Acoustic and electroglottographic (EGG) characteristics of tracheoesophageal speech of Cantonese

Abstract

This study investigated the acoustic and electroglottographic (EGG) characteristics of tracheoesophageal (TE) speech of Cantonese. Seven TE speakers and seven aged-matched laryngeal (NL) speakers produced sustained vowel phonation and passage reading. Both acoustic and EGG signals were recorded and analyzed using software programs Praat and Voce Vista. Results indicated that there was a significantly lower fundamental frequency (F0) for TE than NL speakers in passage reading. Significantly higher jitter, shimmer and closed quotient (CQ), and lower harmonic-to-noise ratio (H/N) values were associated with TE speech in both sustained vowel phonation and passage reading. Higher formant frequencies in sustained /i/ phonation were found for TE speakers. The findings appear to explain the perceptually hoarse, breathy and low-pitch voice of TE speech. Results were discussed in terms of higher position, greater tissue density, slower movement during closing phase and aperiodic vibration of neoglottis of TE speakers than the vocal folds of NL speakers.