I met with Mark Wing-Davey from Graduate Acting a couple days ago, and we bounced around some good ideas for developing educational tools for instant feedback on one’s own vocal resonance/timbre. His main idea was having the timbre of the voice represented by 4 different ranges: violin, viola, cello, bass. Some straight up classical thinking. I feel that. Anyway, the first step I took after that meeting was loosely mapping brightness to video grayscale. Determining the brightness of a sound is the product of measuring the spectral average of a signal weighted by its amplitude. As I continue with this project, I plan on having more involved spectral analysis come into play. In this video, I am changing the resonance of my voice through a held tone–a constructive analysis of speech is going to be more difficult. For one, you have to calibrate the software to the frequency of someones voice. I made a program that does median filtering on sound input over the length of a test sentence to get an estimate of the person’s standard speaking frequency. Anyway, more to come!