A portable do-it-yourself DIY system of apparatus demonstrates how to generate a standard vowel sound like (a, e, i, o, u), say /eI/, see /i/, eye /aI/, go /oU/, you /u/. System components include (1) a periodic glottal source vocal tract excitation GS (t), (2) a vocal tract filter frequency response VT (f) with three formant peaks for vowel selection, and (3) detection and display of the vowel signal VS (t) versus time resulting from the frequency response |VT (f) / GS (f) |. The glottal source excitation is related to the periodic cycle T of the opening and closing of the vocal fold (cord) vibrations. The time domain airflow GS (t) through the glottis is somewhat related to the measurement of the area of the opening versus time. A vowel sound will be generated approximately using a periodic ramp generator GS (t) that is on for 0 t T/2 and off for slightly less than T/2. Three Helmholtz resonators will represent the formant peaks f1, f2, and f3 of VT (f), generating the time domain vowel sound VT (t). See the pioneering work of Gunnar Fant (1919–2009) and Kenneth Stevens (1924–2013) ; and H. Helmholtz (1821–1894) the first synthesizer and R. Koenig (1832–1901) sound analyzers).
Twitty et al. (Wed,) studied this question.