Open-Amp: Synthetic Data Framework for Audio Effect Foundation Models

Alec Wright; Alistair Carson; Lauri Juvela

REVIEW 4 major objections 4 minor 47 references

Open-Amp: Synthetic Data Framework for Audio Effect Foundation Models

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Open-Amp shows that a crowdsourced collection of learned emulations of guitar amps and pedals can be turned into synthetic training data that transfers to real audio-effects tasks, including classification and the emulation of unseen…

desk verdict A genuinely useful synthetic data framework built on crowd-sourced neural amp captures; the core idea is solid, but the SOTA claim is oversold and the evaluation protocol needs more detail before the numbers can be trusted. read the letter →

arxiv 2411.14972 v1 pith:REPOPHL7 submitted 2024-11-22 eess.AS cs.AIcs.LGcs.SD

Alec Wright , Alistair Carson , Lauri Juvela This is my paper · ORCID

classification eess.AScs.AIcs.LGcs.SD

keywords audioeffectssyntheticdataaugmentationguitaramplifiermodellingcontrastivelearningone-to-manyclassificationneuralemulation

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Open-Amp turns a crowd-sourced collection of neural emulations of guitar amplifiers and distortion pedals into a source of synthetic training audio. The paper's central claim is that machine-learning models trained on this synthetic data transfer to real audio-effects tasks: a guitar-effects encoder trained only on Open-Amp renders outperforms the published baselines on most evaluated classification settings and generalises across datasets better than encoders trained on the domain-specific datasets themselves. The paper also claims that a one-to-many guitar-effects model trained on Open-Amp synthetic devices can enrol previously unseen analog pedals through its learned embedding space, matching dedicated per-device models particularly when training data is scarce. If these claims hold, large-scale training for audio-effects models no longer requires physical access to hundreds of devices, since user-contributed captures can stand in for them.

What carries the argument

The load-bearing object is the Open-Amp rendering pipeline: a Python package that takes arbitrary input audio, runs it through a collection of user-uploaded recurrent neural-network captures of guitar amplifiers and pedals (59 amplifiers and 101 pedals, all single-layer LSTMs with 40 hidden units), and produces effect-processed audio online during training via multiprocess dataloading. Because the effects are software models rather than fixed recordings, the input signal can be chosen freely and augmented on the fly. The classification experiments use a contrastive learning objective in which two different clips processed by the same effect form a positive pair and clips from other effects form negatives. The emulation experiments use a temporal convolutional network with feature-wise linear modulation conditioned on a learned lookup table of 394 device embeddings, so enrolling a new device means freezing the network and fine-tuning only a new embedding vector.

What would settle it

A direct test would be to take a physical distortion pedal with no capture in the tone library, record clean/processed audio pairs, then compare enrolling that pedal into the Open-Amp foundation model against training a dedicated per-device model on the same pairs; if the enrolment route does not match the per-device model at comparable data sizes, the claimed transferability to unseen analog effects is refuted.

Watch

Extended reading notes

Core claim

The paper's discovery is that a diverse collection of learned emulations of real guitar gear can be used as a synthetic data source that transfers to real-device tasks. Concretely, the authors claim that a contrastively trained encoder, fed audio rendered through 160 user-created amplifier and pedal models, achieves new state-of-the-art accuracy on multiple guitar-effects classification benchmarks and beats encoders trained on the domain-specific datasets themselves. They further claim that a one-to-many guitar-effects model trained on 394 synthetic device variants can embed three unseen analog distortion pedals and reproduce their sound comparably to a dedicated one-to-one model trained on recordings of that same pedal, with the advantage growing when training data is scarce.

Load-bearing premise

The load-bearing premise is that the user-uploaded emulations in the tone-library collection actually reproduce the sound of the real amplifiers and pedals they claim to represent, and the paper reports no fidelity check; if many of them are inaccurate, the synthetic data will not reflect real hardware and the reported transfer may not generalise beyond this particular collection.

