonpathological voice phonation. Finally. 2007 . and /i/) with a change of laryngeal mechanism at signal and the glottal contact area in the case of three different pitches (293 Hz—D4. Three databases (DBs) of simulta.16–18 It is evident that some observations ryngeal mechanisms. works.7 or proprioceptive observations. are briefly presented.1 electrophysiological. Switzerland and sive vibratory cycles. 11 amateur. The subjects review of literature). Acoustic discussed in the third section. and sustained vowels (/a/. Forty-two subjects participated in this study: 21 scopic analysis pertains to the shape of the male subjects (seven professional. Sound-pressure level (SPL) was tor can take throughout the human voice frequency measured in dBA by a sound level meter (1560-P range. Vol. concerned with the overall shape of the envelope of the EGG amplitude-time waveform and its varia. -. In this paper. subjects (seven trained and three untrained singers) and geal activity is well known and has been used since nine female subjects (seven trained and two untrained its conception by Fabre in 195719 (see Childers and singers). which brings great confusion to this domain. the concept of laryn. This exploratory technique allows gen. There is The three databases to which we refer and which need for a better understanding of the notion of vocal were recorded for the purpose of previous studies register. and six untrained nonsingers are used in Material and Methods sec. ARTICLE IN PRESS 2 BERNARD ROUBEAU ET AL varied workers have developed widely different in. nal. There is likely to be a good correlation between this /o/. the register terminology is organized depend- nating cavities or the sensations characteristic of ing on the relation of the registers to the four prin- the proprioceptive stimulations because of muscle cipal laryngeal vibratory mechanisms. West Concord. For this purpose.6. The purpose of this study was to establish tion.21. Therefore.5. whereas others include the action of the reso. confu. the terms used are similar.14 acoustic. defined by the rupture phenomena. The EGG signal is based on were asked to produce ascending and descending glis- the monitored conductance between the vocal folds. The descriptions which have been given of detected in a macroscopic analysis of the EGG sig- these ‘‘registers’’ can be derived from laryngo. MATERIAL AND METHODS sion that is often mentioned by the authors. France).5. Audio and EGG sig- a normal voice it enables one to evaluate indirectly nals were recorded simultaneously in a soundproof the amplitude of vocal fold contact during succes. on the vowel [a]. we focus on the laryngeal source level. as well as the main phases Frokjaer-Jensen EG 830 electroglottograph. The borders between the mechanisms are General Radio.23 The macroscopic analysis is (Paris. sandos in a spontaneous way. respectively).25 The micro. The vocal-register literature is have more to do with the way the laryngeal source reviewed to point out the sources of confusion. nism and the EGG glottal-pulse shape is tory.24. electroglottography (EGG) Database 1 (DB1) seems to us an appropriate experimental technique.22 for a detailed shift between laryngeal mechanisms. the notion and electrophysiological observations are very often of vocal registers is revisited in the light of the la- combined. 311 Hz—D#4. subjects). Copenha- of this contact. In spite of the diversity of approaches. The purpose of this study was to analyze the Krishnamurthy20 or Henrich et al21.

asked to indicate in which laryngeal mechanism thus defining with precision four frequency areas they were singing. and sustained vowels ([a]. two systems that one can collect valuable informa- plifier (Brue¨l & Kjær NEXUS 2690). with no specific concern for es. an ascending 1500 Hz) in the vocal range. two tenors. A few studies have stressed in an indirect way. pre.23–25. EG2 electroglottograph).24 for phenomena have been much less investigated. duce glissandos. and three counter-tenors) and six female singers (three mezzo- sopranos and three sopranos). Version 6. The technical computing although these different mechanisms have been environment MATLAB. several parts. M1. and at three different an upward jump in fundamental frequency.28 However.10.29 mechanisms An ascending glissando is a vocal production EGG analysis of the M1–M2 transition during during which the frequency progressively goes glissandos from the lowest pitch (sometimes around 20 Hz) In the zone covering the frequencies most used in to the highest (in some cases up to 1000. and conditioning am. which could be related to Journal of Voice. ter and the sources of confusion. It is characterized by an upward fre- thetic quality. This break appears In database DB1. quite extensively described as the result of a varie- ital signal processing. one can observe several events dis. the M1 and M2 mechanisms. Audio and EGG signals were re. (Figure 2). A descending glis. singers quency jump. two-channel tion on the systems themselves.1 was used for dig. The database comprises FIGURE 1. the most remarkable elements be- CHANGES WITH LARYNGEAL VIBRATORY ing the modification of EGG signal amplitude and MECHANISM TRANSITIONS the modification of the shape of the waveform From transition phenomena to laryngeal itself.9. male and female subjects pro. that we call laryngeal mechanisms M0. quiet studio room. -. There loudness levels. in which the singers were asked to pro. 2007 . Roubeau demonstrated in section Historical review of regis- et al26 for DB2. glissando presents one unique break separating sando is the inverse production. in the acoustic signal and in the EGG signal duced glissandos. ARTICLE IN PRESS LARYNGEAL VIBRATORY MECHANISMS 3 from the subject’s mouth. even speech and song by both genders. and Henrich et al21. [e]. M2. Female subject. Sonogram of an ascending vocal glissando with the successive use of four laryngeal vibratory mechanisms. The purpose of this study was to explore the variations of open quotient in western operatic singing.5. changes of glottal con- figuration thanks to concomitant transformations of MACROSCOPIC STUDY OF EGG SIGNAL the EGG signal. No. a reduction of the amplitude of the and nonsingers. and M3. We shall return to the justification for condenser microphone (Brue¨l & Kjær 4165 Naerum. this terminology in the sequel to this article. It is at the critical points of equilibrium between amplifier (Brue¨l & Kjær 2669). EGG signal. The re- rupting the evolution of the frequency. ty of observational techniques. as more technical details concerning DB1. as shown in duction in amplitude of the EGG signal may come the spectral analysis presented in Figure 1. These from a reduction of the contact surface area be- events are visible on the spectrographic analysis tween the vocal folds. crescendos. -. and [u]) for different pitches covering their as disruptions of the harmonic curves related to comfortable frequency range. the singers were are at most three of these transitory phenomena. Vol. Denmark) placed 50 cm from the singer’s mouth.28. corded simultaneously in a soundproof room (1/2 in. and a modification of its shape. When necessary. The recording took place in a large. the transitional Readers are referred to Roubeau et al23. Database 3 (DB3) Eighteen trained singers participated in this study: 12 male singers (seven baritones. With most of the subjects.27 for DB3.

