This page is maintained by Nathalie HENRICH (GIPSA-Lab), Cédric GENDROT (Laboratoire de Phonétique et Phonologie) and Alexis MICHAUD (Langues et Civilisations à Tradition Orale).
The tools offered for download below have in common the use of the derivative of the electroglottographic signal, in light of research discussed in Henrich et al. (2004).
In a nutshell: this figure illustrates the presence of alternating positive and negative peaks on the derivative of the electroglottographic signal. The positive peak on the DEGG signal corresponds to the glottal closing instant (GCI), i.e. the instant when the vocal fold contact area increases with greatest velocity. The negative peak, which is the point at which the EGG signal falls most steeply, corresponds to the glottal opening instant (GOI), i.e. the instant when the vocal fold contact area decreases with greatest velocity.
Visualization of the closing and opening phases by simultaneous high speed and electroglottographic recordings. Vertical bar indicates the moment in time at which the visual image occurs with respect to the EGG and DEGG signals. This example corresponds to a nonpathological male phonation in mechanism 1, with f0=110 Hz. The EGG sampling frequency is 44 444 Hz and the high-speed camera sampling frequency is 3704 frames/s. Author: N. Henrich. The high-speed and EGG datafiles were kindly provided by a team of the Department of Phoniatrics and Pediatric Audiology of the University of Erlangen-Nuremberg (Germany).
The physiological correlates of the peaks observed on the DEGG signal were mainly studied by Childers et al., using simultaneous and synchronized measurements of EGG and DEGG signals, inverse-filtered derived glottal flow and glottal area measured from ultra-high speed cinematography (Childers et al., 1990; Childers and Krishnamurthy, 1985; Childers and Larar, 1984; Childers et al., 1983a,b), along with a theoretical study of the relation between the EGG/DEGG signals and the glottal contact area calculated from a physical model (Childers et al., 1986).
On the substantial basis of these studies, DEGG peaks may be considered as reliable indicators of glottal opening and closing instants; the latter being defined by reference to the glottal flow, as the instants when the flow starts to increase greatly from the baseline (opening) and decrease greatly to the baseline (closing).
The duration between two consecutive glottal closing instants corresponds to a fundamental period; its inverse gives the fundamental frequency of the voice. In the same manner, the duration between the glottal opening instant and the consecutive glottal closing instant corresponds to the open time. The open quotient can then be derived from these two measures as the ratio between open time and fundamental period.
In cases where the peaks do not
stand out clearly, the open
quotient cannot be calculated reliably; indeed, it can
be argued that in cases where opening takes place gradually and at
points in the glottis, resulting in double/multiple opening peaks, it
not make sense to talk of an "open quotient" at all.
A review of the methods based on EGG and DEGG signals for open quotient measurements is presented here :
- DEGG methods : threshold-based methods and correlation-based method
- EGG threshold methods
- EGG + DEGG method, a method that combines detection of positive peaks on the DEGG signal and an approximation of the opening instant on the EGG signal by thresholding,
Which of the software tools available for download here is suitable for you?
|name||author||type of method||expectations on periodicity of the data||environment||notes|
|peakdet||Michaud||DEGG||none||Matlab||Main research purpose:
investigation of phonation type
during rhymes (vowels and final consonants). Documentation available.
A version of this script without the interface for user verification of the results is available online from the GitHub repository COVAREP since 2014: click here to access it.
|peakdet2||Abadal & Recasens||DEGG||none||Matlab||Modified version of <peakdet>, by Sergi Abadal and Daniel Recasens (2009); main research purpose: study of the onset of voicing. An Instructions Manual is available here.|
|(variant of peakdet)||Brunelle et al.||DEGG||none||Matlab||Modified version of <peakdet>, by Marc Brunelle and his research assistant (2009); main research purpose: investigation of the phonation types used in the prosodic systems of Southeast Asian languages|
|decom||Henrich||DEGG||quasi-periodic (i.e. suitable for the singing voice rather than speech)||Matlab||Matlab function to be applied on one frame of DEGG signal|
|oq1praat||Gendrot||DEGG||none||Praat||Translated from <oq1>|
|quasi-periodic (i.e. suitable for the singing voice rather than speech)||Matlab||Matlab function to be applied on one frame of EGG signal|
interface for Matlab : MOQ
The variety of the tools reflects the differences in the data to be analyzed. In particular, N. Henrich's functions were developed for the singing voice. They assume a quasi-periodic signal. The others were developed for speech, with attention to rapid changes in voice quality: breathy voice, glottal constriction, laryngealization ("creaky voice"), i.e. phenomena which result in large changes in cycle length from one glottal cycle to the next. Peaks are detected individually by applying a threshold to the derivative signal.
