Reinaldo Kazuo Yazaki 1 , Paulo Pontes 2

1 UNIFESP-EPM, Otolaryngology- Head and Neck Surgery, Division of Laryngology and Voice, Sao Paulo, Brazil

2 UNIFESP-EPM, Otolaryngology-Head and Neck Surgery, Sao Paulo

Introduction: Laryngeal respiratory posture is not a well understood event and glottic behaviour during breathing in singing is not known. It is stablished that both thyroarytenoid and cricothyroid muscles are electromyographically associated to glottal abduction, during more complex phonatory tasks as in speech. In singing, during low sustained emissions, there is a predominance of thyroarytenoid muscles action, with shorter glottal membranous portion and, in high emissions, there is predominance of crycothyroid muscles action, with longer membranous portion, as evidenced by electromyography. Will the glottal length change during breathing just after these distinct emissions?

Objective: Analize comparatively glottal configuration during inspiration after low and high vocal emissions in singers, considering glottal proportion and glottal opening angle.

Method: Choral singers from our university were submitted to telelaryngoscopy. Laryngeal images were captured during inspiration, after a sung sustained emission of the vowel /E/ as in "bed", in a high and low manner. In each of them, lengths of phonatory and respiratory portions of the glottis were measured. From the mathematical relationship of them, glottal proportion was calculated for the high and low emissions, which was compared between singers and between the two groups of emission. Glottal opening angles were also measured.

Results: Glottal proportion was signifficantly smaller during inspiration after the low emissions (p=0.01). The opening angle was smaller after high-pitched emissions (p=0.02).

Conclusions: There is difference in glottal behaviour during inspiration after low and high emissions in singers. After high vocal emissions, phonatory portion is proportionally longer than after low ones, with less abduction.

 

Key-words: Proportion, Larynx, Voice, Music, Laryngeal muscles, Stroboscopy

Sylvain Lamesch 1 , Boris Doval 1 , Michèle Castellengo 1

1 LAM-IJLRA, , Paris

The human larynx is known to be able to vibrate according to four different configurations called laryngeal vibratory mechanisms. In the western classical singing, only the two main laryngeal mechanisms M1 and M2 are used.  Although there is a great overlapping area in terms of frequency and vocal intensity between M1 and M2, the transition between them is often a major inconvenient for classical singers, who seek a homogenous vocal quality among their whole tessitura. To sing in this overlapping area, they have developed two different strategies: either using only one laryngeal mechanism, or using both and getting trained to smooth the transition.

As a consequence, some singers are trained to use both laryngeal mechanisms, and some others are not. The aim of this paper is to explore the influence of these strategies on the spectrum of vocal productions and glottal vibration. What are the differences between M1 and M2 productions at the same frequency and vocal intensity? Do these differences depend on the vowel? On the vocal training? On gender or vocal category?

21 singers (8 females and 13 males) were recorded, singing crescendos and decrescendos on different notes in M1 and M2, on the vowels /a/, /i/ and /o/. The singers were all advanced amateurs or professional singers. Sound and electroglottographic signals were recorded. The study focuses on the repartition of the vocal energy among the spectrum in the overlapping area. Two frequency bands were defined, depending on the gender: the singer’s formant frequency band (FB2) from 2 to 4.5 kHz for males and from 2.4 to 5.4 kHz for females, and another one (FB3) from 4.5 to 8 kHz for males and from 5.4 to 9.6 kHz for females. Open quotient values were also computed.

The results confirm that there are systematic differences in the averaged open quotient in the overlapping area between M1 and M2 productions. Females and counter-tenors keep an energy ratio in the singer’s formant region very close in M1 and M2 on the vowels /a/ and /i/. But on /o/, a 6 dB difference was observed. For the other male singers, who are used to sing only in M1, this energy ratio is 6 dB higher in M1 than in M2 on the three vowels. The energy ratio in FB3 was globally lower in M2 than in M1. Finally, singers who are trained to sing in both mechanisms are able to obtain the same singer’s formant level at a given frequency and intensity, especially on /a/ and /i/.

Christian Herbst 1 , Jan Svec 1 , Qingjun Qiu 2 , Harm Schutte 2

1 Palacký University Olomouc, Laboratory of Biophysics, Dept. Experimental Physics, Olomouc

2 Cymo B.V., , Groningen, The Netherlands

While it has been recognized that glottal adduction is an important parameter in speech, relatively little has been known on the adjustment of the glottal adduction when changing the voice quality in singing. Our previous pilot data on a single subject suggest that the cartilaginous and membranous parts of glottis play different roles in singing – while the membranous part is expected to play an important role for switching between the chest and falsetto registers, the cartilaginous part is expected to play a primary role for adjusting the sound quality within the desired register. The goal of this study was to design singing exercises that enable both trained and untrained singers to independently manipulate cartilaginous and membranous glottal adduction and to verify these exercises on a set of subjects laryngoscopically.
 
