A poor use case for voice synthesis
Switch branches/tags
Nothing to show
Clone or download
Latest commit 00f441c Oct 19, 2018



A poor use case for vocal synthesis

Demo at jons.website/projects/jon-trombone.


I've always wanted to be a web app!

This nonsense uses Three.js to show a 3D scanned model of me, which is animated to have a flapping jaw, which may or may not reflect reality. A modified version of Pink Trombone is used to synthesize vocals to match the movement of the jaw.

It comes with really, really basic MIDI parsing and playback. Someday I'll update this with support for multiple voices. If you want to try out a MIDI file of your own, just drag it onto my head.

Build Instructions

I can't imagine why you'd want to do this, but you can build this yourself with a few simple commands. It's a Node project, so all you need to do is run npm install in the root, and then you can use the following npm tasks (e.g. npm run build):

  • build - Builds the project into /dist
  • dev:watch - Builds and presents the project's test page (live reloads)

Everything you need to run is in /dist. For an example of an integration, see the test page.

Jon-Trombone will use Guify if it's available to give you a way to fiddle with the parameters.



MIT License. See license.md for details.