dc.contributor.author | Hayes, B | |
dc.contributor.author | Saitis, C | |
dc.contributor.author | Fazekas, G | |
dc.contributor.author | Proceedings of the 22nd International Society for Music Information Retrieval | |
dc.date.accessioned | 2021-07-19T15:32:56Z | |
dc.date.available | 2021-07-09 | |
dc.date.available | 2021-07-19T15:32:56Z | |
dc.identifier.uri | https://qmro.qmul.ac.uk/xmlui/handle/123456789/73126 | |
dc.description.abstract | We present the Neural Waveshaping Unit (NEWT): a novel, lightweight, fully causal approach to neural audio synthesis which operates directly in the waveform domain, with an accompanying optimisation (FastNEWT) for efficient CPU inference. The NEWT uses time-distributed multilayer perceptrons with periodic activations to implicitly learn nonlinear transfer functions that encode the characteristics of a target timbre. Once trained, a NEWT can produce complex timbral evolutions by simple affine transformations of its input and output signals. We paired the NEWT with a differentiable noise synthesiser and reverb and found it capable of generating realistic musical instrument performances with only 260k total model parameters, conditioned on F0 and loudness features. We compared our method to state-of-the-art benchmarks with a multi-stimulus listening test and the Fréchet Audio Distance and found it performed competitively across the tested timbral domains. Our method significantly outperformed the benchmarks in terms of generation speed, and achieved real-time performance on a consumer CPU, both with and without FastNEWT, suggesting it is a viable basis for future creative sound design tools. | en_US |
dc.subject | neural audio synthesis | en_US |
dc.subject | signal processing | en_US |
dc.subject | deep learning | en_US |
dc.subject | machine learning | en_US |
dc.subject | audio synthesis | en_US |
dc.title | Neural Waveshaping Synthesis | en_US |
dc.type | Conference Proceeding | en_US |
dc.rights.holder | © Copyright 2021 Ben Hayes. | |
pubs.author-url | https://benhayes.net/ | en_US |
pubs.notes | Not known | en_US |
pubs.publication-status | Accepted | en_US |
dcterms.dateAccepted | 2021-07-09 | |
qmul.funder | UKRI Centre for Doctoral Training in Artificial Intelligence and Music::Engineering and Physical Sciences Research Council | en_US |