Distortion Audio Effects: Learning How to Recover the Clean Signal

TL;DR

This paper presents neural network models for removing distortion effects from guitar tracks, improving quality and speed over traditional methods, and demonstrating versatility across various distortion types.

Contribution

It introduces neural network-based methods for effective and fast removal of distortion audio effects, outperforming sparse optimization techniques.

Findings

Models outperform state-of-the-art sparse optimization methods.

Effective for multiple distortion effects including declipping.

Faster inference with comparable or better quality.

Abstract

Given the recent advances in music source separation and automatic mixing, removing audio effects in music tracks is a meaningful step toward developing an automated remixing system. This paper focuses on removing distortion audio effects applied to guitar tracks in music production. We explore whether effect removal can be solved by neural networks designed for source separation and audio effect modeling. Our approach proves particularly effective for effects that mix the processed and clean signals. The models achieve better quality and significantly faster inference compared to state-of-the-art solutions based on sparse optimization. We demonstrate that the models are suitable not only for declipping but also for other types of distortion effects. By discussing the results, we stress the usefulness of multiple evaluation metrics to assess different aspects of reconstruction in distortion effect removal.