Deep Learning for Tube Amplifier Emulation

Eero-Pekka Damskägg; Lauri Juvela; Etienne Thuillier; Vesa Välimäki

doi:10.1109/ICASSP.2019.8682805

Deep Learning for Tube Amplifier Emulation

Eero-Pekka Damskägg, Lauri Juvela, Etienne Thuillier, Vesa Välimäki

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

30 Citations (Scopus)

Abstract

Analog audio effects and synthesizers often owe their distinct sound to circuit nonlinearities. Faithfully modeling such significant aspect of the original sound in virtual analog software can prove challenging. The current work proposes a generic data-driven approach to virtual analog modeling and applies it to the Fender Bassman 56F-A vacuum-tube amplifier. Specifically, a feedforward variant of the WaveNet deep neural network is trained to carry out a regression on audio waveform samples from input to output of a SPICE model of the tube amplifier. The output signals are pre-emphasized to assist the model at learning the high-frequency content. The results of a listening test suggest that the proposed model accurately emulates the reference device. In particular, the model responds to user control changes, and faithfully restitutes the range of sonic characteristics found across the configurations of the original device.

Original language	English
Title of host publication	44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019; Brighton; United Kingdom; 12-17 May 2019 : Proceedings
Publisher	IEEE
Pages	471-475
Number of pages	5
ISBN (Electronic)	978-1-4799-8131-1
ISBN (Print)	978-1-4799-8132-8
DOIs	https://doi.org/10.1109/ICASSP.2019.8682805
Publication status	Published - 1 May 2019
MoE publication type	A4 Article in a conference publication
Event	IEEE International Conference on Acoustics, Speech, and Signal Processing - Brighton, United Kingdom Duration: 12 May 2019 → 17 May 2019 Conference number: 44

Publication series

Name	Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing
ISSN (Print)	1520-6149
ISSN (Electronic)	2379-190X

Conference

Conference	IEEE International Conference on Acoustics, Speech, and Signal Processing
Abbreviated title	ICASSP
Country/Territory	United Kingdom
City	Brighton
Period	12/05/2019 → 17/05/2019

Keywords

Integrated circuit modeling
Neural networks
Convolution
Computational modeling
Predictive models
SPICE
Computer architecture
Audio systems
feedforward neural networks
music
nonlinear systems
supervised learning

Access to Document

10.1109/ICASSP.2019.8682805

https://arxiv.org/abs/1811.00334

OpenUrl availability

Full text

Cite this

Damskägg, E-P., Juvela, L., Thuillier, E., & Välimäki, V. (2019). Deep Learning for Tube Amplifier Emulation. In 44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019; Brighton; United Kingdom; 12-17 May 2019 : Proceedings (pp. 471-475). [8682805] (Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing). IEEE. https://doi.org/10.1109/ICASSP.2019.8682805

Damskägg, Eero-Pekka ; Juvela, Lauri ; Thuillier, Etienne et al. / Deep Learning for Tube Amplifier Emulation. 44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019; Brighton; United Kingdom; 12-17 May 2019 : Proceedings. IEEE, 2019. pp. 471-475 (Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing).

@inproceedings{405e0d894477440688046c42e4816a4c,

title = "Deep Learning for Tube Amplifier Emulation",

abstract = "Analog audio effects and synthesizers often owe their distinct sound to circuit nonlinearities. Faithfully modeling such significant aspect of the original sound in virtual analog software can prove challenging. The current work proposes a generic data-driven approach to virtual analog modeling and applies it to the Fender Bassman 56F-A vacuum-tube amplifier. Specifically, a feedforward variant of the WaveNet deep neural network is trained to carry out a regression on audio waveform samples from input to output of a SPICE model of the tube amplifier. The output signals are pre-emphasized to assist the model at learning the high-frequency content. The results of a listening test suggest that the proposed model accurately emulates the reference device. In particular, the model responds to user control changes, and faithfully restitutes the range of sonic characteristics found across the configurations of the original device.",

keywords = "Integrated circuit modeling, Neural networks, Convolution, Computational modeling, Predictive models, SPICE, Computer architecture, Audio systems, feedforward neural networks, music, nonlinear systems, supervised learning",

author = "Eero-Pekka Damsk{\"a}gg and Lauri Juvela and Etienne Thuillier and Vesa V{\"a}lim{\"a}ki",

year = "2019",

month = may,

day = "1",

doi = "10.1109/ICASSP.2019.8682805",

language = "English",

isbn = "978-1-4799-8132-8",

series = "Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing",

publisher = "IEEE",

pages = "471--475",

booktitle = "44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019; Brighton; United Kingdom; 12-17 May 2019 : Proceedings",

address = "United States",

note = "IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP ; Conference date: 12-05-2019 Through 17-05-2019",

