Download Guitar Preamp Simulation Using Connection Currents This paper deals with a method of decomposition of a nonlinear audio circuit based on so called connection currents. These currents are used to connect inner blocks of the audio circuit with regards to preserve mutual interaction between adjoined blocks. Although this approach requires usage of numerical algorithm to solve the nonlinear equations, it reduces number of nonlinear equations to be solved if the solution of inner blocks is approximated while the accuracy of simulation is comparable to numerical solution of the whole nonlinear audio circuit.
Download WaveNet-Style Guitar Amplifier Model Pruning for Real-Time iOS Deployment WaveNet-style convolutional networks emulate tube amplifiers and distortion pedals with high fidelity, but their computational cost has confined them to desktops or dedicated DSP hardware. We present a sparse-enabled WaveNet inference engine for iOS that runs heavily pruned neural guitar amplifier models in real time on iPhones. Aggressive iterative magnitude pruning removes 90% of the network weights with no perceptible loss in quality. A custom sparse C++ engine turns this sparsity directly into compute savings, sustaining low-latency real-time operation on a CPU-only iPhone implementation where the dense model cannot. On-device output matches the trained model to within int16 quantization error. At the demonstration, visitors will play a guitar through the app on iPhone hardware and A/B the on-device pruned model against the physical pedal it emulates. Source code and audio examples are available online.
Download Simulating Idiomatic Playing Styles in a Classical Guitar Synthesizer: Rasgueado as a Case Study This paper presents our research efforts to synthesize complex instrumental gestures using a score-based control scheme. Our specific goal is to simulate the rasgueado technique that is popular especially in flamenco music. This technique is also used in the classical guitar repertoire. Rasgueado is especially challenging as ordinary music notation is not adequate to represent the dense stream of notes required for a convincing simulation. We will take two approaches to realize our task. First, we use the practical knowledge of how the actual performance is accomplished by the human player. A second, complementary, approach is to analyze an excerpt from real guitar playing. Our main focus here is to extract the onset times and the amplitudes of the recoded gesture. Next we combine the results from the two analysis steps using a constraintbased approach to find possible pitch and fingering sequences. Finally we translate the findings to our macro-note scheme that allows us to fill algorithmically a musical score.
Download Real-Time Detection of Finger Picking Musical Structures MIDIME is a software architecture that houses improvisational agents that react to MIDI messages from a finger-picked guitar. They operate in a pipeline whose first stage converts MIDI messages to a map of the state of instrument strings over time, and whose second stage selects rhythmic, modal, chordal, and melodic interpretations from the superposition of interpretations latent in the first stage. These interpretations are nondeterministic, not because of any arbitrary injection of randomness by an algorithm, but because guitar playing is nondeterministic. Variations in timing, tuning, picking intensity, string damping, and accidental or intensional grace notes can affect the selections of this second stage. The selections open to the second stage, as well as the third stage that matches second stage selections to a stored library of composition fragments, reflect the superposition of possible perceptions and interpretations of a piece of music. This paper concentrates on these working analytical stages of MIDIME. It also outlines plans for using the genetic algorithm to develop improvisational agents in the final pipeline stage.
Download Block-oriented Gray Box Modeling of Guitar Amplifiers In this work, analog guitar amplifiers are modeled with an automated procedure using iterative optimization techniques. The digital model is divided into functional blocks, consisting of lineartime-invariant (LTI) filters and nonlinear blocks with nonlinear mapping functions and memory. The model is adapted in several steps. First the filters are measured and afterwards the parameters of the digital model are adapted for different input signals to minimize the error between itself and the analog reference system. This is done for a small number of analog reference devices. Afterwards the adapted model is evaluated with objective scores and a listening test is performed to rate the quality of the adapted models.
Download Simplified Guitar Bridge Model for the Displacement Wave Representation in Digital Waveguides In this paper, we present a simplified model for the string-bridge interaction in guitars or other string instruments simulated by digital waveguides. The bridge model is devised for the displacement wave representation in order to be integrated with other models for string interactions with the player and with other parts of the instrument, whose simulation and implementation is easier in this representation. The model is based on a multiplierless scattering matrix representing the string-bridge interaction. Although not completely physically inspired, we show that this junction is sufficiently general to accommodate a variety of transfer functions under the sole requirement of passivity and avoids integration constants mismatch when the bridge is in turn modeled by a digital waveguide. The model is completed with simple methods to introduce horizontal and vertical polarizations of the string displacement and sympathetic vibrations of other strings. The aim of this paper is not to provide the most general methods for sound synthesis of guitar but, rather, to point at low computational cost and scalable solutions suitable for real-time implementations where the synthesizer is running together with several other audio applications.
Download Morphing Instrument Body Models In this study we present morphing methods for musical instrument body models using DSP techniques. These methods are able to transform a given body model gradually into another one in a controlled way, and they guarantee stability of the body models at each intermediate step. This enables to morph from a certain sized body model to a larger or smaller one. It is also possible to extrapolate beyond original models, thus creating new interesting (out of this world) instrument bodies. The opportunity to create a time-varying body, i.e., a model that changes in size over time, results in an interesting audio effect. This paper exhibits morphing mainly via guitar body examples, but naturally morphing can also be extended to other instruments with reverberant resonators as their bodies. Morphing from a guitar body model to a violin body model is viewed as an example. Implementation and perceptual issues of the signal processing methods are discussed. For related sound demonstrations, see www.acoustics.hut.fi/demo/ dafx2001-bodymorph/.
Download Simulation of Fender Type Guitar Preamp using Approximation and State-Space Model This paper deals with usage of approximations for simulation of more complex audio circuits. A Fender type guitar preamp was chosen as a case study. This circuit contains two tubes and thus four nonlinear functions as well as it is a parametric circuit because of an integrated tone stack. A state-space approach was used for simulation and further, precomputed solution is approximated using nonuniform cubic splines.
Download Sound Matching with a Differentiable Karplus-Strong Algorithm We present a self-supervised, event-based sound matching model using a differentiable extended Karplus-Strong algorithm. To avoid relying on external onset and fundamental frequency detectors, we explore training methodologies combining parameter losses on synthetic data with audio losses. We demonstrate that time-domain fractional delay interpolation provides gradient accuracy comparable to frequency-sampling while avoiding time-aliasing in highly resonant time-varying scenarios. Through systematic gradient analysis, we reveal that standard spectral losses provide no meaningful directional gradients for onset times, heavily degrading joint training. Training exclusively with parameter losses on synthetic data effectively learns fundamental frequency, timbral parameters, and onset times, but struggles to generalise to monophonic studio recordings of plucked guitar. External detectors combined with audio losses generalise best, isolating the model to timbre optimisation. While our Karplus-Strong decoder recovers interpretable parameters and naturally captures the transient characteristics of plucked guitar, Harmonics plus Noise baselines yield higher reconstruction fidelity by most metrics.
Download Structurally Passive Scattering Element for Modelling Guitar Pluck Action In this paper we propose new models for the plucking interaction of the player with the string for use with digital waveguide simulation of guitar. Unlike the previously presented models, the new models are based on structurally passive scattering junctions, which have the main advantage of being properly scaled for use in fixed-point waveguide implementations and of guaranteeing stability independently of the plucking excitation. In a first model we start from the Cuzzucoli-Lombardo equations [1], within the Evangelista-Eckerholm [2] propagation formulation, in order to derive the passive scattering junction by means of bilinear transformation. In a second model we start from equations properly modeling the finger compliance by means of a spring. In a third model we formalize the interaction in terms of driving impedances. The model is also extended using nonlinear (feathering) compliance models.