Music De-limiter Networks via Sample-wise Gain Inversion

TL;DR

This paper introduces a neural network framework for estimating uncompressed music signals from heavily compressed ones, inspired by the principle of sample-wise gain reduction, and demonstrates high-quality reconstruction on a new dataset.

Contribution

It proposes the sample-wise gain inversion framework and a large dataset for training music de-limiter networks, advancing the ability to recover original music from compressed signals.

Findings

Achieved 24.0 dB SI-SDR in reconstructing original music signals.

Created the musdb-XL-train dataset with 300k segments for training.

Provided open-source code and models for the music de-limiter task.

Abstract

The loudness war, an ongoing phenomenon in the music industry characterized by the increasing final loudness of music while reducing its dynamic range, has been a controversial topic for decades. Music mastering engineers have used limiters to heavily compress and make music louder, which can induce ear fatigue and hearing loss in listeners. In this paper, we introduce music de-limiter networks that estimate uncompressed music from heavily compressed signals. Inspired by the principle of a limiter, which performs sample-wise gain reduction of a given signal, we propose the framework of sample-wise gain inversion (SGI). We also present the musdb-XL-train dataset, consisting of 300k segments created by applying a commercial limiter plug-in for training real-world friendly de-limiter networks. Our proposed de-limiter network achieves excellent performance with a scale-invariant source-to-distortion ratio (SI-SDR) of 24.0 dB in reconstructing musdb-HQ from musdb-XL data, a limiter-applied version of musdb-HQ. The training data, codes, and model weights are available in our repository (https://github.com/jeonchangbin49/De-limiter).