Editorial extensions

If this is right

The state-of-the-art classification results are achieved without the encoder ever seeing the benchmark's original recordings, so transfer from synthetic renders is the source of the performance.
The Open-Amp encoder generalises across datasets better than encoders trained on the target dataset itself, so the framework can serve as pretraining data.
The foundation-enrolment results show that an unseen analog pedal can be added to a model with only a fraction of the paired data a per-device model needs, which matters when hardware access or recording time is limited.
Because rendering happens online, the input signal can be any audio, so downstream tasks such as transcription or automatic mixing can draw effect-processed training material without committing to a fixed dataset.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the user-submitted captures are faithful, the key bottleneck for audio-effects foundation models shifts from collecting physical hardware to auditing capture quality and coverage; model zoos could be evaluated as data assets rather than rebuilt device by device.
The same embedding-enrolment recipe is plausibly testable on time-based effects such as delay, reverb, or modulation, where the contrastive encoder's effect identity would need to disentangle the dry input from time-varying processing.
A stronger external validation than the reported benchmarks would be to enrol directly recorded analog hardware that has no corresponding capture in the tone library and measure whether the trained embeddings still separate it correctly.
Because input audio is arbitrary, Open-Amp could be used to stress-test effect models with out-of-distribution signals, revealing whether learned effect embeddings are device-specific or partly input-dependent.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The central idea here is real: use crowd-sourced neural amp captures as a synthetic data source for audio effects. That is new, and it is useful. The Open-Amp package is released, it renders audio online during training, and the diversity of devices (160 captures from the GuitarML Proteus Tone Pack) is far beyond what Pedalboard or DASP offer for distortion. I think the paper deserves a serious referee.

The strongest evidence is the one-to-many emulation experiment. Pretraining a TCN on synthetic devices and then enrolling real analog pedals from EGFX works: with limited training data, the embedding-initialized models match or beat one-to-one baselines. That is a meaningful transfer result and it is not circular, because the target pedals are real analog hardware.

The classification experiments are more of a mixed bag. The contrastive encoder does beat FxNet on the GFX task overall (87.9 vs 86.9), with far fewer parameters and without seeing GFX during training. That is a fair data point. But the margin is 1 point, the KNN variant loses across the board, and the paper gives no seeds, error bars, or MLP optimization details. The stress-test note is right: the SOTA claim is not established as written.

Table II is the weakest part. The label spaces of GFX, EGFX, EGDB, and Open30 are incompatible (13, 12, 6, and 30 classes), yet the paper reports cross-dataset MLP accuracy without explaining how the classifier maps between them. That is a protocol gap, not a small omission. Also, Open30 is just a random subset of the same Proteus Tone Pack used to train the encoder, so the 99.8% accuracy there is not evidence of generalization to unseen data. The paper should stop calling it that.

One other thing: the paper says 65 parameter-conditioned models are each treated as five separate models, which should give 420 synthetic devices, not the stated 394. Minor arithmetic discrepancy, but worth fixing.

The unverified fidelity of user-uploaded LSTM captures is a real concern, but it is partly mitigated by the EGFX enrollment result. If the synthetic captures were badly biased, it is hard to see why the learned embedding would transfer to real analog pedals as well as it does.

Overall: the framework is a contribution, the experiments point in the right direction, and the missing details are fixable. I would send this to peer review, but with a clear request for code/data, error bars, an explicit account of the cross-dataset label mapping, and a more modest claim than "new state-of-the-art" until those are in place.

Referee Report

4 major / 4 minor

Summary. The paper introduces Open-Amp, a Python framework that generates synthetic audio effects data by rendering audio through crowdsourced neural network emulations of guitar amplifiers and effects pedals, sourced from the GuitarML Proteus Tone Pack. The authors use this framework to train a contrastive guitar-effects encoder and evaluate it on existing datasets (GFX, EGFX, EGDB) plus a synthetic Open30 set, reporting state-of-the-art classification accuracy versus the FxNet baseline. They also train a one-to-many TCN-based effects model with FiLM conditioning on synthetic devices and demonstrate enrollment of unseen analog EGFX pedals by fine-tuning only a learned embedding.