Vol. the cover may decouple from the deep layer. the change of mechanism is characterized by an abrupt in- crease in the amplitude of the EGG signal and an abrupt fall in frequency. No. Its perception by a listener. thanks to variation of the fundamental frequency (F0). It is this abrupt reduction that probably FIGURE 3. As suggested by Titze. and glissandos was aimed at minimizing phenomena variation of the fundamental frequency (F0). during the production of ascending glissandos and descending glissandos. EGG. -. The latter is no lon- ger part of the vibration. tension and rigidity are not equally distributed in the different layers. linked to vocal training. A descending glissando causes a symmetrical phenomenon (Figure 3). During the transition from mechanism M1 to mechanism M2. These events characterize the transition from mechanism M2 to mechanism M1.30 This is significantly lower for men than for women. as a characteristic of the switch from mechanism M1 to mechanism M2. which explains the fall in fundamental frequency. the heterogeneities of the structures may induce a decoupling between these layers in the midst of the vocal fold. of the vibrating mass. however. -. from a biomechanical point of view.10 one cannot rule out the possibility that a resonance phenomenon is added to this purely biomechanical phenomenon. 2007 . Ascending glissando: acoustic signal. may depend on the level of vocal training. This time the vibrating mass of the vocal fold becomes more significant. Male subject. The pitch at which the transition is activated is a reduction in the thickness of the fold. We could infer that. this reduction also leads to an increase in frequency. and this leads to a thinning of the vocal fold and an abrupt reduction.4 When the frequency increases. when a critical point is reached. Finally. EGG. In this case. Journal of Voice. ARTICLE IN PRESS 4 BERNARD ROUBEAU ET AL weak levels of energy. Descending glissando: acoustic signal. These different densities have important con- sequences for the nature of vocal fold vibration. and causes the jump in frequency. The pitch of the transition from one mechanism to another and the extent of the accompanying fre- quency jump have been measured on the EGG sig- nals from database DB1. Male subject. Hirano’s research has established links between the histological heterogeneities of the different layers constituting the vocal fold and their densi- ties. The choice of productions as poorly controlled as FIGURE 2. This phenomenon does not depend on the gender of the subjects nor on their level of vocal training.

this gap between the pitches of the change of mech.5 2 cending production occurs at a higher pitch than during a descending production. ARTICLE IN PRESS LARYNGEAL VIBRATORY MECHANISMS 5 much for the ascending glissandos and for the de. This phenomenon is probably related The magnitude of the frequency jump is pre.5 semitones) (4. and maintains its current state of equi. depending on the direction of the transition. sando (ascending or descending) and on the gender of the subjects. with women. to the morphological differences characterizing sented in Table 2 depending on the type of glis. It occurs at the same fre- quencies for both genders. which means that in this zone of the vocal range. nal is a criterion for identifying the change of librium. TABLE 2. is illustrated in Figure 5. The amplitude ratio is greater during the transi- anism during an ascending production and during tion M2–M1 than in the other direction but this a descending production confirms the known notion difference is more accentuated with men than of partial overlap of the range of the mechanisms.29 This difference is much less than Ascending Descending one octave (Table 1). the overlap zone of the mechanisms is considerable (one octave on average). Similarly to a hys- teresis phenomenon. 2007 . No. -.9 semitones) FIGURE 4. range of the overlap. as in the M2–M1 direction. Pitches of Transition M1–M2 and M2–M1 During Glissandos Ascending Descending Glissando Glissando Transition from Transition from M1 to M2 M2 to M1 Male subjects 238 Hz—Bb3 195 Hz—G3 (3.11.6 semitones) Female subjects 312 Hz—Eb4 279 Hz—C#4 (4. Frequency Jump in Semitones During the scending glissandos. Vol.11 is probably more tion point has been carried out and analyzed marked when producing glissandos than when pro. The ratio of amplitudes at the transi- also been noted by Svec et al. The border between the Glissando (M1–M2) Glissando (M2–M1) mechanisms is mobile. female and male voi- ces have the same possibilities of choice of produc- tion mechanism.25. the jump in frequency is greater with men than women. As illustrated in Figure 4 and Table 3. This constancy phenomenon. a point already made by several authors. the system thus seems to delay The abrupt change of amplitude of the EGG sig- the rupture.9. Journal of Voice. Average voice range profiles for men and women voices in each of both mechanisms. Finally. in agreement with previous Transition M1–M2 and M2–M1 studies. which has mechanism. In the M1–M2 direction. TABLE 1. and it ducing scales which are more controllable. female and male larynges. Thus the transition from Male subjects 5.8 semitones) (6. -. Representation of the The standard deviations are given in parenthesis.29 The range of each mechanism and the range of the overlap phenomenon have been studied on data- base DB2.5 4 mechanism M1 to mechanism M2 during an as- Female subjects 1.

frequency. women. The frequency range overlap of mechanisms M1 Both phases of the frequency perturbation can be and M2 allows the production of sounds of the explained by the mechanical phenomenon of de- same pitch in one or the other mode of production. men. The switch from mechanism M2 to mechanism To examine this type of production.5) M1 to mechanism M2. witnesses the modification of the vibrating W Mean limits D3–G#4 G#3–C5 G#3–G#5 structures typical of the switch from mechanism Extent in semitones 19 (3) 30 (5. the whole EGG analysis of the M1–M2 transition during of the phenomenon lasting no longer than 100 sustained sounds milliseconds. is almost instantaneous. which of mechanism without interruption of the produc. No.13 then by Van Deinse31 causes the upward frequency jump. W. Average values (standard deviation). The change of mechanism on a constant pitch but A pilot study23 shows that the amplitude of the with varying intensity has allowed Vilkman et al28 frequency jump varies in relation to the intensity of to determine precisely the concept of critical mass the production at the time of change of mechanism. This singing task was required in database DB1. The transition from M1 to M2 can be detected by a mod- FIGURE 5. as well as a reduction in amplitude of the Extent in semitones 29 (5. Vol. The amplitude zone of both mechanisms M1 and M2 is explored of the EGG signal increases and the fundamental for a given subject. besides possible modifications in tim- Mechanism Mechanism bre. although tion. variation of the fundamental frequency.4) 26 (6. The reduction in amplitude is related to a very Abbreviations: M. A change of the EGG wave shape can be ob- M Mean limits Eb2–F#4 F3–F5 F3–F#5 served. On this note. in the ‘‘chest register’’. then change the mechanism without interrupting sound production. Range of Mechanisms M1 and M2 and of During the switch from mechanism M1 to mech- Their Overlap Zone anism M2. the common M1 involves inverse modifications. This decoupling. already mentioned by Garcia.1 has been used allowing reduction of the vibrating mass.8) 15 (5. Journal of Voice. abrupt frequency jump followed by a slower readjustment to the initial frequency. FIGURE 6. The duration production in one mechanism. he/she must start the adjustment to the original frequency. -. ARTICLE IN PRESS 6 BERNARD ROUBEAU ET AL TABLE 3. then the recov- to analyze the modifications in timbre that occur ery under neuromuscular control of the initial during the switch of registers from chest to falsetto.9) 17 (6.5) signal. Sustained sound with change of laryngeal mech- anism. coupling of the tissue layers of the vocal fold Sustaining a sound at a constant pitch with a change during the switch in the direction M1–M2. 2007 . Amplitude variation of the EGG signal during ification of the amplitude of the EGG signal and an abrupt transitions between laryngeal mechanisms M1 and M2. Male subject. first and described by Large. -. A note within this zone is given frequency undergoes an abrupt fall followed by a re- to the subject.13.31 we can also observe important and abrupt M1 M2 Overlap modifications of the EGG signal (Figure 6).

