Download Differentiable Active Acoustics - Optimizing Stability via Gradient Descent Active acoustics (AA) refers to an electroacoustic system that actively modifies the acoustics of a room. For common use cases, the number of transducers—loudspeakers and microphones—involved in the system is large, resulting in a large number of system parameters. To optimally blend the response of the system into the natural acoustics of the room, the parameters require careful tuning, which is a time-consuming process performed by an expert. In this paper, we present a differentiable AA framework, which allows multi-objective optimization without impairing architecture flexibility. The system is implemented in PyTorch to be easily translated into a machine-learning pipeline, thus automating the tuning process. The objective of the pipeline is to optimize the digital signal processor (DSP) component to evenly distribute the energy in the feedback loop across frequencies. We investigate the effectiveness of DSPs composed of finite impulse response filters, which are unconstrained during the optimization. We study the effect of multiple filter orders, number of transducers, and loss functions on the performance. Different loss functions behave similarly for systems with few transducers and low-order filters. Increasing the number of transducers and the order of the filters improves results and accentuates the difference in the performance of the loss functions.
Download An Interdisciplinary Approach to Audio Effect Classification The aim of this paper is to propose an interdisciplinary classification of digital audio effects to facilitate communication and collaborations between DSP programmers, sound engineers, composers, performers and musicologists. After reviewing classifications reflecting technological, technical and perceptual points of view, we introduce a transverse classification to link disciplinespecific classifications into a single network containing various layers of descriptors, ranging from low-level features to high-level features. Simple tools using the interdisciplinary classification are introduced to facilitate the navigation between effects, underlying techniques, perceptual attributes and semantic descriptors. Finally, concluding remarks on implications for teaching purposes and for the development of audio effects user interfaces based on perceptual features rather than technical parameters are presented.
Download Sound texture synthesis using Convolutional Neural Networks The following article introduces a new parametric synthesis algorithm for sound textures inspired by existing methods used for visual textures. Using a 2D Convolutional Neural Network (CNN), a sound signal is modified until the temporal cross-correlations of the feature maps of its log-spectrogram resemble those of a target texture. We show that the resulting synthesized sound signal is both different from the original and of high quality, while being able to reproduce singular events appearing in the original. This process is performed in the time domain, discarding the harmful phase recovery step which usually concludes synthesis performed in the time-frequency domain. It is also straightforward and flexible, as it does not require any fine tuning between several losses when synthesizing diverse sound textures. Synthesized spectrograms and sound signals are showcased, and a way of extending the synthesis in order to produce a sound of any length is also presented. We also discuss the choice of CNN, border effects in our synthesized signals and possible ways of modifying the algorithm in order to improve its current long computation time.
Download Audio Nonlinear Modeling through Hyperbolic Tangent Functionals In the present contribution we present the preliminary results of a black box nonlinear system (NLS) modeling. The NLS is composed by a nonlinear sigmoid-type input-output relationship (NLTF) followed by a linear system (LTI), as in a Hammerstein nonlinear system. Here, the used NLTF is derived from a deformation of the Hyperbolic Tangent power expansion. The advantage of using the hyperbolic tangent function is that nonlinearity depends on the linear and cubic terms that measure curvature (and thus nonlinearity) of the transfer function. The hyperbolic tangent model is extended to other types of nonlinear systems by expanding the nonlinear system in linear and increasingly nonlinear contributions, where the expansion parameters are deformed to enhance or suppress specific nonlinear modes of the expansion. Simulations were performed using Matlab 2012a. The preliminary results show fairly good agreement between the system obtained by parametric inference and a reference system, with mean square error (MSE)=0.035.
Download A Frequency-Domain Reverberator Plug-In We present FDverb, a frequency-domain artificial reverberator, based on the idea of a vocoder with a noise carrier signal. Using a short-time Fourier transform (STFT) for analysis and synthesis, FDverb generates late reverberation by weighting spectral noise components with envelopes. We extend FDverb with early reflections, nonlinear decay, and pitch shifting. These extensions enable creative sound-design applications. We provide FDverb as an open-source DAW plug-in, using the JUCE framework.