Significance. If the claims are upheld, this is a valuable contribution: it provides a scalable, open-source alternative to physical-device data collection for audio effects research, and the code release is a concrete asset to the community. The use of external datasets (GFX, EGFX, EGDB) for evaluation is a strength, as is the enrollment experiment on real analog pedals, which goes beyond purely synthetic validation. However, the central 'state-of-the-art' claim currently rests on an underspecified and in places internally inconsistent evaluation protocol (Tables I and II), and the reported gains are small and lack variance estimates. The idea is promising, but the evidence as presented is not yet sufficient to support the abstract's claims.

major comments (4)

[Section III-A, Table I] The evaluation protocol for the central classification claim is underspecified. The paper does not state how the train/test splits for the GFX subsets were constructed (e.g., per-clip random split, per-guitar or per-conditioning split, split ratio), how the KNN classifier was configured (k, distance metric, feature normalization), or how the MLP was optimized (optimizer, number of epochs, learning rate, regularization, early stopping, number of seeds averaged). The overall accuracy gain over the FxNet baseline is only 1.0 percentage point (87.9% vs 86.9%), the KNN is worse in all rows, and the MLP loses on two of eight rows. Without confidence intervals or significance tests, the claimed improvement is not established.
[Section III-A, Table II] The cross-dataset evaluation in Table II is not interpretable as described. The datasets have incompatible label spaces: GFX has 13 distortion classes, EGFX has 12 effects across distortion/reverb/modulation/delay, EGDB has 6 amplifiers, and Open30 has 30 devices. The text says an MLP classifier is fit with 'various permutations of training and test set,' but it never specifies how a classifier trained on, for example, 13 GFX classes can be evaluated on 12 EGFX classes or 6 EGDB classes. No class mapping or relabeling procedure is given. If the intended procedure is instead to train an encoder on the source dataset and then train a separate MLP on the target dataset, that needs to be stated explicitly; as written, the numbers in Table II do not support the claim that the Open-Amp encoder 'performs best on all datasets.'
[Section III-A, Open30] The 99.8% Open-Amp-to-Open30 result is not evidence of transferability. Open30 consists of 30 devices 'randomly selected' from the same Proteus Tone Pack used to train the Open-Amp encoder, and the paper does not state that these 30 devices were excluded from the contrastive training set. Without a clear hold-out guarantee, the near-perfect accuracy likely reflects that the encoder was trained on the same device identities. This result should be either removed or re-run with explicitly held-out devices.
[Section III, first paragraph] The framework's validity depends on the fidelity of the crowdsourced LSTM captures as emulations of the physical devices they are named after, yet the paper provides no verification of this fidelity. All models are described as single-layer LSTMs with hidden size 40, but no comparison of any capture against a real device recording is reported, and no reference to existing validation of GuitarML captures is given. The authors should either provide a quantitative fidelity check (e.g., ESR or spectral loss for a capture versus the real device) or explicitly discuss this as a limitation, since the transferability claims rest on the assumption that these synthetic models adequately represent real analog and digital effects.

minor comments (4)

[Section III-A, Table II] The table would benefit from a note that the Open-Amp row uses the 100,000-iteration encoder, not the 200,000-iteration encoder from Table I; this is mentioned in the text but is easy to miss and could confuse readers.
[Section III-B, Table III] The selection of the five devices in Table III is based on the Emb-64 model's combined loss; this selection procedure should be stated in the caption or text so the reader understands that the displayed spread is not uniformly random across all 394 devices.
[Section III-B, Fig. 2] The enrollment results in Fig. 2 show no error bars or statistical significance tests; since the differences between the baseline and embedding methods are sometimes small (especially at low data fractions), the authors should add variance information or state how many runs were averaged.
[General] The term 'Open30' is close to 'Open-Amp' and could be confused with the framework itself; consider renaming it to something like 'Proteus-30' or 'OpenAmp-30' for clarity.

Circularity Check

1 steps flagged · score 2.0 of 10

External benchmarks (GFX, EGFX, EGDB) make the main transfer claims non-circular; only the internal Open30 ablation is same-distribution and adds no generalization evidence.

fitted input called prediction [Section III-A, Table II and preceding paragraph (Open30 ablation)]
"As a further ablation we carry out additional classification tasks, on four different datasets, the previously introduced EGFX, EGDB and GFX datasets (see Sec. II-B), as well as 'Open30', a dataset consisting of guitar audio processed by 30 randomly selected devices from the Proteus Tone Pack, synthesised using Open-Amp. ... The results show that the encoder trained using Open-Amp performs best on all datasets, further demonstrating transferability to unseen data."