known in an explored in the literature. folds. abrupt fall in the frequency (Figure 8). EGG analysis of M2–M3 and M0–M1 The origin of the discontinuity M2–M3 is not transitions completely identified. Vol. and its characteristics empirical way by singers that the amplitude of are still not well known. As for the transition M1–M2. with a much-reduced vibrating part compared ily understand that perceptual or acoustic analyses with mechanism M2. an abrupt reduction FIGURE 7. Female subject. can be ‘‘negotiated’’ in dif. such as the the laryngeal vibrating system. tion.32 acoustic spectrum does not particularly decrease during the switch to mechanism M3. This observation is confirmed alone are insufficient to identify and authenticate by the weak amplitude of the EGG signal. In the present case. One should note that the of the singer. ARTICLE IN PRESS LARYNGEAL VIBRATORY MECHANISMS 7 of this readjustment (60 milliseconds) is less than of the amplitude of the EGG signal and an upward the one observed in the preceding situation. that is. It is characterized by very thin and stretched vocal dicate its presence. the transition M2– Both examples presented in Figures 6 and 7 show M3 corresponds to a reduction in a component of that a physiological phenomenon. Spectrographic and EGG. See an important nisms M1. by a significant decrease in intensity. jump in frequency. of the highest-pitched sounds of the human voice. This transition in the high-pitch range of the voice presents the same characteristics as those separating mechanism M1 from mechanism M2. M2. Descending-ascending glissandos with mecha- damental frequency during the transition. In the opposite direction. From this example. This phe- In this case. the change of mechanism is preceded nomenon can be observed in men and women. Sustained sound with change of laryngeal mech- anism from M1 to M2 but without abrupt variation of the fun. one can still observe an abrupt in the amplitude of the EGG signal and an equally modification in the amplitude of the EGG signal. this time by reduc- change of mechanism. tion of the vibrating length. This observa. FIGURE 8. is not perceptible and only the EGG signal can in. macro- decrease of the acoustical amplitude. Journal of Voice. there When the switch from mechanism M1 to mech. Male subject. It is not to be excluded that Transition M2–M3 The production of an ascending glissando reach- ing the highest frequency zones allows one to ob- serve another transition separating mechanisms M2 and M3 (Figure 1). No. and M3. the change of mechanism but it is seldom used either in speech or in singing. The laryngeal mechanism M3 has been poorly tion would confirm the hypothesis. -.33 It allows the production the jump depends on the intensity of sound produc. scopic aspect. contrary to what is commonly described (Figure 8). is a symmetrical switch from mechanism M3 to anism M2 is not accompanied by a frequency mechanism M2: one observes an abrupt increase jump (Figure 7).33 a change of mechanism. -. one can eas. which allows access ferent ways depending on the esthetic objectives to the highest frequencies. 2007 .

4 mechanism. Transition from M1 to M0 (top) and from M0 to The EGG signal is modulated depending on the M1 (bottom) illustrated by EGG during a descending glissando contact between the vocal folds: the greater the (top) and an ascending glissando (bottom). and in singers ing ascending glissandos.35 A physiological hypothesis of participation of a lat- Transition M0–M1 eral compression of the vocal folds would explain The production of an ascending glissando start. Qualitative description and quantitative measures of a glottal cycle by EGG FIGURE 9. On the other hand. All the results pre- sented here come from the analysis of database DB3. contact. the physiological muscular data specific to this Mechanism M0 differentiates itself clearly from mechanism must be completed. The relationship between laryngeal mechanism and open quotient is also discussed. is used much Contrary to other mechanisms an overlap be- more in speech. and relatively little in singing. which may suggest that the con- tween M2 and M3 is equally mobile and that an tact surface (instantaneous maximum) between the overlap between the range of mechanism M2 and vocal folds does not change. the EGG signal also provides qualitative and quantitative information which makes it possi- ble to characterize the different laryngeal mecha- nisms. Vol. we find again the characteristic event LOCAL DESCRIPTION OF THE EGG of the change of mechanism which is the upward SIGNAL IN RELATION TO LARYNGEAL frequency jump during the transition from a given VIBRATORY MECHANISMS mechanism to that of a higher rank. It is also probable waveform shape changes radically (long contact that the everyday use of these mechanisms by phase in M0 with two maxima). the higher the amplitude of the EGG Journal of Voice. and cant but distributed in time. crease in a continuous way but shows an upward jump during the transition to mechanism M1. transition M0–M1. found in both men and women. its characteristics are examined within a given laryngeal mechanism. and nonsingers. These be that the folds’ contact surface remains signifi- hypotheses need to be investigated specifically. -. this mechanism can be As we can observe for the other transitions dur. One reason may trained subjects has an effect on this overlap.34 the other mechanisms. ing with the lowest voice frequencies (around of the inertia of the vibrator which allows the pro- 20 Hz) involves mechanism M0 (Figure 9). Male subject. At the level of a glot- tal cycle. the frequency cannot in. 2007 .36 tion where the period consists of a long closed It is important to note that. the importance of the vibrating mass and therefore. Unlike the other transitions. characterization of the laryngeal mechanisms is based on transition phenomena highlighted in the envelope of the EGG signals. It tween M0 and M1 does not seem to exist. as described for all phase and a very short open phase. the amplitude of the dition to the mechanical phenomenon which is EGG signal is not abruptly modified during the purely laryngeal. except corresponds to a specific type of laryngeal oscilla. quite well described. in rare male voices which are particularly low. Therefore. and the abrupt The preceding discussion has shown that the fall in frequency in the other direction. ARTICLE IN PRESS 8 BERNARD ROUBEAU ET AL it is triggered by a resonance phenomenon. -.4. After a brief description of the general EGG shape of a glottal cycle. the others through various physical parameters. No. It is probable that the frontier be. This duction of such low frequencies. the that of mechanism M3 exists.18.10 in ad.

