Paper Archive

We release synth1B1, a multi-modal audio corpus consisting of 1 billion 4-second synthesized sounds, paired with the synthesis parameters used to generate them. The dataset is 100x larger than any audio dataset in the literature. We also introduce torchsynth, an open source modular synthesizer that generates the synth1B1 samples on-the-fly at 16200x faster than real-time (714MHz) on a single GPU. Finally, we release two new audio datasets: FM synth timbre and subtractive synth pitch. Using these datasets, we demonstrate new rank-based evaluation criteria for existing audio representations. Finally, we propose a novel approach to synthesizer hyperparameter optimization.

Download

Controllable timbre synthesis has been a subject of research for several decades, and deep neural networks have been the most successful in this area. Deep generative models such as Variational Autoencoders (VAEs) have the ability to generate a high-level representation of audio while providing a structured latent space. Despite their advantages, the interpretability of these latent spaces in terms of human perception is often limited. To address this limitation and enhance the control over timbre generation, we propose a regularized VAE-based latent space that incorporates timbre descriptors. Moreover, we suggest a more concise representation of sound by utilizing its harmonic content, in order to minimize the dimensionality of the latent space.

Download

Recent advances in data-driven expressive performance rendering have enabled automatic models to reproduce the characteristics and the variability of human performances of musical compositions. However, these models need to be trained with aligned pairs of scores and performances and they rely notably on score-specific markings, which limits their scope of application. This work tackles the piano performance rendering task in a low-informed setting by only considering the score note information and without aligned data. The proposed model relies on an adversarial training where the basic score notes properties are modified in order to reproduce the expressive qualities contained in a dataset of real performances. First results for unaligned score-to-performance rendering are presented through a conducted listening test. While the interpretation quality is not on par with highly-supervised methods and human renditions, our method shows promising results for transferring realistic expressivity into scores.

Download

Conventional Deep Learning (DL) approaches to Infinite Impulse Response (IIR) filter coefficients estimation from arbitrary frequency response are quite limited. They often suffer from inefficiencies such as tight training requirements, high complexity, and limited accuracy. As an alternative, in this paper, we explore the use of Kolmogorov-Arnold Networks (KANs) to predict the IIR filter—specifically biquad coefficients—effectively. By leveraging the high interpretability and accuracy of KANs, we achieve smooth coefficients’ optimization. Furthermore, by constraining the search space and exploring different loss functions, we demonstrate improved performance in speed and accuracy. Our approach is evaluated against other existing differentiable IIR filter solutions. The results show significant advantages of KANs over existing methods, offering steadier convergences and more accurate results. This offers new possibilities for integrating digital infinite impulse response (IIR) filters into deep-learning frameworks.

Download

Advances in deep learning have led to novel, state-of-the-art techniques for blind source separation, particularly for the application of non-stationary noise removal from speech. In this paper, we show how a simple reformulation allows us to adapt blind source separation techniques to the problem of speech dereverberation and, accordingly, train a bidirectional recurrent neural network (BRNN) for this task. We compare the performance of the proposed neural network approach with that of a baseline dereverberation algorithm based on spectral subtraction. We find that our trained neural network quantitatively and qualitatively outperforms the baseline approach.

Download

In this work, we present a data-driven approach to modeling tone stack circuits in guitar amplifiers and distortion pedals. To this aim, the proposed modeling approach uses a feedforward fully connected neural network to predict the parameters of a coupledform state-space filter, ensuring the numerical stability of the resulting time-varying system. The neural network is conditioned on the tone controls of the target tone stack and is optimized jointly with the coupled-form state-space filter to match the target frequency response. To assess the proposed approach, we model three popular tone stack schematics with both matched-order and overparameterized filters and conduct an objective comparison with well-established approaches that use cascaded biquad filters. Results from the conducted experiments demonstrate improved accuracy of the proposed modeling approach, especially in the case of over-parameterized state-space filters while guaranteeing numerical stability. Our method can be deployed, after training, in realtime audio processors.

Download

The sound of magnetic recording media, such as open-reel and cassette tape recorders, is still sought after by today’s sound practitioners due to the imperfections embedded in the physics of the magnetic recording process. This paper proposes a method for digitally emulating this character using neural networks. The signal chain of the proposed system consists of three main components: the hysteretic nonlinearity and filtering jointly produced by the magnetic recording process as well as the record and playback amplifiers, the fluctuating delay originating from the tape transport, and the combined additive noise component from various electromagnetic origins. In our approach, the hysteretic nonlinear block is modeled using a recurrent neural network, while the delay trajectories and the noise component are generated using separate diffusion models, which employ U-net deep convolutional neural networks. According to the conducted objective evaluation, the proposed architecture faithfully captures the character of the magnetic tape recorder. The results of this study can be used to construct virtual replicas of vintage sound recording devices with applications in music production and audio antiquing tasks.

Download

This paper deals with the automatic and robust analysis, and the realistic and low-cost synthesis of percussive drilling like sounds. The two contributions are: a non-supervised removal of quasistationary background noise based on the Non-negative Matrix Factorization, and a granular method for analysis/synthesis of this drilling sounds. These two points are appropriate to the acoustical properties of percussive drilling sounds, and can be extended to other sounds with similar characteristics. The context of this work is the training of operators of working machines using simulators. Additionally, an implementation is explained.

Download

Voice conversion has gained increasing popularity within the field of audio manipulation and speech synthesis. Often, the main objective is to transfer the input identity to that of a target speaker without changing its linguistic content. While current work provides high-fidelity solutions they rarely focus on model simplicity, high-sampling rate environments or stream-ability. By incorporating speech representation learning into a generative timbre transfer model, traditionally created for musical purposes, we investigate the realm of voice conversion generated directly in the time domain at high sampling rates. More specifically, we guide the latent space of a baseline model towards linguistically relevant representations and condition it on external speaker information. Through objective and subjective assessments, we demonstrate that the proposed solution can attain levels of naturalness, quality, and intelligibility comparable to those of a state-of-the-art solution for seen speakers, while significantly decreasing inference time. However, despite the presence of target speaker characteristics in the converted output, the actual similarity to unseen speakers remains a challenge.

Download

In this paper, we present an approach to the neural modeling of overdrive guitar pedals with conditioning from a cross-circuit and cross-setting latent space. The resulting network models the behavior of multiple overdrive pedals across different settings, offering continuous morphing between real configurations and hybrid behaviors. Compact conditioning spaces are obtained through unsupervised training of a variational autoencoder with adversarial training, resulting in accurate reconstruction performance across different sets of pedals. We then compare three Hyper-Recurrent architectures for processing, including dynamic and static HyperRNNs, and a smaller model for real-time processing. Additionally, we present pOD-set, a new open dataset including recordings of 27 analog overdrive pedals, each with 36 gain and tone parameter combinations totaling over 97 hours of recordings. Precise parameter setting was achieved through a custom-deployed recording robot.

Download