The Open-Amp encoder is trained on the same 'Proteus Tone Packs' collection used to render Open30; no holdout of the 30 devices from contrastive training is stated. Because the contrastive objective is to separate devices, classifying Open30 clips is a same-distribution recognition task and the 99.8% Open-Amp-to-Open30 result is expected rather than evidence of generalisation. This is a minor internal ablation; the external GFX, EGFX and EGDB evaluations do not share this circularity.

full rationale

The central classification claim is benchmarked against the GFX dataset and the FxNet baseline; the paper explicitly states the Open-Amp encoder has not seen GFX data, so the MLP/KNN evaluation is a genuine transfer test. The one-to-many emulation claim is tested on real analog EGFX pedals (Rat, BD-2, TS-mini) and compared against per-device one-to-one TCN baselines, so it is not equivalent to the synthetic training input. Self-citations ([7], [8], [28], [37], [44], [46] and others) supply background architectures and sampling methods, not the target result, and are not load-bearing in a circular sense. The only same-distribution element is the Open30 ablation in Table II, which samples from the same Proteus Tone Pack used for Open-Amp training without a stated holdout; this makes the 99.8% figure internally unsurprising, but it does not drive the paper's main external-benchmark conclusions. Hence the overall circularity burden is low.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

No new physical entities are introduced. The framework rests on the fidelity of crowd-sourced neural emulations, the representativeness of the input audio, and the validity of contrastive learning with device identity as the supervisory signal.

free parameters (2)

Conditioning parameter spacing = 5 values linearly spaced from 0 to 1
Models with a single conditioning parameter (typically gain) were expanded into five separate synthetic devices; this choice is ad hoc and directly shapes the one-to-many training set (Section III-B).
Embedding dimension = 16, 64, 256
The one-to-many model uses embedding dimensions of 16, 64, and 256; performance improves with dimension, so the headline result of being comparable to one-to-one depends on the E=256 model (Table III).

assumptions (3)

domain assumption The crowd-sourced Proteus Tone Pack models are accurate emulations of real guitar amplifiers and pedals.
The entire synthetic data pipeline assumes that user-uploaded LSTM captures faithfully reproduce the timbre and nonlinearity of the target devices; Section III states these models are used without independent verification.
domain assumption The IDMT-Guitar dataset provides a representative distribution of electric guitar input signals for downstream tasks.
All synthetic data for the encoder and the one-to-many model is rendered from unprocessed IDMT guitar audio (Section III).
domain assumption Contrastive learning with device identity as the class produces embeddings that organize audio effects by device timbre.
The SimCLR-style objective (Section III-A) assumes that two different clips through the same device form a valid positive pair and that device identity is a sufficient label for transferable effect representations.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Open-Amp: Synthetic Data Framework for Audio Effect Foundation Models." pith.science (2026). https://pith.science/paper/REPOPHL7

@misc{pith2026241114972,
  author       = {Pith},
  title        = {Pith review of: Open-Amp: Synthetic Data Framework for Audio Effect Foundation Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/REPOPHL7}},
  note         = {Machine review of arXiv:2411.14972}
}

read the original abstract

This paper introduces Open-Amp, a synthetic data framework for generating large-scale and diverse audio effects data. Audio effects are relevant to many musical audio processing and Music Information Retrieval (MIR) tasks, such as modelling of analog audio effects, automatic mixing, tone matching and transcription. Existing audio effects datasets are limited in scope, usually including relatively few audio effects processors and a limited amount of input audio signals. Our proposed framework overcomes these issues, by crowdsourcing neural network emulations of guitar amplifiers and effects, created by users of open-source audio effects emulation software. This allows users of Open-Amp complete control over the input signals to be processed by the effects models, as well as providing high-quality emulations of hundreds of devices. Open-Amp can render audio online during training, allowing great flexibility in data augmentation. Our experiments show that using Open-Amp to train a guitar effects encoder achieves new state-of-the-art results on multiple guitar effects classification tasks. Furthermore, we train a one-to-many guitar effects model using Open-Amp, and use it to emulate unseen analog effects via manipulation of its learned latent space, indicating transferability to analog guitar effects data.

