Inspiration

The muppets, a theremin, and laugh tracks.

What it does

This sound synthesizer allows anyone to generate vocal vowel sounding n in-air capacitive sensor picks up the 3D position of your finger in the air and translates that into pitch and vowel sounds.

How We built it

We used a 3D in air capacitive sensor matrix in the form of a Raspberry Pi HAT which we plugged into a Raspberry Pi. This then acted as our sensor input which we used to generate audio with the PyAudio Python audio synthesis library. Depending on the sensor x, y, and z values, we modulated the pitch being generated, and made various almost working vowel sounds.

Challenges I ran into

  • The Skywriter HAT for the raspberry pi doesn't give hand posture information which would have been useful to provide a more intuitive way of controlling vowel sounds.
  • It is difficult to create a continuously frequency modulating sound wave.
  • We ran into trouble when trying to consistently create good sounds from Python's PyAudio library compared to matlab for example

Accomplishments that I'm proud of

  • We actually made a customisable multi dimensional theremin
  • We were able to synthesize vowel sounds that were disturbingly humanlike

What I learned

  • Signal Processing is hard

What's next for SoundsGood

  • Making it vowel sound determination work with hand posture instead of x and y finger position
Share this project:

Updates