Skip to main content

This Raspberry Pi Guitar Pedal Uses Machine Learning for Effects

Raspberry Pi
(Image credit: Keith Bloemer)

When the soldering iron gets hot, makers have been known to cook up some sweet music on the Raspberry Pi. In some cases, we mean that literally, like today with Keith Bloemer's NeuralPi project.

NeuralPi is a Raspberry Pi-based guitar pedal that uses machine learning to create custom effects. We've always insisted the best Raspberry Pi projects are the ones you can personalize—in this case, you can train the system with existing models or a new one of your own to get the sound you want.

Recreating this project will set you back just a little over $160. Bloemer used a Raspberry Pi 4 alongside a HiFiBerry ADC +DAC housed inside a HiFiBerry case. A male RCA to 1/4-inch female audio adapter is necessary for guitar pedal output, and he used a dual 1/4-inch female to a 1/8-inch male stereo adapter to connect the guitar to the HiFiBerry module.

NeuralPi is designed to work specifically with Elk Audio OS, an open-source OS catered to optimizing audio processing on embedded devices. You can find the NeuralPi source code on GitHub, along with a couple of Machine learning models you can use to get started right away.

If you're interested in recreating this project, check out the full tutorial at Towards Data Science written by Bloemer himself. Be sure to follow him for more Pi projects and updates on this one.