Figures

Figures reproduced from arXiv: 2411.14972 by the authors.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 43 canonical work pages

[1]

Time-variant gray-box model- ing of a phaser pedal,

R. Kiiski, F. Esqueda, and V . V ¨alim¨aki, “Time-variant gray-box model- ing of a phaser pedal,” in Proc. Int. Conf. Digital Audio Effects (DAFx) , Brno, Czech Republic, Sep 2016, pp. 31–38

work page 2016
[2]

Simulation of analog flanger effect using bbd circuit,

J. Ma ˇc´ak, “Simulation of analog flanger effect using bbd circuit,” in Proc. Int. Conf. Digital Audio Effects (DAFx) , Brno, Czech Republic, Sep 2016, pp. 99–102

work page 2016
[3]

Modeling of an optocoupler-based audio dynamic range control circuit,

F. Eichas and U. Z ¨olzer, “Modeling of an optocoupler-based audio dynamic range control circuit,” in Novel Optical Systems Design and Optimization XIX, vol. 9948. SPIE, 2016, pp. 47–62

work page 2016
[4]

Lightweight and inter- pretable neural modeling of an audio distortion effect using hypercondi- tioned differentiable biquads,

S. Nercessian, A. Sarroff, and K. J. Werner, “Lightweight and inter- pretable neural modeling of an audio distortion effect using hypercondi- tioned differentiable biquads,” in Proc. IEEE ICASSP, Toronto, Canada, June 2021, pp. 890–894

work page 2021
[5]

Wave digital simulation of a vacuum- tube amplifier,

M. Karjalainen and J. Pakarinen, “Wave digital simulation of a vacuum- tube amplifier,” in Proc. IEEE ICASSP , Toulouse, France, May 2006, pp. 153–156

work page 2006
[6]

Deep learning for tube amplifier emulation,

E.-P. Damsk ¨agg, L. Juvela, E. Thuillier, and V . V¨alim¨aki, “Deep learning for tube amplifier emulation,” in Proc. IEEE ICASSP , Brighton, UK, May 2019, pp. 471–475

work page 2019
[7]

Real-time black-box modelling with recurrent neural networks,

A. Wright, E.-P. Damsk ¨agg, and V . V ¨alim¨aki, “Real-time black-box modelling with recurrent neural networks,” in Proc. Int. Conf. Digital Audio Effects (DAFx) , Birmingham, UK, Sep. 2019, pp. 173–180

work page 2019
[8]

Real-time guitar amplifier emulation with deep learning,

A. Wright, E.-P. Damsk ¨agg, L. Juvela, and V . V ¨alim¨aki, “Real-time guitar amplifier emulation with deep learning,” Appl. Sci., vol. 10, no. 3, 2020

work page 2020

Show all 47 references

[9]

A review of neural network-based emulation of guitar amplifiers,

T. Vanhatalo, P. Legrand, M. Desainte-Catherine et al. , “A review of neural network-based emulation of guitar amplifiers,” Appl. Sci., vol. 12, no. 12, p. 5894, 2022

2022
[10]

Deep learning for audio effects modeling,

M. A. Mart ´ınez Ram´ırez, “Deep learning for audio effects modeling,” Ph.D. dissertation, Queen Mary University of London, London, UK, Nov. 2020

2020
[11]

Exposure bias and state matching in recurrent neural network virtual analog models,

A. Peussa, E.-P. Damsk ¨agg, T. Sherson, S. I. Mimilakis, L. Juvela, A. Gotsopoulos, and V . V ¨alim¨aki, “Exposure bias and state matching in recurrent neural network virtual analog models,” in Proc. Int. Conf. Digital Audio Effects (DAFx) , 2021, pp. 284–291

2021
[12]

Neural modeling of phaser and flanging effects,

A. Wright and V . V ¨alim¨aki, “Neural modeling of phaser and flanging effects,” J. Audio Eng. Soc. , vol. 69, no. 7/8, pp. 517–529, Jul. 2021

2021
[13]

Differentiable grey-box modelling of phaser effects using frame-based spectral pro- cessing,

A. Carson, C. Valentini-Botinhao, S. King, and S. Bilbao, “Differentiable grey-box modelling of phaser effects using frame-based spectral pro- cessing,” in Proc. Int. Conf. Digital Audio Effects (DAFx) , Copenhagen, Denmark, Sep 2023, pp. 219–226

2023
[14]

Neural modeling of magnetic tape recorders,

O. Mikkonen, A. Wright, E. Moliner, and V . V¨alim¨aki, “Neural modeling of magnetic tape recorders,” in Proc. Int. Conf. Digital Audio Effects (DAFx), Copenhagen, Denmark, Sep 2023, pp. 196–203

2023
[15]