opening of the glottis has little influence on The closing being usually faster than the the electrical signal. this corre- tween the vocal folds translate into marked peaks sponds to the moment when the vocal folds on the derivative of the EGG signal (DEGG sig. the contact varies very little and movement. opening. culation of various quotients relative to the open ically. -. begin to separate at their upper edges. We prefer the open main in contact over their whole length. they still give an indication relating to the general shape observed in a glottal cycle. tact of the vocal folds. the vi- brating mass is greater in laryngeal mechanism M1 than in mechanism M2. one can  Open phase (4–5): Once the vocal folds are thus observe the different phases of vocal fold separated. Cases of laryngeal mechanisms M1 and M2 As was mentioned in the previous section. Vol. Conversely. This often leads to a greater amplitude open quotient. which is characterized  Opening phase (3–4): The vocal folds separate by an increase of the contact between vocal folds. this phase is marked by a steep gra- dient in the EGG signal. Characterization of glottal cycles according to laryngeal mechanisms As there is a variation in the vibrating mass in- volved in phonation depending on the laryngeal mechanism with which the sound is produced. as this parameter is often related observed corresponding to the degree of con. the local minimum tal opening is detected by a progressive decrease of of which is associated with the moment of amplitude. Thus. In this work we are particularly the vocal folds connect on the whole of the interested in the open quotient. T0 represents the fundamental period and Oq the movement. moments allows quantification of the duration of the local maximum of which is associated a glottal cycle (fundamental period T0). These rapid variations in contact be. because of the participation FIGURE 10. the glot. Furthermore. of the EGG signals. a variation in the EGG signal can be (Cq 5 1 Oq). quotient to its equivalent. -. the closed quotient ever. Physiolog. How. this corresponds to the moment when and contact phases. between the duration of the open phase and the  Closed phase (2–3): The vocal folds then re. No. 1–2: closing phase. The derivative of the The detection of the glottal opening and closing EGG signal presents a very significant peak. 4–5: of the deep layers of the fold in glottal vibratory open phase. duration of the glottal cycle. there is Journal of Voice. 2–3: closed phase. nal22). glottal opening. Physiologically. Description of a glottal cycle by EGG.20 presented schematically in Figure 10: therefore. 2007 . to the glottal flow. the shape and amplitude of the EGG signals are related to the laryngeal mecha- nisms. The electroglottograph cannot give any infor-  Closing phase (1–2): The vocal folds connect mation relative to this phase as the degree of from their lower edges to their upper edges. If these parameters do not allow a formal identification of the laryngeal mechanism used. is detected on the EGG signal by a fast increase The EGG derivative presents a negative peak of the amplitude of the signal. 3–4: opening phase. The glottal closing. In a cycle of the glottal signal. more or less well marked. and the cal- with the moment of glottal closing. a relatively flat signal is observed. from their lower edges to their upper edges. defined as the ratio lower edge. ARTICLE IN PRESS LARYNGEAL VIBRATORY MECHANISMS 9 signal. there is also a variation of the contact area between the vocal folds.

the closing As was seen in the previous section.6 The contact phase is very long in relation to signal in laryngeal mechanism M1: a large ampli. a pronounced asymmetry. which of the lowest sounds of human phonation. Physiologically. No. In the case of laryngeal mechanism M2. the T0 corresponds to the repetition of a pair of cy- the EGG signals are reduced in amplitude and are cles. The vocalis Figure 12 shows glottal cycles in pairs. even if it can remain contracted. The opening and closing closing. the duration of a glottal cycle. FIGURE 11. or nonperiodic-impul- brating mass involved is very small. take part in the vibration. the shape of a glottal cycle is not necessarily phase shown by a strongly marked peak on the reproducible from one period to the next period. or multiple cycles (doubles and triples). is accompanied by a second DEGG glottal peaks have comparable amplitudes. Thus.38 vibrating only on their superficial part. Male and female subjects. -. Vol. As shown in Figure 11. -. the laryn- phase is shorter and the closure more abrupt in geal mechanism M0. DEGG signal.37 Journal of Voice. is charac- leads to a marked asymmetry of the EGG signals. this means that the first glottal much more symmetrical. As illustrated in Fig- tude. with a very low frequency. the first cy- muscle. does not cle being more marked than the next. the vi. that is much less pronounced. an abrupt closing ure 9. which allows the production mechanism M1 than in mechanism M2. one can observe a periodic glottal cycle.18. terized by very short vocal folds. very marked. very thick and Figure 11 illustrates the typical shape of an EGG lax. ARTICLE IN PRESS 10 BERNARD ROUBEAU ET AL a difference between the opening and closing Case of the laryngeal mechanism M0 phases of the vocal folds: in general. 2007 . the vocal folds sions. in M1 and in M2. In this case. Characteristic examples of EGG signals and their DEGG derivatives on two periods of the glottal cycle.

Visualization of the shape of an EGG signal and used in the literature.21.3 and 0. Vol.21. the open quotient nism is sometimes difficult to detect by EGG. The open quotient is also measured on the glottal flow signal. the EGG signal shows a very symmetrical shape.21 and it is thus more commonly FIGURE 13. which corresponds to the ratio be- tween the contact duration and the total duration of a glottal cycle. ryngeal mechanism. has values varying between 0. for a vocal production in M3. the glottal opening is reduced. No. in this mechanism EGG signal. 2007 . and the open quotient (Oq 5 1 Cq). Therefore. important differ- ences in the contact time within a glottal cycle can be observed.27 Thus. Therefore. which corresponds to the ratio between the open duration and the total duration of a glottal cycle. this glottal parameter here. ARTICLE IN PRESS LARYNGEAL VIBRATORY MECHANISMS 11 FIGURE 12. This laryngeal mecha. near that of mechanism M2. Two equivalent glottal parameters al- low these differences to be quantified: the closed quotient. for two glottal cycle periods. baritone and mezzo-soprano.8 in When contact is sufficient to be detected. -. -. Shape of EGG signals and their DEGG derivatives in mechanism M0. made on database DB3 have shown that the value and it is possible that there is no contact between of the open quotient varies depending on the la- the folds during phonation.22 Measurements stretched. or by detection of opening and clos- the vocal folds are very thin and very tightly ing peaks on the derivative. An example of glottal cycle detected by the EGG in mechanism M3 is presented in Figure 13. by a threshold-based method on the study of EGG signal changes. Cq. Journal of Voice. Laryngeal mechanisms and the open quotient As the vibrating masses in contact vary depend- ing on the laryngeal mechanisms. we shall refer to its derivative (DEGG) signal on two periods of glottal cycle.37 Case of the laryngeal mechanism M3 The open quotient can be very easily measured As mentioned in the section on macroscopic by EGG.

