A study on generation of the voice bar in voiced stops
Jianwu Dang, Takayoshi Nakai, Yongnan Piao, Shinobu Masaki · The Journal of the Acoustical Society of America · 2006
The voice bar of the voiced stops is generally considered as a radiation sound from the vibration of the laryngeal wall while the mouth is closed. In this study, we investigated the mechanisms involved in generation of the voice bar of the voiced stops by means of acoustic and mechanical measurements and MRI observations for three Japanese speakers. Radiated sounds from the mouth and the nostrils were recorded separately, while vibration of the laryngeal wall was measured by an accelerometer. The correlation of the nostril radiation to the voice bar is higher than that of the laryngeal wall vibration to the voice bar. It implies that the major component of the voice bar is likely radiated from the nostrils but not the laryngeal wall, where the nostrils’ radiation is possibly caused by a transvelar coupling of the yielding velum. The authors have proposed a two-layer board model to imitate the transvelar function. To obtain physiological evidences, articulation of stop-vowel sequences was observed by MRI movies. It found that the thickness of the velum varies with the adjacent vowels. Applying this measurement to the two-layer board model, the generation of the voice bar was evaluated by comparing the simulations and observations.