Neural modelling of audio effects,

A. Wright, “Neural modelling of audio effects,” Ph.D. dissertation, Aalto University, Espoo, Finland, Dec 2023

2023
[16]

GuitarML

K. Bloemer, “GuitarML.” [Online]. Available: https://guitarml.com/
[17]

Neural amp modeler

S. Atkinson, “Neural amp modeler.” [Online]. Available: https: //www.neuralampmodeler.com/
[18]

Style transfer of audio effects with differentiable signal processing,

C. J. Steinmetz, N. J. Bryan, and J. D. Reiss, “Style transfer of audio effects with differentiable signal processing,” J. Audio Eng. Soc., vol. 70, no. 9, pp. 708–721, Sep. 2022

2022
[19]

Music mixing style transfer: A contrastive learning approach to disentangle audio effects,

J. Koo, M. A. Mart ´ınez-Ram´ırez, W.-H. Liao, S. Uhlich, K. Lee, and Y . Mitsufuji, “Music mixing style transfer: A contrastive learning approach to disentangle audio effects,” in Proc. IEEE ICASSP , 2023, pp. 1–5

2023
[20]

General purpose audio effect removal,

M. Rice, C. J. Steinmetz, G. Fazekas, and J. D. Reiss, “General purpose audio effect removal,” inProc. IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA) , 2023, pp. 1–5

2023
[21]

Audio effect chain estimation and dry signal recovery from multi-effect-processed musical signals,

O. Take, K. Watanabe, T. Nakatsuka, T. Cheng, T. Nakano, M. Goto, S. Takamichi, and H. Saruwatari, “Audio effect chain estimation and dry signal recovery from multi-effect-processed musical signals,” in Proc. Int. Conf. Digital Audio Effects (DAFx) , Guildford, Surrey, Sep 202...

2024
[22]

Distortion recovery: A two-stage method for guitar effect removal,

Y .-S. Lee, Y .-P. Peng, J.-T. Wu, M. Cheng, L. Su, and Y .-H. Yang, “Distortion recovery: A two-stage method for guitar effect removal,” in Proc. Int. Conf. Digital Audio Effects (DAFx) , Guildford, United Kingdom, Sep 2024, pp. 177–184

2024
[23]

Towards zero-shot amplifier modeling: One-to- many amplifier modeling via tone embedding control,

Y .-H. Chen, Y .-T. Yeh, Y .-C. Cheng, J.-T. Wu, Y .-H. Ho, J.-S. R. Jang, and Y .-H. Yang, “Towards zero-shot amplifier modeling: One-to- many amplifier modeling via tone embedding control,” arXiv preprint arXiv:2407.10646, 2024

2024 arXiv
[24]

Towards automatic transcription of polyphonic electric guitar music: A new dataset and a multi-loss transformer model,

Y .-H. Chen, W.-Y . Hsiao, T.-K. Hsieh, J.-S. R. Jang, and Y .-H. Yang, “Towards automatic transcription of polyphonic electric guitar music: A new dataset and a multi-loss transformer model,” in Proc. IEEE ICASSP, 2022, pp. 786–790

2022
[25]

Synthtab: Leveraging synthesized data for guitar tablature transcription,

Y . Zang, Y . Zhong, F. Cwitkowitz, and Z. Duan, “Synthtab: Leveraging synthesized data for guitar tablature transcription,” in Proc. IEEE ICASSP, 2024, pp. 1286–1290

2024
[26]

Introducing a dataset of guitar amplifier sounds for nonlinear emulation benchmarking,

T. Schmitz and J.-J. Embrechts, “Introducing a dataset of guitar amplifier sounds for nonlinear emulation benchmarking,” in Proc. 145th Audio Eng. Soc. Conv. New York, USA, Oct. 2018

2018
[27]

Profiling audio compressors with deep neural networks,

S. Hawley, B. Colburn, and S. I. Mimilakis, “Profiling audio compressors with deep neural networks,” in Proc. 147th Audio Eng, Soc. Conv., New York, USA, Oct. 2019, Conference Paper

2019
[28]

End-to-end amp mod- eling: from data to controllable guitar amplifier models,

L. Juvela, E.-P. Damsk ¨agg, A. Peussa, J. M ¨akinen, T. Sherson, S. I. Mimilakis, K. Rauhanen, and A. Gotsopoulos, “End-to-end amp mod- eling: from data to controllable guitar amplifier models,” in Proc. IEEE ICASSP, 2023, pp. 1–5

