Human vocal tract resonances and the corresponding mode shapes investigated by three-dimensional finite-element modelling based on CT measurement

Abstract

Resonance frequencies of the vocal tract have traditionally been modelled using one-dimensional models. These cannot accurately represent the events in the frequency region of the formant cluster around 2.5–4.5 kHz, however. Here, the vocal tract resonance frequencies and their mode shapes are studied using a three-dimensional finite element model obtained from computed tomography measurements of a subject phonating on vowel [a:]. Instead of the traditional five, up to eight resonance frequencies of the vocal tract were found below the prominent antiresonance around 4.7 kHz. The three extra resonances were found to correspond to modes which were axially asymmetric and involved the piriform sinuses, valleculae, and transverse vibrations in the oral cavity. The results therefore suggest that the phenomenon of speaker's and singer's formant clustering may be more complex than originally thought.

Key words::

• Acoustic mode shapes of vibration
• biomechanics of human voice
• speaker's and singer's formant
• voice production modelling

Acknowledgements

The authors are very grateful to Doc. MUDr. Petr Krupa from the Hospital U Svaté Anny in Brno for providing the CT data of a human vocal tract during phonation.

Declaration of interest: The authors report no conflicts of interest.

The research was supported by the grant GAČR No. P101/12/1306 and by the Academy of Finland No 106139 Biomechanical study on the traumatizing mechanisms in vocal fold vibration and by the COST Action 2103 Advanced Voice Function Assessment. The research of Dr Švec has been supported by the European Social Fund projects no. CZ.1.07/2.4.00/17.0009 and CZ.1.07/2.3.00/20.0057.