Download Synthetic Transaural Audio Rendering (STAR): a Perceptive Approach for Sound Spatialization The principles of Synthetic Transaural Audio Rendering (STAR) were first introduced at DAFx-06. This is a perceptive approach for sound spatialization, whereas state-of-the-art methods are rather physical. With our STAR method, we focus neither on the wave field (such as HOA) nor on the sound wave (such as VBAP), but rather on the acoustic paths traveled by the sound to the listener ears. The STAR method consists in canceling the cross-talk signals between two loudspeakers and the ears of the listener (in a transaural way), with acoustic paths not measured but computed by some model (thus synthetic). Our model is based on perceptive cues, used by the human auditory system for sound localization. The aim is to give the listener the sensation of the position of each source, and not to reconstruct the corresponding acoustic wave or field. This should work with various loudspeaker configurations, with a large sweet spot, since the model is neither specialized for a specific configuration nor individualized for a specific listener. Experimental tests have been conducted in 2015 and 2019 with different rooms and audiences, for still, moving, and polyphonic musical sounds. It turns out that the proposed method is competitive with the state-of-the-art ones. However, this is a work in progress and further work is needed to improve the quality.
Download Sound Matching Using Synthesizer Ensembles Sound matching allows users to automatically approximate existing sounds using a synthesizer. Previous work has mostly focused on algorithms for automatically programming an existing synthesizer. This paper proposes a system for selecting between different synthesizer designs, each one with a corresponding automatic programmer. An implementation that allows designing ensembles based on a template is demonstrated. Several experiments are presented using a simple subtractive synthesis design. Using an ensemble of synthesizer-programmer pairs is shown to provide better matching than a single programmer trained for an equivalent integrated synthesizer. Scaling to hundreds of synthesizers is shown to improve match quality.
Download Bit Bending: an Introduction We introduce the technique of "Bit Bending," a particularly fertile technique for circuit bending which involves short circuits and manipulations upon digital serial information. We present a justification for computer modeling of circuit-bent instruments, with deference to the movement's aversion to "theory-true" design and associations with chance discovery [1]. To facilitate software modeling of Bit Bending, we also present a software library for modeling certain classes of digital integrated circuits. A synthesis architecture case study (frequency modulation via numerically controlled oscillators) demonstrates software modeling of Bit Bending in action.
Download Toward the Perfect Audio Morph? Singing Voice Synthesis and Processing This paper reviews the popular methods and models used for the synthesis of the singing voice, discussing strengths and weaknesses of each technique. Then a brief review is given of research on cross-modal visual/auditory perception of the human voice. The paper concludes with comments related to the singing synthesis systems discussed, addressing multi-modal perception, audio morphing, and the categorical perception of sound.
Download Implementing Digital Audio Effects Using a Hardware/Software Co-Design Approach Digital realtime audio effects as of today are realized in software in almost all cases. The hardware platforms used for this purpose reach from multi purpose processors like the Intel Pentium class over embedded processors (e.g. the ARM family) to specialized DSP. The upcoming technology of complete systems on a single programmable chip contrasts such a software centric solution, because it combines software and hardware via some co-design methodology and makes for a promising alternative for the future of realtime audio. Such systems are able to combine the vast amount of computing power provided by dedicated hardware with the flexibility offered by software in a way the designer is free to influence. While the main realization vehicles for these systems – FPGAs – were already promising but unfortunately offered limited possibilities a decade ago [1] they have made rapid progress over the years being one of the product classes that drive the silicon technology of tomorrow. We describe an example for such a realtime digital effects system which was developed using a hardware/software co-design method. While digital realtime audio processing takes place in low latency dedicated hardware units the control and routing of audio streams is done by software running on a 32 bit NIOS II softcore processor. Implementation of the hardware units is done using a DSP centric methodology for raising the abstraction level of VHDL descriptions while still making use of standard of the shelf FPGA synthesis tools. The physical implementation of the complete system uses a rapid prototyping board tailored for communications and audio applications based on an Altera Cyclone II FPGA.
Download A Frequency Domain Adaptive Algorithm for Wave Separation We propose a frequency domain adaptive algorithm for wave separation in wind instruments. Forward and backward travelling waves are obtained from the signals acquired by two microphones placed along the tube, while the separation filter is adapted from the information given by a third microphone. Working in the frequency domain has a series of advantages, among which are the ease of design of the propagation filter and its differentiation with respect to its parameters. Although the adaptive algorithm was developed as a first step for the estimation of playing parameters in wind instruments it can also be used, without any modifications, for other applications such as in-air direction of arrival (DOA) estimation. Preliminary results on these applications will also be presented.
Download Source-Filter based Clustering for Monaural Blind Source Separation In monaural blind audio source separation scenarios, a signal mixture is usually separated into more signals than active sources. Therefore it is necessary to group the separated signals to the final source estimations. Traditionally grouping methods are supervised and thus need a learning step on appropriate training data. In contrast, we discuss unsupervised clustering of the separated channels by Mel frequency cepstrum coefficients (MFCC). We show that replacing the decorrelation step of the MFCC by the non-negative matrix factorization improves the separation quality significantly. The algorithms have been evaluated on a large test set consisting of melodies played with different instruments, vocals, speech, and noise.
Download Practical Modeling of Bucket-Brigade Device Circuits This paper discusses the sonic characteristics of the bucket-brigade device (BBD) and associated circuitry. BBDs are integrated circuits which produce a time-delayed version of an input signal. In order to reduce aliasing, distortion, and noise, BBDs are typically accompanied by low-pass filters and compander circuitry. Through circuit analysis and measurements, each component of the BBD system can be accurately modeled.
Download A Quaternion-Phase Oscillator An approach to designing dynamical systems with a three-dimensional state space is described that can be used to build a variety of non-periodic oscillators. The state space is taken to be a 3sphere, which is identified with the manifold of unit quaternions. Any such system can be described as a quaternion-valued ordinary differential equation, which is digitally realized using an approximation as a finite difference e quation. Two examples are shown. Compared to previous applications of dynamical systems used to generate audio samples, the approach described here offers a wide choice of specific flows which can neither diverge nor approach a stable limit point.
Download On the Definition of Musical Notes from Pitch Tracks for Melody Detection in Polyphonic Recordings The present study addresses the problem of defining musical notes from pitch tracks, in the context of a system for melody detection in polyphonic musical signals. This is an important issue for melody transcription, as well as melody-based music information retrieval. Previous work in the area tackled mainly the extraction of melodic pitch lines, without explicit determination of musical notes. Therefore, in this paper we propose an approach for the creation of musical notes based on a two-stage segmentation of pitch tracks. In the first step, frequency-based segmentation is carried out through the detection of frequency variations in pitch tracks. In the second stage, salience-based segmentation is performed so as to split consecutive notes with equal value, by making use of salience minima and note onsets.