New Technology for Teaching Voice Science and Pedagogy: The Madde Synthesizer (Svante Granqvist)
Kenneth Bozeman · Journal of Singing · 2012
(ProQuest: ... denotes formula omitted.) TEACHING VOICE ACOUSTICS to undergraduates is often met with resistance from those students less science inclined. This may be due in part to real aptitude differences, but perhaps is more often related to students' discomfort and/or lack of facility with the modes of communication typically used in scientific writing. Converting visual aids to musician-friendly formats can bypass either problem. For example, the locations of the first formants of the vocal tract are crucial to vocal resonance because of their interaction with the harmonics of the pitches we sing. It is therefore helpful to know first formant locations, at least approximately, when working with various voice types. For this author's work on male passaggio, a more user friendly way to keep track of formant locations, at least for some cardinal vowels, was generated (Figure 1). In this image, typical vowel formant locations for a lyric tenor are indicated in boxes on the treble clef, with the locations of turning/closing an octave lower on the bass clef. (There is a timbral shift when the second harmonic of the voice source crosses the first formant of the vocal tract, the F1/H2 crossing, which is variously called, covering, turning over, closing [voce chiusa], etc.)1 Such a visual presentation is readily understood by musicians and therefore more immediately usable. MADDE SYNTHESIZER Swedish engineer, Svante Granqvist, made a free, downloadable voice synthesizer available on his website a number of years ago, and teachers and scientists interested in voice acoustics have been taking advantage of his Madde Synthesizer ever since. For those somewhat comfortable with voice science, his original version was already quite helpful. Among other things, it allowed one to set: 1) number and frequency locations of up to six vocal tract formants (both factors relevant for vocal Fach); 2) bandwidth of the formants (frequency width of the resonance areas); 3) vibrato rate and extent; 4) net number of source harmonics (up to 40; relevant for mode of phonation/glottal resistance/laryngeal register); 5) net roll-off in power of the voice source harmonic spectrum (power loss per octave of the harmonic spectrum; also relevant for mode of phonation/glottal resistance/laryngeal register). Once parameters were set, one could play various pitches or scales on a short keyboard (C^sub 2^-B^sub 6^) to hear the synthesized sung results. With a bit of quick math, one could determine when harmonics crossed formants, but there was no visual representation of those intersections. In the summer of 2010 at the Salt Lake City NATS Conference, Scott McCoy shared with me a clever visual program that he had created, the Formant Calculator, which placed a set of average male or female formants for a set of cardinal vowels on a keyboard with color shading, then allowed one to select a pitch, from which the first twelve harmonics would be built and displayed. This was a visual aid that I had only crudely done for my pedagogy classes with overhead transparencies. The Formant Calculator program offered an excellent solution to this pedagogic need. However, shortly after exploring Formant Calculator, I realized that Scott had created an ideal mate for Madde, where formants for any vowel and voice type could be created, manipulated, and heard. Madde already had a keyboard, albeit an octave short of the singers formant cluster. With Scott's permission and blessing, I contacted Svante Granqvist, who within a few weeks added the top octave of the keyboard (C^sub 7^-B^sub 7^) to include the most relevant higher formants of the singer's formant cluster, a red color shading for any of the six programmable formants, with formant bandwidth variability, and a display of as many as fifty harmonics up from any selected pitch on the keyboard (Figure 2). With this new edition of Madde, Grant and Scott have provided the pedagogy world an outstanding tool for explaining and exploring voice acoustics in pedagogy classes. …