}

Damskägg, E-P, Juvela, L , Thuillier, E & Välimäki, V 2019, Deep Learning for Tube Amplifier Emulation. in 44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019; Brighton; United Kingdom; 12-17 May 2019 : Proceedings., 8682805, Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing, IEEE, pp. 471-475, IEEE International Conference on Acoustics, Speech, and Signal Processing, Brighton, United Kingdom, 12/05/2019. https://doi.org/10.1109/ICASSP.2019.8682805

Deep Learning for Tube Amplifier Emulation. / Damskägg, Eero-Pekka; Juvela, Lauri ; Thuillier, Etienne et al.
44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019; Brighton; United Kingdom; 12-17 May 2019 : Proceedings. IEEE, 2019. p. 471-475 8682805 (Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Deep Learning for Tube Amplifier Emulation

AU - Damskägg, Eero-Pekka

AU - Juvela, Lauri

AU - Thuillier, Etienne

AU - Välimäki, Vesa

N1 - Conference code: 44

PY - 2019/5/1

Y1 - 2019/5/1

N2 - Analog audio effects and synthesizers often owe their distinct sound to circuit nonlinearities. Faithfully modeling such significant aspect of the original sound in virtual analog software can prove challenging. The current work proposes a generic data-driven approach to virtual analog modeling and applies it to the Fender Bassman 56F-A vacuum-tube amplifier. Specifically, a feedforward variant of the WaveNet deep neural network is trained to carry out a regression on audio waveform samples from input to output of a SPICE model of the tube amplifier. The output signals are pre-emphasized to assist the model at learning the high-frequency content. The results of a listening test suggest that the proposed model accurately emulates the reference device. In particular, the model responds to user control changes, and faithfully restitutes the range of sonic characteristics found across the configurations of the original device.

AB - Analog audio effects and synthesizers often owe their distinct sound to circuit nonlinearities. Faithfully modeling such significant aspect of the original sound in virtual analog software can prove challenging. The current work proposes a generic data-driven approach to virtual analog modeling and applies it to the Fender Bassman 56F-A vacuum-tube amplifier. Specifically, a feedforward variant of the WaveNet deep neural network is trained to carry out a regression on audio waveform samples from input to output of a SPICE model of the tube amplifier. The output signals are pre-emphasized to assist the model at learning the high-frequency content. The results of a listening test suggest that the proposed model accurately emulates the reference device. In particular, the model responds to user control changes, and faithfully restitutes the range of sonic characteristics found across the configurations of the original device.

KW - Integrated circuit modeling

KW - Neural networks

KW - Convolution

KW - Computational modeling

KW - Predictive models

KW - SPICE

KW - Computer architecture

KW - Audio systems

KW - feedforward neural networks

KW - music

KW - nonlinear systems

KW - supervised learning

UR - http://www.scopus.com/inward/record.url?scp=85069003738&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2019.8682805

DO - 10.1109/ICASSP.2019.8682805

M3 - Conference contribution

SN - 978-1-4799-8132-8

T3 - Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing

SP - 471

EP - 475

BT - 44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019; Brighton; United Kingdom; 12-17 May 2019 : Proceedings

PB - IEEE

T2 - IEEE International Conference on Acoustics, Speech, and Signal Processing

Y2 - 12 May 2019 through 17 May 2019

ER -

Damskägg E-P, Juvela L , Thuillier E , Välimäki V. Deep Learning for Tube Amplifier Emulation. In 44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019; Brighton; United Kingdom; 12-17 May 2019 : Proceedings. IEEE. 2019. p. 471-475. 8682805. (Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing). doi: 10.1109/ICASSP.2019.8682805

Deep Learning for Tube Amplifier Emulation

Abstract

Publication series

Conference

Keywords

Access to Document

OpenUrl availability

Other files and links

Fingerprint

Aalto Acoustics Lab

Science-IT

Cite this

Deep Learning for Tube Amplifier Emulation

Abstract

Publication series

Conference

Keywords

Access to Document

OpenUrl availability

Other files and links

Fingerprint

Equipment

Aalto Acoustics Lab

Science-IT

Cite this