the british voice association

About the Association

Van Lawrence Prize 2012


Analyzing the female "middle register" with EGG wavegrams – (abstract)

Christian T. Herbst, Laboratory of Bio-Acoustics, Dept. of Cognitive Biology, University of Vienna, Austria
Jan G. Švec, Department of Biophysics, Faculty of Science, Palacký University Olomouc, Czech Republic
J. Schlömicher-Thier, Austrian Voice Institute, Neumarkt am Wallersee, Austria
W. Tecumseh Fitch, Laboratory of Bio-Acoustics, Dept. of Cognitive Biology, University of Vienna, Austria


Introduction: The choice of singing register and the degree of vocal fold adduction are two concepts that are not easily discriminated by inexperienced singers. This is particularly true for the mid range (pitch C4 – C5) of untrained female classical singers, where adducted falsetto, the desired sound quality in this range, is rarely observed. As an underlying physiological principle, vocal fold adduction can be separately controlled by (a) cartilaginous adduction, i.e. the adduction of the posterior glottis via the arytenoids (controlled by the singer with the degree of "breathiness"/"pressedness"); and by (b) membranous medialization through vocal fold bulging (controlled by the choice of vocal register, i.e. chest vs. falsetto) [1].

Methods: In this study, singing exercises for adjusting adductory settings (cartilaginous adduction vs. membranous medialization) in the female mid-range were performed by 14 female classical singers, both trained and untrained. Phonation in aBducted falsetto (FaB), aDducted falsetto (FaD), aBducted chest (CaB) and aDducted chest (CaD) was monitored by acoustic recording, electroglottography (EGG) and laryngeal endoscopy. EGG wavegrams [2] [3], a novel method for displaying EGG signals, were utilized for data analysis. A blinded, randomized evaluation of the dEGG wavegram data (based on the 1st derivative of the recorded EGG signals) was performed ten months after data acquisition. Data was rated by (a) occurrence of vocal fold contact; (b) precision of contacting/decontacting events; (c) duration of vocal fold contact per glottal cycle (i.e. contact quotient); and (d) "fuzziness" (i.e. visual noise) in the wavegram appearance. To verify the wavegram ratings, the following averaged parameters were calculated from the original EGG data for each phonation sample: EGG signal amplitude (no automated gain controller was employed in the hardware); harmonics-to-noise ratio of the EGG signal (HNREGG); EGG contact quotient (CQEGG).

Results: EGG wavegram data revealed distinct differences between the targeted phonation types for each individual. The observed differences established themselves as (a) presence/absence of vocal fold contact; (b) duration of vocal fold contact per glottal cycle; (c) changes in the overall EGG signal amplitude; (d) distinctness of opening/closing events; (e) noisy appearance of the wavegrams.

Results in detail: A general agreement between subjective ratings and calculated analysis data was observed. The more adducted phonations had a larger HNREGG and a lower "fuzziness" rating than their abducted counterparts. The same trend was observed for chest vs. falsetto register. The contact quotient as estimated from the wavegrams followed the same trends as the videokymographic closed quotient, which was determined in a previous study [1]: Both increased adduction and the change from falsetto to chest resulted in an increased contact quotient. In eight of the 14 subjects, however, the EGG signal assumed a quasi-sinusoidal shape superimposed by noise, and was not suitable for CQEGG calculation. In these cases, no vocal fold contact was observed in the respective wavegrams. In average, the EGG signal amplitude of the abducted falsetto phonation type was at least 15 dB lower than that of all the other phonations.

Discussion: The singing exercises employed in this study enabled all participating subjects to produce all four targeted phonation types (including adducted falsetto, which is the desired phonation type in the female classical mid range). The gathered data suggests that EGG wavegrams are well suitable for distinguishing the four analyzed phonation types individually in each subject. Consequently, EGG wavegrams were shown to be useful for assessing the underlying physiological adjustments in the larynx, i.e. cartilaginous adduction and membranous medialization. Inter-subject data variation indicates that the degree of vocal fold contact in singing is not only influenced by physiological control of laryngeal configuration, but also by the individual singer's anatomy.

Conclusion: EGG wavegrams proved to be useful in documenting changes of both singing register and glottal adduction. Untrained singers can learn to increase glottal adduction when singing in falsetto register in their mid range, thus avoiding to employ the chest register above their zona di passaggio (i.e the primary register transition around D4). In this context, EGG wavegrams promise to be a powerful tool in both the singing studio and in clinical settings, particularly when utilized as a real-time feedback technology.



[1] C. T. Herbst, et al., "Membranous and cartilaginous vocal fold adduction in singing," J Acoust Soc Am, vol. 129, pp. 2253-2262, 2011.

[2] C. T. Herbst, et al., "Electroglottographic wavegrams: a technique for visualizing vocal fold dynamics noninvasively," J Acoust Soc Am, vol. 128, pp. 3070-8, Nov 2010.

[3] http://en.wikipedia.org/wiki/Electroglottographic_wavegram



Back to 'Awards & prizes'

Calendar icon What's on:
bva Seminars and Events plus much more

Button iconJoin us Now!
Great savings on event tickets plus other benefits

Newsletter (cover) our newsletter
Published 3 times a year – Download a sample

Leaflet (cover) free voice care leaflets & information – download here

Give as you live (logo) Help our work by donating while you shop