Skip to main content
School of Electronic Engineering and Computer Science

Christopher Mitcheltree




Project title: Representation Learning for Audio Production Style and Modulations 

C4DM theme affiliation: Audio Engineering, Machine Listening, Music Informatics 

Abstract: Today anyone with a laptop can create studio quality music and release it to the world via the internet. As a result, formerly niche genres of electronic music are reaching wider audiences and giving birth to new genres that are pushing the limits of audio production and redefining what music can sound like.  

This project is centred around the following two research questions: 

  • How can time varying parameters (also known as modulations) of synthesizers and audio effects be learned, controlled, and augmented with machine learning 
  • Can modulation information be leveraged to improve representations of audio production style (i.e. what makes an electronic music artist sound unique through their choice of modulations, synthesizers, audio effects, mixing, and samples). 


Back to top