2023
[29]

Pedalboard,

P. Sobot, “Pedalboard,” Jul. 2021. [Online]. Available: https://doi.org/ 10.5281/zenodo.7817838

2021 doi
[30]

Differentiable audio signal processors in pytorch,

C. Steinmetz, “Differentiable audio signal processors in pytorch,” Jan
[31]

DDSP: Differentiable digital signal processing,

J. Engel, L. Hantrakul, C. Gu, and A. Roberts, “DDSP: Differentiable digital signal processing,” in Proc. Int. Conf. Learning Representations (ICLR), Addis Ababa, Ethiopia, 2020

2020
[32]

Torchaudio 2.1: Advancing speech recognition, self-supervised learning, and audio processing components for pytorch,

J. Hwang, M. Hira, C. Chen, X. Zhang, Z. Ni, G. Sun, P. Ma, R. Huang, V . Pratap, Y . Zhang, A. Kumar, C.-Y . Yu, C. Zhu, C. Liu, J. Kahn, M. Ravanelli, P. Sun, S. Watanabe, Y . Shi, Y . Tao, R. Scheibler, S. Cornell, S. Kim, and S. Petridis, “Torchaudio 2.1: Advancing speech ...

2023
[33]

Black-box modeling of distortion circuits with block-oriented models,

F. Eichas and U. Z ¨olzer, “Black-box modeling of distortion circuits with block-oriented models,” in Proc. Int. Conf. Digital Audio Effects (DAFx- 16), Brno, Czech Republic, Sept. 2016, pp. 39–45

2016
[34]

Guitar effects recognition and parameter estimation with convolutional neural networks,

M. Comunit `a, D. Stowell, and J. D. Reiss, “Guitar effects recognition and parameter estimation with convolutional neural networks,” J. Audio Eng. Soc., vol. 69, no. 7/8, pp. 594–604, July 2021

2021
[35]

EGFxSet: Electric guitar tones processed through real effects of distortion, modulation, delay and reverb,

H. Pedroza, G. Meza, and I. R. Roman, “EGFxSet: Electric guitar tones processed through real effects of distortion, modulation, delay and reverb,” ISMIR Late Breaking Demo , 2022

2022
[36]

Amp-space: A large-scale dataset for fine-grained timbre transformation,

J. Naradowsky, “Amp-space: A large-scale dataset for fine-grained timbre transformation,” in Proc. Int. Conf. Digital Audio Effects (DAFx), Vienna, Austria, Sep. 2021, pp. 57–64

2021
[37]

Sampling the user controls in neural modeling of audio devices,

O. Mikkonen, A. Wright, and V . V ¨alim¨aki, “Sampling the user controls in neural modeling of audio devices,” EURASIP J. Audio, Speech, and Music Processing, vol. 2024, no. 1, p. 26, 2024

2024
[38]

Automatic tablature transcription of electric guitar recordings by estimation of score- and instrument-related parameters,

C. Kehling, J. Abeßer, C. Dittmar, and G. Schuller, “Automatic tablature transcription of electric guitar recordings by estimation of score- and instrument-related parameters,” in Proc. Int. Conf. Digital Audio Effects (DAFx), Erlangen, Germany, Sep. 2014, pp. 219–226

2014
[39]

A simple framework for contrastive learning of visual representations,

T. Chen, S. Kornblith, M. Norouzi, and G. Hinton, “A simple framework for contrastive learning of visual representations,” in Proc. Int. Conf. Machine Learning. PMLR, 2020, pp. 1597–1607

2020
[40]

End-to-end music remastering system using self-supervised and adversarial training,

J. Koo, S. Paik, and K. Lee, “End-to-end music remastering system using self-supervised and adversarial training,” in Proc. IEEE ICASSP , 2022, pp. 4608–4612

2022
[41]

Efficient neural networks for real-time modeling of analog dynamic range compression,

C. Steinmetz and J. Reiss, “Efficient neural networks for real-time modeling of analog dynamic range compression,” in Proc. 152nd Audio Eng. Soc. Conv., The Hague, Netherlands, May 2022

2022
[42]

