Audio declipping performance enhancement via crossfading

TL;DR

This paper introduces a postprocessing technique to improve the perceptual quality of inconsistent audio declipping methods, making them comparable to consistent methods with lower computational cost.

Contribution

It proposes a simple sample replacement method and an improved variant to enhance inconsistent declipping techniques, validated through perceptual metrics.

Findings

Most inconsistent declipping methods benefit from the proposed approach.

The SS PEW method with the proposed enhancement performs comparably to top consistent methods.

The improved method achieves similar perceptual quality at a lower computational cost.

Abstract

Some audio declipping methods produce waveforms that do not fully respect the physical process of clipping, which is why we refer to them as inconsistent. This letter reports what effect on perception it has if the solution by inconsistent methods is forced consistent by postprocessing. We first propose a simple sample replacement method, then we identify its main weaknesses and propose an improved variant. The experiments show that the vast majority of inconsistent declipping methods significantly benefit from the proposed approach in terms of objective perceptual metrics. In particular, we show that the SS PEW method based on social sparsity combined with the proposed method performs comparable to top methods from the consistent class, but at a computational cost of one order of magnitude lower.