A baritone, who was previously found capable of independently manipulating the cartilaginous and membranous glottal adduction, served as an instructor in this study. 6 female and 6 male subjects, singers and non-singers, were asked to imitate the instructor in producing 4 phonation types, i.e. (A) ‘naïve falsetto'; (B) ‘quality falsetto’; (C) ‘light chest’; and (D) ‘heavy chest’, at a pitch located within the range of the chest/falsetto register transition (C#4 to F4). In order to maintain the desired register, the target notes for chest and falsetto were reached by singing an ascending and descending, respectively, scale of five notes starting in the desired register. The subjects were asked not to “blend or mix the registers”. The phonation was monitored by videostroboscopy, videokymography (VKG), electroglottography (EGG) and audio recording.

The results showed distinct laryngeal configurations and vocal fold vibration characteristics for the four phonation types. As expected, all the subjects showed a less adducted posterior, i.e. cartilaginous, glottis in phonation types A and C than in phonations types B and D. Changes in the membranous part of the vocal folds were reflected in videokymographic imaging which revealed that the chest phonations, as compared to the falsetto phonations, had larger mucosal waves, sharper lateral peaks and longer closed quotient.

The findings indicate that the singers succeeded in independently manipulating the membranous and cartilaginous adduction of the glottis. Individual control over these two types of glottal adduction is expected to be a key factor for the experienced singer to create different vocal timbres. The designed singing exercises were found useful in training the subjects for achieving this goal.

Hanny van Lankeren 1 , Donald Miller 2

1 Zingen Zien en Horen, Voice Research, Wageningen

2 Cymo, B.V., Voice Research, Groningen

A shortage of male tenors in amateur choirs in Holland, forces choir leaders to supplement the tenor group with low female alto singers. Unfortunately, their contributions are not always satisfactory. It is said that female singers lack the low range, produce a different sound quality and are unable to sing as loud as males do. This raises the question whether we can select and train female tenor voices to become valuable members of the tenor group. Many mezzo-soprano and alto singers are able to extend their lower range considerably, suggesting that in these cases, size of the larynx is not a limiting factor. In addition, Miller (2008) recently demonstrated that male and female voices can adopt identical resonance strategies within the same vocal range. Finally, the limited relative loudness of female voices due to lower Closed Quotients (CQ) of the vocal folds might not be relevant in this particular case, since the big volumes sung by a robust tenor on his highest operatic notes are generally not required of choral tenors.

A short study was conducted to compare voice characteristics of 4 male and 4 female subjects, all active members of the tenor section in an amateur choir. The protocol included assessment of their vocal range, sung scale in A-flat ( 208-415 Hz) on vowels [i] and [a], in soft and loud phonation and a messa di voce on both vowels at the bottom and top of the scale. Audio and EGG signals were processed by means of VoceVista voice analysis software.

 

Results show that closed quotients depended on the individual singer, the sung quality and the relative loudness. If the subject managed to keep up a modal voice source to the top note in the tenor range (B-flat = 466 Hz)  CQ's were raised to more than 60 % in both male and female voices. Presence and strength of a singer's formant varied widely with the individual singer, but often improvement was achieved in response to direct bio-feedback. In all subjects, the most frequently observed resonance strategy on the vowel [a] was the first formant enhancing the third or second harmonic.

 

These findings indicate that when singing with identical voice qualities in the same range, voice characteristics in male and female tenors are comparable. This also suggests that a simple and effective training method can be developed, supported by feed-back through voice analysis software, that will help female as well as male singers to become better tenors in any amateur choir.

Gillyanne Kayes 1 , Johan Sundberg 2 , Graham F. Welch 3

1 Vocal Process & Institute of Education, Culture and Pedagogy, London

2 Royal Institute of Technology, Music Speech and Hearing, Stockholm

3 Institute of Education, University of London, Arts and Humanities, London

Pitch range is an important aspect in all types of vocal art. Each singing voice may have a typical pitch range of maximum phonatory comfort. In Western Lyric singing, one aspect of vocal ‘fach'is a match between ‘comfortable pitch range' and song material. On the other hand in ‘CCM' singing, desirable expressive effects might well result from a mis-match between comfortable pitch range and song material.

 In this study six female singers were selected: all were experienced performers or teachers of ‘CCM' genres. Each singer performed a reading test and sang two songs from her performance repertoire. In addition each singer demonstrated her ‘comfortable pitch range' in singing and was interviewed about her experience of vocal effort needed to sing the songs. We compare the centre of gravity of F0 histograms of songs with the mean F0 in habitual speech, taking into account the effect of loudness change on mean F0. The difference is compared with singers' perception of vocal effort and expert listeners' perception of effort in performance.

Johan Sundberg 1 , Lide Gu 2 , Jiang Huang 2 , Ping Huang 2

1 KTH, Speech Music Hearing, Stockholm

2 Singing Voice Research Institute of China Conservatory, , Beijing