imprecision jump that is characteristic of the transition between from which some confusion could originate. It has. depending on the vocal intensity and the fundamental frequency of the sound produced. established on ho- open quotient will have lower values in mecha. mechanisms. whose nature differs essentially from another series FIGURE 14. No. Garcia’s definition of register is the on a glissando. and it accompanies the frequency one most cited. -.27 In the case of a transition skillfully When Manuel Garcia1 presented his work to the masked by the singer (no frequency jump). the The concept of mechanism. values. ARTICLE IN PRESS 12 BERNARD ROUBEAU ET AL mechanism M1. This jump in the open quotient is detected Nowadays. 2007 . to largely revise the notion of vocal register (which nisms where the open quotient can take on similar is still characterized by considerable confusion). and without any cultural notions. Vowel [a] sung by a counter-tenor at the same pitch (D4) in mechanisms M1 and M2. makes it possible There exists an overlap zone between mecha.39 he gave quotient jump is. detected. provided the following definition of vocal registers which that the values measured in M1 are well below is still nowadays the most referred to: ‘‘By the the characteristic values of mechanism M2. as shown in Figure 15. as illustrated gardless of gender or the level of vocal technique. nevertheless. mogeneous physiological observations (EGG) re- nism M1 than in mechanism M2. The transition between laryngeal mechanisms is Historical review of register and the sources often accompanied by a marked jump in the open of confusion quotient. Vol. therefore. the open quotient variations can. the open French Acade´mie des Sciences in 1840. and nism. -. nevertheless. in Figure 14. indicate a change of laryngeal mecha. produced by the same mechanical principle.5 in mechanism M2. At a same pitch.27 Journal of Voice. whereas they are always greater THE NOTION OF REGISTER REVISITED than 0. and homogeneous tones going from low to high. For word register we mean a series of consecutive a given singer.

mogeneity of sound to identify the registers. will be mechanical. whatever otherwise Henrich 41). strohbass. Similarly the term loft refers more or less explicitly on Garcia’s definition. and based of the sound produced.37 of tones equally consecutive and homogeneous pro. All the ing. Indeed. The terms heavy. The number of registers is variable subjects them. ARTICLE IN PRESS LARYNGEAL VIBRATORY MECHANISMS 13 FIGURE 15. As is of the register in question for the male spoken known.12 and quently of the same nature. Therefore. No. Several points of Garcia’s definition re. and classifying the different regis- tones belonging to the same register are conse. Musical phrase sung by a female soprano trying to use the laryngeal mechanism M1. from Paschke40). or even both. -. terms referring to different notions. and pulse refer to the impulses char- ‘‘homogeneous’’ character then becomes very am. the terms able to attribute the most diverse meanings. has led to much confusion. thin.’’ This leads normal and modal refer to the normality of the use to different interpretations of this definition. acterizing the perception of very low frequencies. The terms ‘‘nature’’ of the ‘‘mechanical principle. which must here be considered on cific to the author. thick. This definition. (singing teaching. -. The term falsetto refers to the acoustic quality From the XIX century to till date. but mogeneous tones [. however coherent going from the mirror laryngoscope applied with for this author. An overview of the main research car- may be the modifications of timbre or of the force ried out on registers is presented in Table 4.’’ Thus the notion of homogeneity depending on the authors. First trial: after a few notes. the choice of If this detail is not systematically and explicitly which reflects sometimes the type of approach stipulated. 2007 . The populations studied Garcia mentions a ‘‘series of consecutive and ho. success by Garcia to EGG. ters (for a review. biguous. both to the timbre produced and to the use of high Journal of Voice. and light refer to the main imprecise. the singer switches to mechanism M2. research on registers has mainly consisted in nam- duced by another mechanical principle. Thus we can find a group of the glottal level as the follow-up of his work shows.] whatever may be the modifi. numbering. ranging from two (most he evokes here seems only to refer to the principle frequently) to four. the different authors who refer to the ho. head. singers or nonsingers. physiology and vocal therapy. Second trial: the singer remains in mechanism M1 on the whole sentence.23 Miller. Vol. and acoustic). an acoustical phenomenon can originate voice. these populations are not systematically compared cations of timbre or of the force to which one to each other. Their designation is often spe- of production. are men and/or women. and the fry. see Roubeau. The ex- to which one subjects them’’ (translation of Garcia1 ploratory methods differ depending on the authors. The terms chest and head refer to the vibra- at the level of the larynx just as it can originate at tory sensations felt at the level of the chest or the the level of the resonance cavities. mainly the level and the morphological aspect of the vocal cords.

often evident in the literature. electromyography. and 4 Mechanisms for male and (singers and sustained tones EMG female (0. from which the author tackles the notion of register. frequencies. fusion. 2. Male and female Isoparametric tones Airflow rate 3 For male voice (chest. EGG and 2 Chest and head (trained glissandos. 1972 and falsetto) 2 For female voice (chest and middle) Gay et al3 1972 Male and female Sustained tones EMG 2 Chest falsetto Lecluse14 1977 Male Sustained tones EGG 2 Chest and falsetto Van Deinse31 1981 Male and female Sustained tones EMG 2 For male voice: chest. 1. Glissandos and EGG and acoustics 4 Mechanisms for male and singers sustained tones female (0.44 1840 Male and female Sustained tones Breath support 2 For male and female voice (poitrine and fausset-teˆte) 1855 Male and female Sustained tones Laryngoscope 3 (Poitrine. head.48 1970. This confusion first comes from the angle characterizes each one. -. mid. The terms flageolet. Male: singers and Sustained tones Acoustics and 2 (Modal and falsetto) 1973 nonsingers perception Large et al17.7 1972. Glissandos and EGG. modal. Sustained tones Tomography 2 For male and female voice singers (first and second registers) Van den Berg8 1960 Excised human larynx Direct observation (male) Hirano et al16 1970 Male and female: 2 Sustained tones and EMG 3 For male and female: For both gender scales Chest. and This proliferation of terms shows the great con- sifflet refer to the high pitch of the frequencies pro. ARTICLE IN PRESS 14 BERNARD ROUBEAU ET AL TABLE 4. and 3) Abbreviations: EMG. acoustics. -. when it duced. Chest. 3 (Pulse. head. and falsetto 4 For female voice: chest. fausset. and airflow rate Colton15. 2. flute. and falsetto (female) Hollien6 1974 Male Sustained tones Perception. little. assimilated to those that one can obtain with comes to the designation and identification of reg- these musical instruments. and photoglottography singer) 1 Trained scales Chest and falsetto (untrained 1 Untrained singer) Roubeau23 1993 Male and female.43. 2007 . and loft) X-rays. whistle. head (male). 1. Vol. head. No. and the tonality which isters. and whistle Kitzing5 1982 Male Sustained tones. and 3) nonsingers) Henrich21 2001 Male and female. acoustics.47. thin. Synthesis of the Main Studies Carried Out on Registers Since 1840 Authors Year Subjects Production Analysis Registers Mu¨ller42 1840 Excised human larynx Sustained tones Direct observation 2 (Chest and falsetto) (male) glissandos with strain variation Garcia1. and teˆte Battaille45 1861 Not precise Sustained tones Laryngoscope 2 (Poitrine and fausset) Benhke2 1880 Male and female Sustained tones Laryngoscope 2 For male voice (thick and thin) 3 For female voice (thick. Journal of Voice. and small) Husson and Djian46 1952 Male and female.

