Download Fast Temporal Convolutions for Real-Time Audio Signal Processing This paper introduces the possibilities of optimizing neural network convolutional layers for modeling nonlinear audio systems and effects. Enhanced methods for real-time dilated convolutions are presented to achieve faster signal processing times than in previous work. Due to the improved implementation of convolutional layers, a significant decrease in computational requirements was observed and validated on different configurations of single layers with dilated convolutions and WaveNet-style feedforward neural network models. In most cases, equivalent signal processing times were achieved to those using recurrent neural networks with Long Short-Term Memory units and Gated Recurrent Units, which are considered state-of-the-art in the field of black-box virtual analog modeling.
Download Efficient Simulation of the Bowed String in Modal Form The motion of a bowed string is a typical nonlinear phenomenon resulting from a friction force via interaction with the bow. The system can be described using suitable differential equations. Implicit numerical discretisation methods are known to yield energy consistent algorithms, essential to ensure stability of the timestepping schemes. However, reliance on iterative nonlinear root finders carries significant implementation issues. This paper explores a method recently developed which allows nonlinear systems of ordinary differential equations to be solved non-iteratively. Case studies of a mass-spring system and an ideal string coupled with a bow are investigated. Finally, a stiff string with loss is also considered. Combining semi-discretisation and a modal approach results in an algorithm yielding faster than real-time simulation of typical musical strings.
Download Real-Time Implementation of the Dynamic Stiff String Using Finite-Difference Time-Domain Methods and the Dynamic Grid Digital musical instruments based on physical modelling have gained increased popularity over the past years. This is partly due to recent advances in computational power, which allow for their real-time implementation. One of the great potentials for digital musical instruments based on physical models, is that one can go beyond what is physically possible and change properties of the instruments which are static in real life. This paper presents a real-time implementation of the dynamic stiff string using finitedifference time-domain (FDTD) methods. The defining parameters of the string can be varied in real time and change the underlying grid that these methods rely on based on the recently developed dynamic grid method. For most settings, parameter changes are nearly instantaneous and do not cause noticeable artefacts due to changes in the grid. A reliable way to prevent artefacts for all settings is under development.
Download Physical Modeling Using Recurrent Neural Networks with Fast Convolutional Layers Discrete-time modeling of acoustic, mechanical and electrical systems is a prominent topic in the musical signal processing literature. Such models are mostly derived by discretizing a mathematical model, given in terms of ordinary or partial differential equations, using established techniques. Recent work has applied the techniques of machine-learning to construct such models automatically from data for the case of systems which have lumped states described by scalar values, such as electrical circuits. In this work, we examine how similar techniques are able to construct models of systems which have spatially distributed rather than lumped states. We describe several novel recurrent neural network structures, and show how they can be thought of as an extension of modal techniques. As a proof of concept, we generate synthetic data for three physical systems and show that the proposed network structures can be trained with this data to reproduce the behavior of these systems.
Download Control Parameters for Reed Wind Instruments or Organ Pipes with Reed Induced Flow Sound synthesis of a pipe coupled with a reed requires to finely tune the physical parameters of the underlying model. Although the pipe geometry is often well known, the 1 degree of freedom reed model’s parameters are effective coefficients (mass, section, etc) and are difficult to assess. Studies of this coupled system have essentially focused on models without the reed induced flow, and have exhibited two dimensionless parameters γ and ζ, which respectively describe the ratio between feeding pressure and closing reed pressure, and a dimensionless opening of the reed at rest. Including the reed flow in the model, then performing a scaling of the equations, leads to a new third dimensionless quantity that we will call κ. Varying the reed frequency with constant (γ, ζ, κ) on different pipe dimensions shows a certain stability of the model once put under this form. Using a real-time sound synthesis tool, the parameter space (γ, ζ, κ) is explored while the damping of the reed is also varied.
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 A Structural Similarity Index Based Method to Detect Symbolic Monophonic Patterns in Real-Time Automatic detection of musical patterns is an important task in the field of Music Information Retrieval due to its usage in multiple applications such as automatic music transcription, genre or instrument identification, music classification, and music recommendation. A significant sub-task in pattern detection is the realtime pattern detection in music due to its relevance in application domains such as the Internet of Musical Things. In this study, we present a method to identify the occurrence of known patterns in symbolic monophonic music streams in real-time. We introduce a matrix-based representation to denote musical notes using its pitch, pitch-bend, amplitude, and duration. We propose an algorithm based on an independent similarity index for each note attribute. We also introduce the Match Measure, which is a numerical value signifying the degree of the match between a pattern and a sequence of notes. We have tested the proposed algorithm against three datasets: a human recorded dataset, a synthetically designed dataset, and the JKUPDD dataset. Overall, a detection rate of 95% was achieved. The low computational load and minimal running time demonstrate the suitability of the method for real-world, real-time implementations on embedded systems.
Download Two Datasets of Room Impulse Responses for Navigation in Six Degrees-of-Freedom:a Symphonic Concert Hall and a Former Planetarium This paper presents two datasets of room impulse responses (RIRs) for navigable virtual acoustics. The first is a set of 240 mono and Ambisonic RIRs recorded at the Maison Symphonique, a symphonic concert hall in Montreal renowned for its great acoustic characteristics. The second is a set of 67 third-order Ambisonic RIRs which was recorded in the former planetarium of Montreal (currently known as the Centech), a space where the room acoustic includes an acoustic focal point where extreme reverberation times occur. The article first describes the two datasets and the methods that were used to capture them. A use case for these RIRs is then presented: an audio rendering of scene navigation using interpolation among RIRs.
Download On the Challenges of Embedded Real-Time Music Information Retrieval Real-time applications of Music Information Retrieval (MIR) have been gaining interest as of recently. However, as deep learning becomes more and more ubiquitous for music analysis tasks, several challenges and limitations need to be overcome to deliver accurate and quick real-time MIR systems. In addition, modern embedded computers offer great potential for compact systems that use MIR algorithms, such as digital musical instruments. However, embedded computing hardware is generally resource constrained, posing additional limitations. In this paper, we identify and discuss the challenges and limitations of embedded real-time MIR. Furthermore, we discuss potential solutions to these challenges, and demonstrate their validity by presenting an embedded real-time classifier of expressive acoustic guitar techniques. The classifier achieved 99.2% accuracy in distinguishing pitched and percussive techniques and a 99.1% average accuracy in distinguishing four distinct percussive techniques with a fifth class for pitched sounds. The full classification task is a considerably more complex learning problem, with our preliminary results reaching only 56.5% accuracy. The results were produced with an average latency of 30.7 ms.
Download Flutter Echo Modeling Flutter echo is a well-known acoustic phenomenon that occurs when sound waves bounce between two parallel reflective surfaces, creating a repetitive sound. In this work, we introduce a method to recreate flutter echo as an audio effect. The proposed algorithm is based on a feedback structure utilizing velvet noise that aims to synthesize the fluttery components of a reference room impulse response presenting flutter echo. Among these, the repetition time defines the length of the delay line in a feedback filter. The specific spectral properties of the flutter are obtained with a bandpass attenuation filter and a ripple filter, which enhances the harmonic behavior of the sound. Additional temporal shaping of a velvet-noise filter, which processes the output of the feedback loop, is performed based on the properties of the reference flutter. The comparison between synthetic and measured flutter echo impulse responses shows good agreement in terms of both the repetition time and reverberation time values.