Objective: Speakers and singers commonly use a lip trill production as a voice warm up and training exercise. The oscillation of the two lips may change the vibratory characteristics of the vocal folds and induce laryngeal movement due to oscillating supraglottal pressures. This study examines electroglottographic (EGG) signal variation during lip trill production for both normal adduction and during breathy voice quality. Human data were compared to the results using the computer model BG3M (a 3-mass vocal fold model with sub- and supraglottal acoustic tracts and M5 intraglottal pressures).
Methods: EGGW25 (a measure similar to closed quotient) and EGG DC shift data were obtained for four professional voice users as they produced a vowel, a lip trill, and a vowel on the same breath group. EGG, a microphone system, and the oral air pressure transducer with mouth tube (Glottal Enterprises) were used. Five trials of the vowel-trill-vowel exercise were obtained for both normal and breathy voice conditions at a comfortable loudness level and on modal pitches (men: Eb3 [155 Hz], women: C4 [261 Hz]).
Results: For three of the participants, lip trill produced with normal phonation temporarily reduced EGGW25 by at least 0.05, but this reduction was not maintained following the cessation of lip trill. This trend (less vocal fold contact time with narrow lip opening) was predicted by the BG3M model. Two of the participants exhibited the same trend during breathy voice production. In contrast, one participant increased EGGW25 during lip trill production regardless of phonatory condition, and one demonstrated an increase in EGGW25 throughout the utterance during breathy phonation, likely due to increased glottal adduction to retain sufficient subglottal pressure. Two participants’ larynges moved in time with the oral air pressure variations seen in the lip trill, as hypothesized by the large EGG DC shifts observed, while two participants did not have timed variations in their EGG DC shifts. BG3M modelling suggested that the vocal folds came together at the end of glottal closing in a parallel, vertical pattern during the minimum lip aperture, but came together with a divergent pattern when the vocal tract was nearly uniform. For the nearly occluded opening, BG3M predicted 3.8° glottal flow skewing, but 36.9° skewing for the nearly – uniform vocal tract. This suggests that vocal tract closing during SOVT gestures may reduce glottal flow skewing.
Discussion and Conclusions: For the subjects of this study, it can be hypothesized that lip trill does not change the muscle contraction levels of the laryngeal musculature due to the nearly-constantly observed return of the EGGW25 to pre-trill widths. Rather, the changes to EGGW25 observed during lip trill may be related to the increase of aerodynamic pressures of the supraglottal vocal tract needed to sustain simultaneous oscillation of the vocal folds and the lips. The increased supraglottal air pressures may also result in the EGG DC shifts observed in two participants. Laryngeal height changes may be a clinical benefit of lip trill.