Voice efficiency for different voice qualities combining experimentally derived sound signals and numerical modeling of the vocal tract

Abstract

Purpose: Concerning voice efficiency considerations of different singing styles, from western classical singing to contemporary commercial music, only limited data is available to date. This single-subject study attempts to quantify the acoustic sound intensity within the human glottis depending on different vocal tract configurations and vocal fold vibration.Methods: Combining Finite-Element-Models derived from 3D-MRI data, audio recordings, and electroglottography (EGG) we analyzed vocal tract transfer functions, particle velocity and acoustic pressure at the glottis, and EGG-related quantities to evaluate voice efficiency at the glottal level and resonance characteristics of different voice qualities according to Estill Voice Training®.Results: Voice qualities Opera and Belting represent highly efficient strategies but apply different vowel strategies and should thus be capable of predominate orchestral sounds. Twang and Belting use similar vowels, but the twang vocal tract configuration enabled the occurrence of anti-resonances and was associated with reduced vocal fold contact but still partially comparable energy transfer from the glottis to the vocal tract. Speech was associated with highly efficient glottal to vocal tract energy transfer, but with the absence of psychoactive strategies makes it more susceptible to noise interference. Falsetto and Sobbing apply less efficiently. Falsetto mainly due to its voice source characteristics, Sobbing due to energy loss in the vocal tract. Thus technical amplification might be appropriate here.Conclusion: Differences exist between voice qualities regarding the sound intensity, caused by different vocal tract morphologies and oscillation characteristics of the vocal folds. The combination of numerical analysis of geometries inside the human body and experimentally determined data outside sheds light on acoustical quantities at the glottal level.

An Acoustic Simulation Method of the Japanese Vowels /i/ and /u/ by Using the Boundary Element Method

“…The recent development of acoustic wave-based techniques, such as the finite element, finite difference, and boundary element methods, has made it possible to visualize and conduct a detailed analysis of the relationship between the morphology of, and sound produced by, a three-dimensional (3D) shape obtained for a 3D model of the vocal tract [2][3][4][5][6][7]. Many analysis methods using the finite element method have been reported, but the range targeted for the analysis target has been limited to a small area from the oral cavity to the glottis.…”