Advances in Physical Vocal Tract Modeling: Mannequin Head and Torso Singing Simulator
Introduction:
Quantifying the singing voice acoustically, particularly the effects of distance and room, requires repeatable objective measures of vocalizations. Advances have been made in the study of voice directivity in speech and in singing to explain why the voice sounds louder in front of the singer/speaker and how the singer’s formant is radiated in space. Unfortunately, most of these studies rely on data collected in anechoic environments where the singer’s autophonic perception is altered. Because vocal production and perception are linked, these studies have limited real-world applicability.
Singing is a highly variable behavior, and acoustic studies are often limited by high inter- and intrasubject variability. Physical simulations may reduce this variability, but commercially available head and torso simulators and the “artificial singer” by Kob and Jers may be poor analogs for in vivo singing. These simulators do not have 3D modeled vocal tracts, and often have unrealistic (or missing) pinnae and ear canals.
Method:
The construction of a dummy head and torso with anatomically modeled auditory and phonatory structures, coupled with appropriate electronic transducers is feasible. Surprisingly, to date, no such simulator exists.
Such a mannequin head and torso simulator was fabricated in acoustically analogous materials with MRI/CT modeled auditory and phonatory structures, and appropriate electronic transducers and data acquisition hardware. In particular, the dummy head and torso contains two groundbreaking features that have never been previously implemented:
(a) a switchable vocal tract
(b) a realistic ear canal and pinnae
Anatomical structures were derived from segmented CT and MRI scans of two professional singers. High-quality, SPL-calibrated microphones are mounted at the inner ends of the ear canal at the level of the eardrums to provide binaural acoustic signals. These allow the computation of self-to-other ratios at an unprecedented level of accuracy.
Results:
3D modeling is complete and fabrication is underway. The dummy head simulator is currently undergoing testing.