Abstracts Submitted: 459
Number of Users: 527
View Abstracts Submitted
Back to home Page
Eliciting accurate and acoustically meaningful vocal tract information directly from speech sounds in real-time remains difficult, exacerbated by the fact that the voice itself – with harmonic resolution of >100 Hz – samples the frequency domain poorly. Consequently, methods extracting information directly from [radiated] speech sounds offer only limited frequency resolution, while visualization techniques are unsatisfactory because the visual domain does not translate easily into acoustic behavior; further, such methods are either highly invasive (endoscopic study), possibly harmful (x-ray fluoroscopy), difficult to interpret (ultrasound imaging) or too slow and noisy (MRI) to be ‘ecological’. However, external broadband signal excitation applied at the speaker (or singer)’s mouth has been successfully used to estimate acoustic resonances of the vocal tract non-invasively during speech and singing. Accordingly, we present here a modified, low cost, light-weight, pocket-sized and simplified version of this acoustic measurement technique, with reduced sampling time and improved low frequency detection, such that vocal tract measurements may be easily deployed ‘in the field’ and facilitate a more ‘ecological/natural’ tracking of phonatory gestures in real-time. Technical details of this measurement system will be presented, along with empirical data collected with volunteer speakers and also physical models of the vocal-tract, showing that the resulting relative impedance spectrum γ (‘gamma’) measured provides a meaningful estimation of vocal tract resonances under ‘field conditions’ which are comparable in performance with previous studies made under ‘laboratory conditions’.
© Copyright 2017 All Rights Reserved