and the populations any change of laryngeal mechanism. We were able to show that the is subdivided into two registers: chest and belting. in singing distinct regions of vocal quality that can be main.6. As we will next demonstrate. ARTICLE IN PRESS LARYNGEAL VIBRATORY MECHANISMS 15 Some authors define registers from a perceptual mode of observation.49 pp. the primary registers are the chest we were able to explore three situations corre- and falsetto registers. registers middle. singers and nonsingers.’’ Miller. male term register has been used to describe perceptually and female. In the During each production. anism M1 to mechanism M2.48 Frequently. changes in resonance. by the homogeneity of the timbre all styles of vocal expression.12 pp. Richard Miller.) that may occur into the chest. EGG signal The confusion of the designation and identifica. Miller defines registers regarding discontinuity phe- Mechanisms and registers: a pilot investigation nomena as ‘‘readily perceptible (. and it is common to all subjects. -. No. It is physiologically of the sound produced (Titze. falsetto. applicable to all subjects and point of view. In this way to this author. 253: ‘‘the defined. the other registers depending sponding to the different combinations of change on the gender of the singer. signal. to describe signal with no modification of the EGG signal one or two mechanisms but never the entirety. and it does not correspond Therefore. tion of the acoustic signal. upper. and flageolet. 43). It seems that the Production c (voce finta to falsetto) is achieved concept of laryngeal mechanism may be used for without any noticeable change of the acoustic a better understanding of the laryngeal level of pro. neity of the observation tools. quency and the perceptible discontinuities. 44).43 whereas another category of authors registers. We focus sociated with the glottal source. and a jump in tion of the registers also results from the heteroge. Detected perceptually. Thus. where justments at the level of resonance of the vocal tract singers use different labels for registers. we performed an kinds: transitions of primary registers (‘‘first-order experiment which involved the production of sus- register’’31 and ‘‘natural register’’12) which are as- tained sounds with a change of register. frequency. mixed voice. The to a change of mechanism.12 pp. or frequency jump. that is. To establish a connection between this terminol- these phenomena of discontinuity can be of two ogy and the mechanisms involved. studied are mostly masculine. four registers involved only two different laryngeal the primary falsetto register is subdivided into three mechanisms. a modifica- ent phenomena is a source of confusion. of the fundamental fre- of ‘‘register’’ which are illustrated on Figure 16. the results cannot be generalized. without lyrical singing style singers. for example. the Production b (chest voice to mixed voice) is same tool has only been used to observe one single accompanied by a modification of the acoustic phenomenon or one single population. The term ‘‘chest’’ is. According whose terminology we have adopted. it is possible to es- Others define them from laryngeal configura. used to denote at the same Production a (chest voice to falsetto) shows time a primary and a resonance-dependent register. these concepts are mixed up. therefore. and in speech. In most cases. This exper- (‘‘other discontinuities associated with changes in iment was made with the collaboration of a singer the source and a group characterized chiefly by and a teacher of great renown. -. the primary register chest EGG were recorded. and others which here on the middle of the frequency zone covered involve both the modifications of the source and ad- by the laryngeal mechanisms M1 and M2. 2007 . tablish a correspondence between mechanisms and tions. as it proposes a homogeneous is characteristic of the switch from mechanism Journal of Voice. the EGG envelope duction of a register. but on the other hand. characteristic elements of the switch from mech- The use of the same terms to describe these differ. combines the vibratory and resonance aspects. tained over some ranges of pitch and loudness’’). (amplitude and wave shape). the acoustic signals and female case. Vol. and voce finta along an (ideally smooth) continuum of pitch or registers loudness’’ (Miller. It is in fact a change description of registers is mainly based on western of register such as singers describe it.

(a) Transition from chest voice to falsetto. and head for women.51 produced using the mechanism M2. This classification of registers is pre. whereas the falsetto register is compare it to the other modes of production. Journal of Voice. Pulse Normal Head (W) Flageolet Strohbass Chest Loft Flute Registers heavy. mixed fied as produced by a distinct mechanism. Very few studies geal mechanism M1. men. It is Classification of registers depending on laryngeal vibratory mechanisms TABLE 5. and pulse is identi- of sound quality. From top to bottom: envelope of acoustic signal. unlike what the jump in frequency). that we voice. No. loft. The register ensuring the lowest pitch produc- nism. it is Mechanism Mechanism Mechanism Mechanism thus possible to redistribute different register label- M0 M1 M2 M3 ing depending on the identified and authenticated mechanisms. M1 to mechanism M2 (change of amplitude and an intermediate laryngeal process. ARTICLE IN PRESS 16 BERNARD ROUBEAU ET AL FIGURE 16. Examples sung by R. Miller on sustained tones (C4) with switch of registers.50 indispensable to authenticate the change of mecha. Voce finta (M) The voix mixte (mid and middle voice) is most Head operatic (M) often produced in men in mechanism M1 and in women in mechanism M2. -. envelope of the EGG. women. W. known as fry. acoustic characteristics would suggest. and fundamental fre- quency curve. normal. The registers chest voice. strohbass. modal. Classification of Registers Depending involved in their production on the Laryngeal Mechanisms Involved On the basis of previously defined criteria. (c) transition from voce finta to falsetto. From these These three productions present the different essential data more subtle distinctions can be situations where the analysis of the EGG signal is made. Contrebasse Thick Thin Registers falsetto. shall call here mechanism M0. whether it is accompanied by a modification tions. It is not the result of Abbreviations: M. (b) transition from chest voice to mixed voice. -. Fry Modal Falsetto Whistle sented in Table 5. and chest Voix de Heavy Light Sifflet are produced in mechanism M1. thin Voix mixte (M) Voix mixte (W) Mixed (M) Mixed (W) and light are produced in mechanism M2. the register associated with the produc- tion of the highest-pitched sounds (whistle and fla- geolet) seems to correspond to mechanism M3. and voce finta are produced using the laryn. 2007 . thick. Finally. Vol.