- Singing voice, 4 speakers (2 males / 2 females) - N. Henrich, ©2000, LAM-LIMSI
Data used in :
- Vu Ngoc Tuan, C.,
and d'Alessandro, C. (2000)
"Glottal closure detection using EGG and the wavelet transform", in
Proceedings 4th International Workshop on Advances in Quantitative
Laryngoscopy, Voice and Speech Research, Jena, pp. 147-154.
- de Cheveigné, A., Kawahara, H. (2002) "YIN, a fundamental frequency estimator for speech and music", J. Acoust. Soc. Am. 111, 1917-1930.
- Michaud, A.; Vu-Ngoc, T. Glottalized and Nonglottalized Tones under Emphasis: Open Quotient Curves Remain Stable, F0 Curve is Modified. Speech Prosody 2004, Nara, Japan: 745-748 (2004).
- Michaud, A.: Final consonants and glottalization: new perspectives from Hanoi Vietnamese. Phonetica 61(2-3): 119-146 (2004).
- voice material from the VOQUAL'03 workshop (Geneva, 2003)
- audio and EGG files accompanying A. Michaud's PhD
Childers, D. G., Hicks, D. M.,
Moore, G. P., and Alsaka, Y. A.
(1986). “A model for vocal fold vibratory motion, contact
area, and the electroglottogram,” J. Acous. Soc. Am. 80,
Childers, D. G., Hicks, D. M., Moore, G. P., Eskenazi, L., and Lalwani, A. L. (1990). “Electroglottography and vocal fold physiology,” J. Speech Hear. Res. 33, 245–254.
Childers, D. G. and Krishnamurthy, A. K. (1985). “A critical review of electroglottography,” CRC Critical Reviews in Biomedical Engineering 12, 131–161.
Childers, D. G. and Larar, J. N. (1984). “Electroglottography for laryngeal function assessment and speech analysis,” IEEE Trans. Biomedical Engineering BME-31, 807–17.
Childers, D. G., Moore, G. P., Naik, J. M., Larar, J. N., and Krishnamurthy, A. K. (1983a). “Assessment of laryngeal function by simultaneous, synchronized measurement of speech, electroglottography and ultra-high speed film,” in Transcripts of the eleventh symposium : Care of the professional voice., edited by L. V. , New York : Voice Foundation), pp. 234–44.
Childers, D. G., Naik, J. M., Larar, J. N., Krishnamurthy, A. K., and Moore, G. P. (1983b). “Electroglottography, speech and ultra-high speed cinematography,” in Vocal fold physiology and biophysics of voice, edited by I. Titze and R. Scherer , Denver, CO : Denver Center for the
Performing Arts), pp. 202–20.
Henrich N., d'Alessandro C., Castellengo M. and Doval B., 2004, "On the use of the derivative of electroglottographic signals for characterization of nonpathological voice phonation", Journal of the Acoustical Society of America, 115(3), pp. 1321-1332
Howard, D.M.: Variation of electrolaryngographically derived closed quotient for trained and untrained adult female singers. Journal of Voice 9(2): 163-72 (1995).
Howard, D.M.; Lindsey, G.A.; Allen, B.: Toward the quantification of vocal efficiency. Journal of Voice 4(3): 205-12 (1990).
Rothenberg, M.; Mahshie, J.J.:
Monitoring vocal fold abduction
through vocal fold contact area. Journal of Speech and Hearing Research
31: 338-51 (1988).