Modelling black- box audio effects with time-varying feature modulation,

M. Comunit `a, C. Steinmetz, H. Phan, and J. Reiss, “Modelling black- box audio effects with time-varying feature modulation,” in Proc. IEEE ICASSP, Rhodes, Greece, 2023

2023
[43]

auraloss: Audio focused loss functions in PyTorch,

C. J. Steinmetz and J. D. Reiss, “auraloss: Audio focused loss functions in PyTorch,” in Digital Music Research Network One-day Workshop (DMRN+15), 2020

2020
[44]

Sample rate independent recurrent neural networks for audio effects processing,

A. Carson, A. Wright, V . V ¨alim¨aki, J. Chowdhury, and S. Bilbao, “Sample rate independent recurrent neural networks for audio effects processing,” in Proc. Int. Conf. Digital Audio Effects (DAFx), Guildford, Surrey, Sep 2024, pp. 17–24

2024
[45]

Discretization of the’59 fender bassman tone stack,

D. T. Yeh and J. O. Smith, “Discretization of the’59 fender bassman tone stack,” in Proc. Int. Conf. Digital Audio Effects (DAFx) , Montreal, Canada, 2006, pp. 1–8

2006
[46]

Neural grey- box guitar amplifier modelling with limited data,

ˇS. Mikl ´anek, A. Wright, V . V ¨alim¨aki, and J. Schimmel, “Neural grey- box guitar amplifier modelling with limited data,” in Proc. Int. Conf. Digital Audio Effects (DAFx) , Copenhagen, Denmark, Sep 2023, pp. 151–158

2023
[2023]

Available: https://github.com/csteinmetz1/dasp-pytorch

[Online]. Available: https://github.com/csteinmetz1/dasp-pytorch

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

Time-variant gray-box model- ing of a phaser pedal,

R. Kiiski, F. Esqueda, and V . V ¨alim¨aki, “Time-variant gray-box model- ing of a phaser pedal,” in Proc. Int. Conf. Digital Audio Effects (DAFx) , Brno, Czech Republic, Sep 2016, pp. 31–38

work page 2016

[2] [2]

Simulation of analog flanger effect using bbd circuit,

J. Ma ˇc´ak, “Simulation of analog flanger effect using bbd circuit,” in Proc. Int. Conf. Digital Audio Effects (DAFx) , Brno, Czech Republic, Sep 2016, pp. 99–102

work page 2016

[3] [3]

Modeling of an optocoupler-based audio dynamic range control circuit,

F. Eichas and U. Z ¨olzer, “Modeling of an optocoupler-based audio dynamic range control circuit,” in Novel Optical Systems Design and Optimization XIX, vol. 9948. SPIE, 2016, pp. 47–62

work page 2016

[4] [4]

Lightweight and inter- pretable neural modeling of an audio distortion effect using hypercondi- tioned differentiable biquads,

S. Nercessian, A. Sarroff, and K. J. Werner, “Lightweight and inter- pretable neural modeling of an audio distortion effect using hypercondi- tioned differentiable biquads,” in Proc. IEEE ICASSP, Toronto, Canada, June 2021, pp. 890–894

work page 2021

[5] [5]

Wave digital simulation of a vacuum- tube amplifier,

M. Karjalainen and J. Pakarinen, “Wave digital simulation of a vacuum- tube amplifier,” in Proc. IEEE ICASSP , Toulouse, France, May 2006, pp. 153–156

work page 2006

[6] [6]

Deep learning for tube amplifier emulation,

E.-P. Damsk ¨agg, L. Juvela, E. Thuillier, and V . V¨alim¨aki, “Deep learning for tube amplifier emulation,” in Proc. IEEE ICASSP , Brighton, UK, May 2019, pp. 471–475

work page 2019

[7] [7]

Real-time black-box modelling with recurrent neural networks,

A. Wright, E.-P. Damsk ¨agg, and V . V ¨alim¨aki, “Real-time black-box modelling with recurrent neural networks,” in Proc. Int. Conf. Digital Audio Effects (DAFx) , Birmingham, UK, Sep. 2019, pp. 173–180

work page 2019

[8] [8]

Real-time guitar amplifier emulation with deep learning,

A. Wright, E.-P. Damsk ¨agg, L. Juvela, and V . V ¨alim¨aki, “Real-time guitar amplifier emulation with deep learning,” Appl. Sci., vol. 10, no. 3, 2020

work page 2020