The same mechanism can contribute to sung). quency range of the human voice. identified definition: by the analysis of transitions. but only if one observes great anisms used. between the terminology of registers and the As for the sung voice. scopic level. In this of observation of vocal productions. are classified amplitude of the EGG signal characterize the from low to high and numbered from zero to change of mechanism. Laryngeal vibratory mechanisms are the dif- from zero to three. the notions of registers vibratory mechanism applicable to the whole of the and of mechanisms are different. scopic level. ARTICLE IN PRESS LARYNGEAL VIBRATORY MECHANISMS 17 only very exceptionally described in men. it is mainly the esthetic laryngeal mechanisms associated with their context that directs the choice of one or other mech. mechanism can present great variations in geal vibrator. other. No. be they singers or nonsingers. only if nisms. but the frontiers between the registers does not necessarily It is possible to reduce the confusion that still correspond to those observed between mecha- exists in the domain of vocal registers. On the basis M2. These four laryn- geal mechanisms. the gender. for men’s and women’s voi. This approach has confusion corresponds more to the definition of allowed us to demonstrate. mechanism. 2007 . developing homogeneity or on the rigor in the use of each of the two terms: contrary contrasts of vocal timbre when several ‘‘mechanisms’’ and ‘‘registers’’. whereas on the micro. The changes in as. graded from low to high. with populations of singers and nonsingers. Vol. even though human population. men and women. On the macro. the exclusive or privileged use of one mechanism It is easy to establish a correspondence rather than another. four in all. three. because each mechanisms are involved. singers are accustomed to EGG study of transition phenomena should allow distinguishing sound categories regrouped under one to better assess the laryngeal characteristics the name of register. The modification of tim- The ranges produced by the different mecha. The In their vocal practice. in other words. To conclude. the abrupt modifications in These mechanisms. Macroscopic and microscopic The definition of Manuel Garcia. -. bre and the proprioceptive sensations with nisms are not contiguous but overlap each other. in accordance with laryngeal vibratory mechanisms than that of the some of the literature. ensure the production of the ferent configurations of the glottal vibrator whole vocal range. we offer the following tinct laryngeal vibratory mechanisms. it is easy to understand that the sometimes they may be considered as synony- parameters such as the production type (spoken or mous. This is the way several registers can be described as different approach we have adopted here at the laryngeal although being produced by the same laryngeal level. and cultural context will guide the production of several registers. mechanisms are identified thanks The frequency ranges produced by two neigh- to the study of the signal derivative and to the boring mechanisms can partially overlap each analysis of the open quotient. one has its own specific definition. quoted in analyses of the EGG signal recorded at the laryn- Historical review of register and the sources of geal vibrator level were made. or the proprioceptive sensations linked to their pro- duction. pect of the EGG signal are directly linked to the The sounds produced by one and the same changes in mechanical configuration of the laryn. based on the acoustic qualities of this mechanism. the existence of four dis- registers. Journal of Voice. Based on this physiological notion of laryngeal of these definitions. production. the classification of registers one defines with precision the mode and nature does not cover exactly that of mechanisms. timbre and intensity. -. that allow the production of the entire fre- ces. which they are associated contribute to the particularly in the case of mechanisms M1 and determination of the registers. This notion of register juxtaposes itself to CONCLUSION that of laryngeal vibratory mechanism.

