Gray-Box Modeling of Guitar Amplifiers

Eichas, Felix; Zölzer, Udo

doi:10.17743/jaes.2018.0052

Cited by 20 publications

(20 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Each convolutional layer can be viewed as a Wiener model: a linear filter followed by a static nonlinearity. In previously proposed black-box and grey-box models of musical distortion circuits it has often been assumed that the device can be approximated by a Wiener model [8,11], a Hammerstein model [9] or a Wiener-Hammerstein [10,12] model. As a cascade of Wiener models, the WaveNet-style neural network makes fewer assumptions about the design of the modelled device, and so should be applicable to the modelling of a broad range of nonlinear systems.…”

Section: Convolutional Layermentioning

confidence: 99%

“…The output of the device was recorded by connecting it to one of the inputs of the audio interface. The output of the audio interface was also directly connected to another input of the audio interface, in order to estimate the effect of the audio interface in the measurement, as suggested in [12]. The recorded direct signal from the audio interface and the recorded output from the device under study make up the input/target pairs used in the training of the network.…”

Section: Dataset Generationmentioning

confidence: 99%

“…Black-box modelling is based on measuring the circuit's response to some input signals, and creating a model which replicates the observed input-output mapping. Black-box models for VA modelling include block-oriented models, which are based on assumptions about the design of the modelled circuit [8][9][10][11][12]. As an example, a Wiener model [8,11] emulates the circuit as a linear filter followed by a static nonlinearity.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Real-Time Guitar Amplifier Emulation with Deep Learning

Wright

Damskägg²,

Juvela³

et al. 2020

Applied Sciences

View full text Add to dashboard Cite

This article investigates the use of deep neural networks for black-box modelling of audio distortion circuits, such as guitar amplifiers and distortion pedals. Both a feedforward network, based on the WaveNet model, and a recurrent neural network model are compared. To determine a suitable hyperparameter configuration for the WaveNet, models of three popular audio distortion pedals were created: the Ibanez Tube Screamer, the Boss DS-1, and the Electro-Harmonix Big Muff Pi. It is also shown that three minutes of audio data is sufficient for training the neural network models. Real-time implementations of the neural networks were used to measure their computational load. To further validate the results, models of two valve amplifiers, the Blackstar HT-5 Metal and the Mesa Boogie 5:50 Plus, were created, and subjective tests were conducted. The listening test results show that the models of the first amplifier could be identified as different from the reference, but the sound quality of the best models was judged to be excellent. In the case of the second guitar amplifier, many listeners were unable to hear the difference between the reference signal and the signals produced with the two largest neural network models. This study demonstrates that the neural network models can convincingly emulate highly nonlinear audio distortion circuits, whilst running in real-time, with some models requiring only a relatively small amount of processing power to run on a modern desktop computer.

show abstract

Section: Convolutional Layermentioning

confidence: 99%

Section: Dataset Generationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Real-Time Guitar Amplifier Emulation with Deep Learning

Wright

Damskägg²,

Juvela³

et al. 2020

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…The block-oriented model [20] consists of a linear filter followed by a parametric nonlinearity. It was used for modeling the Bassman preamp in [21]. Here, the linear filters and the parameters of the nonlinearity are optimized jointly with the Adam optimizer, using a generated dataset consisting of sinusoids processed with the SPICE model.…”

Section: Discussionmentioning

confidence: 99%

Deep Learning for Tube Amplifier Emulation

Damskägg

Juvela

Thuillier

et al. 2019

ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

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.

show abstract

“…In "black-box" modelling the inputoutput relationship of the device is emulated directly, by using data measured from the device [7,8]. "Grey-box" models fall somewhere between these two approaches [9,10].…”

Section: Introductionmentioning

confidence: 99%

Perceptual loss function for neural modeling of audio systems

Wright

Välimäki

2020

ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

This work investigates alternate pre-emphasis filters used as part of the loss function during neural network training for nonlinear audio processing. In our previous work, the errorto-signal ratio loss function was used during network training, with a first-order highpass pre-emphasis filter applied to both the target signal and neural network output. This work considers more perceptually relevant pre-emphasis filters, which include lowpass filtering at high frequencies. We conducted listening tests to determine whether they offer an improvement to the quality of a neural network model of a guitar tube amplifier. Listening test results indicate that the use of an A-weighting pre-emphasis filter offers the best improvement among the tested filters. The proposed perceptual loss function improves the sound quality of neural network models in audio processing without affecting the computational cost.

show abstract

Gray-Box Modeling of Guitar Amplifiers

Cited by 20 publications

References 0 publications

Real-Time Guitar Amplifier Emulation with Deep Learning

Real-Time Guitar Amplifier Emulation with Deep Learning

Deep Learning for Tube Amplifier Emulation

Perceptual loss function for neural modeling of audio systems

Contact Info

Product

Resources

About