Effortlessly deploy your trained model to a DAW!

Powered by RedCircle

Listen on

All podcast episodes.

Sign up for WolfSound’s newsletter!


In this episode, I was honored to interview Andrew Fyfe; a musician, a software engineer, and a researcher in the field of deep learning applied to audio.

In particular, Andrew is working on an exciting new technology, Neutone. Neutone allows quick deployment of trained neural networks to audio plugins and running them in the digital audio workstation for verification, testing, up to full-blown product releases for users.

At the same time, Andrew is wrapping up his PhD while working full time from Japan. He was kind enough to share his story and his journey to becoming an audio researcher.

Don’t forget to submit your plugins for the Neural Audio Plugin Competition until March the 17th of 2023!

Note: If you like the podcast so far, please, go to Apple Podcasts and leave me a review there. It will benefit both sides: more reviews mean a broader reach on Apple Podcasts and feedback can help me to improve the show and provide better quality content to you. Thank you for doing this 🙏

Episode Contents

In this podcast episode, you will learn:

  • how Andrew learned audio processing and became an audio researcher after a successful rock band musician career,
  • how he explored AI for music creation in his PhD,
  • what is Neutone about and how to use it to deploy trained neural networks from PyTorch to audio plugins,
  • what is Neural Audio Plugin competition,
  • how is living and working in Japan for a person from the UK.


Below you’ll find all people, places, and references mentioned in the podcast episode.

  1. Andrew Fyfe (andrew@qosmo.jp)
    1. LinkedIn profile
  2. Technologies
    1. Pure Data
    2. Python
    3. PyTorch deep learning framework
    4. C
    5. C++
    6. Max/MSP by Cycling ‘74
    7. JUCE
    8. iPlug2 by Oli Larkin
    9. Microsoft ONNX runtime
    10. JavaScript
    11. Go
    12. Rust
  3. Deep learning concepts
    1. Generative Adversarial Networks (GANs)
    2. Variational Autoencoders (VAEs)
    3. Diffusion models
  4. Companies
    1. Qosmo, Inc.
      1. Neutone
        1. Discord server
        2. Presentation explaining Neutone at ADC22 by Andrew Fyfe and Christopher Mitcheltree
        3. Neural Audio Plugin Competition (NAP) at TheAudioProgrammer.com
    2. Krotos Audio
    3. TikTok
    4. GPU Audio
    5. MathWorks
    6. Cycling ’74
  5. Other
    1. University of Glasgow
      1. Electronics with Music Bachelor of Engineering/Master of Engineering program
    2. Neural Audio Synthesis Hackathon (NASH) at Queen Mary University of London.
    3. Audio Developers Conference 2022
    4. ADCx, 1-day meeting in San Francisco

Thank you for listening!