Hirose H. Girard GT. J Acoust Soc Am. The Mechanism of the Human Voice. Behnke E. Vocal intensity in the modal and falsetto regis.12:131–161. 2004. Glottal open quotient in singing: measurements and corre- 10. teristic leap interval between chest and falsetto registers. Etude de la source glottique en voix parle´e et REFERENCES chante´e: mode´lisation et estimation. 1999. Fabre P. Roubeau B. chest and falsetto registers in voice: data from living and 28. and Physiological Aspects of Voice Mechanisms. Hollien K. 1880. rupture.9:66–73. perception [The`se de Doctor- de´mie des Sciences en 1840. J Acoust Soc Am. 1991. [Laryngeal registers as The Netherlands: University of Groningen.25:62–70. falsetto and varying intensity phona- dynamic. d’Alessandro C. Measurement of charac- studies in the theory of the singing voice [PhD thesis]. Henrich N. Curwen & Sons. Folia Phoniatr. -. Schutte HK.and electroglottographical recording of Phoniatr. J 3. Titze IR. Orsay: Universite´ Paris-Orsay. Vocal Fold Physiology. Identification of the ‘‘shift’’ be.. the use of the derivative of electroglottographic signals Warwick Lane. Hollien H. Phoniatr. Vocal mechanisms in singing: laryngological singing. No. -. register. On pitch jumps between 1417–1430. Van Deinse JB. J Acoust Soc Am.23:66–78. Control of Folia Phoniatr.39:280–289. 1982. and phoniatric aspects. Utrecht. A framework for the study of vocal registers. tween vocal registers. ceptual. Brown WS Jr.117: 11. Henrich N. Vennard W. 1993. Strome M. aero. 2002. Miller DG. Brown WS.24:337–344. 6. 165. Large JW.29:200–205. dam: Drukkerij Elinkwijk.4:155– 16. J Voice. 21. Sawashima M. Krishnamurthy AK.81:401–409. Schutte HK. and J Voice. and perceptual properties of modal tions. Fold in Register Control. Svec JG. Arabia C. Inc. Photo. 12. Chen Y. ARTICLE IN PRESS 18 BERNARD ROUBEAU ET AL Acknowledgments: We are very grateful to Evelyn 19. Coleman RF. 1998. phie de haute fre´quence. Ohala J. Van den Berg JW. Per- 7. vocale. The Role of the Layer Structure of the Vocal thesis].14:769–775. 1995. Kitzing P. 1974. 1977. Acoustic. Hirano M. Arabia-Guidet C. Alku P.33:37–50. Folia Phoniatr Logop. Duverger. Henrich N. Roubeau B.2:51–69. Iwata S. ramme par me´canisme larynge´. 1982.115:1321–1332. On 2. sociated with modal. J Phon. d’Alessandro C. 2nd ed. Me´canismes vibratoires larynge´s et controˆle Ann Otol Rhinol Laryngol. Castellengo M. 1970. 1969. vocal intensity. Aerodynamic study. Large J. 2007 . 22. Lecluse FLE. Folia Phoniatr. McGlone RE. 1988. for characterization of nonpathological phonation. 1970. Registers in singing: empirical and systematic 29. Paris: Universite´ Paris. Instructional Film. 1973. An electromyographic inves. Acoust Soc Am. Roubeau B. Shipp T. 9. vocal-fry phonation. Bodin P. Burger D. ou les 15. In: Gauffin J. Laukkanen AM.16:8–19. Bull Acad Natl Med 1957. 1981. register transition in singing. McGlone RE. 1972. 32. 31. San Diego. 1036. 2004. 1993. Hirano M. Me´moire sur la voix humaine pre´sente´ a` l’Aca. Rotter. Folia Phoniatr. range. Phone´tog- Voice production: the vibrating larynx. Castellengo M. 1971. 2000. Svec JG. 1973. Groningen: University of Groningen. 1847. raphy of the intrinsic laryngeal muscles during phonation. 35. Colton RH. Vocal fold vibratory 17.56:321–333. J Voice. 26. Blomgren M. Regulation of register. Shervanian CC. ornementation. Miller DG. 8. et e´lectroglottographiques. Chevrie-Muller C. Chevrie-Muller C. J Voice. Folia Phoniatr. physiologic. mesures acoustiques 1. Crit Rev Biomed Eng.34:234–241. 2001.66–69. Roubeau B. 1960. Vox Humana: University of 24.22:385–396. 34. Vol. London: J. 25. Continuite´. 1972. Vocal fold length as- 18. 6. 12th ed. Hirano M. The primary female patterns of pulse register phonation.2:125–143.103: 36. 50–62. J Voice.22:1–20. 1971. Schutte HK. Childers DG. A critical review of elec- troglottography. 1988. Miller DG. 1987. Vocal-fold collision excised human larynges. Elektroglottografie [thesis]. Doval B. Some physiologic correlates of 2649–2658. Folia Phoniatr. Registers. Hammarberg B. the laryngeal vibratory pattern during different registers. Ng ML. J Acoust Soc Am.106: mass as a differentiator between registers in the low-pitch 1523–1531. laryngeal vibration in register change. Physical definition of the flageolet tigation of laryngeal intrinsic muscles. Electromyog. pitch and intensity of voice. Ragot M. Castellengo M. Doval B. Electro- Jyvaskyla. and vocal fry registers. Folia Phoniatr. Folia Phoniatr. 2005. Vilkman E. USA: Singular Publishing Group. 20. Miller DG. Journal of Voice. Von Leden H. ters. Gilbert HR. 13. neuro-musculaire de la fre´quence fondamentale [PhD 4.25:39–61. Hollien H. lation with laryngeal mechanisms. Paris: Imprimerie d’E. shown in the voice range profile]. 23. Gay T. Un proce´de´ e´lectrique percutane´ d’inscription de Abberton and Adrian Fourcin for the kind help with l’accolement glottique au cours de la phonation: glottogra- the English text. J Speech Hear Res. Hollien H. eds. Vennard W.46:1033– 27.7:206–212. Castellengo M. fundamental frequency. 33. 1991. glottographic study of the changes of voice registers. Folia 5. 1985. 14. Colton RH. Acoustic study of register equalization in 30. Cahiers de musiques traditionnelles. Acoustic. Garcia M. Spectral characteristics of the modal and bons usages de la transition entre deux modes d’e´mission falsetto registers. at]. On vocal registers. Folia 1977.2:183–194.

Englewood 41. Logoped Phoniatr Vocol. J Acoust Soc Am. 44. Vol. London: The Royal mechanisms. Castellengo M. Part One (Translation of Garcia. ARTICLE IN PRESS LARYNGEAL VIBRATORY MECHANISMS 19 37. New Jersey: Prentice Hall.24:19–29. Is voix mixte. 45. Stockholm. Mirroring the voice from Garcia to the present Cliffs. NATS Bull. Nouvelles recherches sur la phonation. 2003. Paris: Asselin. nism? In: ISMA. 1984. 2007 . Translation complemented by Fourcin ory of vocal registers. 1994. Large JW. 2007. terns of the vocal folds during pulse register phonation. Henrich N. Whitehead BH. On the use of 43. -. a clear-sighted observer 46. Garcia M. Sweden. Paris: chez Baille`re. Japan.31:3–14. 1861. -. Henrich N. 1971. The male operatic head 40. Principles of Voice Production. vocal fry phonation. No. Roubeau B. Metz DE. Vibratory pat. 1847). 1855:399–410. V. A. Folia Phoniatr. 2006. Observations on the human voice. New York: Da 48. J Radiol of human voice production. In: Proceed- electroglottography for characterisation of the laryngeal ings of the Royal Society of London. Von Leden H. les sens et les faculte´s 51. Battaille ChA. 47. Large J. 2004. 1984. J Acoust Soc Am. Paschke DVA. day: some insights into singing voice registers. Garcia M.29:18–40. Towards an integrated physiologic-acoustic the- Capo Press. 49. la voix. the Phoniatr Vocol.75:1293–1297. Henrich N. Tomographie et phonation. 2005. Castellengo M. Conference. Regulation of vocal intensity during intellectuelles.49:1905– traduit de l’allemand sur la 3rd ed. Masson.33:127–135. 2003. Mu¨ller J. Logoped 50. la parole. Presented at: Stockholm Music Acoustics Society. an independent mecha- expe´riences sur les diverses classes d’appareils nerveux. Husson R. Whitehead RL. vocal technique used to smoothe the transition across the 42. 1972. Nara.30:163–170. Titze IR. personal communication. par A. Journal of Voice. 1861. Iwata S. Brown WS. 39. Physiologie du syste`me nerveux ou recherches et two main laryngeal mechanisms. humaine. Manuel Garcia Jr. Djian A. 1907. Chuberre B. 1855. Paris: Acade´mie des Sciences. Complete Treatise on the Art of Singing: register versus falsetto. 1840. 1972. Observations physiologiques sur la voix 38. Deux tomes. Jourdan. Castellengo M. Murry T. Electrol. les mouvements.