Download Neural Net Tube Models for Wave Digital Filters Herein, we demonstrate the use of neural nets towards simulating multiport nonlinearities inside a wave digital filter. We introduce a resolved wave definition which allows us to extract features from a Kirchhoff domain dataset and train our neural networks directly in the wave domain. A hyperparameter search is performed to minimize error and runtime complexity. To illustrate the method, we model a tube amplifier circuit inspired by the preamplifier stage of the Fender Pro-Junior guitar amplifier. We analyze the performance of our neural nets models by comparing their distortion characteristics and transconductances. Our results suggest that activation function selection has a significant effect on the distortion characteristic created by the neural net.
Download Neural Sample-Based Piano Synthesis Piano sound emulation has been an active topic of research and development for several decades. Although comprehensive physicsbased piano models have been proposed, sample-based piano emulation is still widely utilized for its computational efficiency and
relative accuracy despite presenting significant memory storage
requirements. This paper proposes a novel hybrid approach to
sample-based piano synthesis aimed at improving the fidelity of
sound emulation while reducing memory requirements for storing samples. A neural network-based model processes the sound
recorded from a single example of piano key at a given velocity.
The network is trained to learn the nonlinear relationship between
the various velocities at which a piano key is pressed and the corresponding sound alterations. Results show that the method achieves
high accuracy using a specific neural architecture that is computationally efficient, presenting few trainable parameters, and it requires memory only for one sample for each piano key.
Download Fourier Neural Operators for Sample-Rate-Independent Virtual Analog Modeling Neural networks that operate directly on time-domain signals are widely used for virtual analog (VA) modeling. A key limitation of these models is their dependence on the sampling rate used during training, which becomes implicitly encoded in the learned parameters, so that changing it generally alters the realized dynamics. Although architectural modifications to recurrent neural networks have been proposed to enable sample-rate independent operation, these approaches are inherently tailored to upsampling and do not accommodate downsampling scenarios. In this manuscript, we present a VA modeling framework based on Fourier Neural Operators (FNOs) adapted to process fixed-duration audio frames. The proposed formulation defines the learned mapping over a fixed temporal support and evaluates it on uniform grids of different densities, so that a model trained at a single sampling rate can be applied at unseen sampling resolutions. Numerical results on a nonlinear transistor circuit show that the proposed model achieves competitive accuracy in upsampling scenarios while remaining directly applicable to downsampling, unlike a sample-rate independent baseline recurrent architecture.
Download Improving the robustness of the iterative solver in state-space modelling of guitar distortion circuitry Iterative solvers are required for the discrete-time simulation of nonlinear behaviour in analogue distortion circuits. Unfortunately, these methods are often computationally too expensive for realtime simulation. Two methods are presented which attempt to reduce the expense of iterative solvers. This is achieved by applying information that is derived from the specific form of the nonlinearity. The approach is first explained through the modelling of an asymmetrical diode clipper, and further exemplified by application to the Dallas Rangemaster Treble Booster guitar pedal, which provides an initial perspective of the performance on systems with multiple nonlinearities.
Download A parallel 3D digital wave guide mesh model with tetrahedral topology for room acoustic simulation Following a summary of the basic principles of 3D waveguide mesh modelling and the context of its application to room acoustic simulation, this paper presents a detailed analysis of the tetrahedral mesh topology and describes its implementation on a parallel computer model. Its structural characteristics are analysed, with particular emphasis on how they influence execution speed. Performance deterioration due to communication overhead in the parallelised model is discussed. Theoretical predictions are compared with data from performance tests carried out on different computer platforms and both are contrasted with the corresponding results from the rectilinear model, in order to assess the practical efficiency of the model. Objective validation tests are reported and discussed.