Download A Fast Mellin Transform with Applications in DAFx Many digital audio effects rely on transformations performed in the Fourier-transformed (frequency) domain. However, other transforms and domains exist and could be exploited. We propose to use the Mellin transform for a class of sound transformations. We present a fast implementation of the Mellin transform (more precisely a Fast Scale Transform), and we provide some examples on how it could be used in digital audio effects.
Download Transforming Singing Voice Expression - The Sweetness Effect We propose a real-time system which is targeted to music production in the context of vocal recordings. The aim is to transform the singer’s voice characteristics in order to achieve a sweet sounding voice. It combines three different transformations namely SubHarmonic Component Reduction (reduction of sub-harmonics, which are found in voices with vocal disorders), Vocal Tract Excitation Modification (to achieve a change in loudness) and the Intonation Modification (to achieve smoother transitions in pitch). The transformations are done in the frequency domain based on an enhanced phase-locked vocoder. The Expression Adaptive Control estimates the amount of present vocal disorder in the singer’s voice. This estimate automatically controls the amount of SubHarmonic Component reduction to assure a natural sounding transformation.
Download Enhanced Time-Stretching Using Order-2 Sinusoidal Modeling In this article, we introduce a 2-level sinusoidal model and demon strate its aptitude for a challenging digital audio effect: time-stretching without audible artifacts. More precisely, sinusoidal modeling is used at the two levels of the new sound model. We consider the frequency and amplitude parameters of the partials of the classic sinusoidal model as (control) signals, that we propose to model again using a sinusoidal model. This way, higher-level musical structures such as the vibrato and tremolo in the original sound are captured in the “partials of partials” of this order-2 sinusoidal model. We propose then a new time-stretching method, based on this new hierarchical model, which preserves not only the pitch of the original sound, but also its natural vibrato and tremolo.
Download An Efficient Phasiness Reduction Technique for Moderate Audio Time-Scale Modification Phase vocoder approaches to time-scale modification of audio introduce a reverberant/phasy artifact into the time-scaled output due to a loss in phase coherence between short-time Fourier transform (STFT) bins. Recent improvements to the phase vocoder have reduced the presence of this artifact, however, it remains a problem. A method of time-scaling is presented that results in a further reduction in phasiness, for moderate time-scale factors, by taking advantage of some flexibility that exists in the choice of phase required so as to maintain horizontal phase coherence between related STFT bins. Furthermore, the approach leads to a reduction in computational load within the range of time-scaling factors for which phasiness is reduced.
Download A Piano Model Including Longitudinal String Vibrations In this paper a mixed-paradigm piano model is presented. The major development is the ability of modeling longitudinal string vibrations. Longitudinal string motion is the reason for the metallic sound of low piano notes, therefore its modeling greatly improves the perceptual quality of synthesized piano sound. In this novel approach the transversal displacement of the string is computed by a finite-difference string model and the longitudinal motion is calculated by a set of second-order resonators, which are nonlinearly excited by the transversal vibration. The soundboard is modeled by a multi-rate filter based on measurements of real pianos. The piano model is able to produce high-quality piano sounds in real-time with about 5–10 note polyphony on an average personal computer.
Download The Feathered Clarinet Reed In this research, a method previously In this research, a method previouslyapplied appliedtotoimprove improve a digital simulation of the avian syrinx is adapted to the geometry of the clarinet reed. The clarinet model is studied with particular attention to the case when the reed beats again the lay of the mouthpiece, closing off air flow to the bore once each period. In place of the standard reed table which gives steady-state volume flow as a function of constant pressure difference across the reed, a more realistic dynamic volume flow model is proposed. The differential equation governing volume flow dynamics is seen to have a singularity at the point of reed closure, where both the volume flow and reed channel area become zero. The feathered clarinet reed refers to the method, first used in the syrinx, to smooth or feather the volume flow cutoff in a closing valve. The feathered valve eliminates the singularity and reduces artifacts in the simulated clarinet output.
Download Digital Sound Synthesis of Brass Instruments by Physical Modeling The Functional Transformation Method (FTM) is an established method for sound synthesis by physical modeling, which has proven its feasibility so far by the application to strings and membranes. Based on integral transformations, it provides a discrete solution for continuous physical problems given in form of initialboundary-value problems. This paper extends the range of applications of the FTM to brass instruments. A full continuous physical model of the instrument, consisting of an air column, a mouthpiece and the player’s lips is introduced and solved in the discrete domain. It is shown, that the FTM is a suitable method also for sound synthesis of brass instruments.
Download Recent Advances in Physical Modeling with K- and W-Techniques Physical (or physics-based) modeling of musical instruments is one of the main research fields in computer music. A basic question, with increasing research interest recently, is to understand how different discrete-time modeling paradigms are interrelated and can be combined, whereby wave modeling with wave quantities (W-methods) and Kirchhoff quantities (K-methods) can be understood in the same theoretical framework. This paper presents recent results from the HUT Sound Source Modeling group, both in the form of theoretical discussions and by examples of Kvs. W-modeling in sound synthesis of musical instruments.
Download Computation of Nonlinear Filter Networks Containing Delay-Free Paths A method for solving filter networks made of linear and nonlinear filters is presented. The method is valid independently of the presence of delay-free paths in the network, provided that the nonlinearities in the system respect certain (weak) hypotheses verified by a wide class of real components: in particular, that the contribution to the output due to the memory of the nonlinear blocks can be extracted from each nonlinearity separately. The method translates into a general procedure for computing the filter network, hence it can serve as a testbed for offline testing of complex audio systems and as a starting point toward further code optimizations aimed at achieving real time.
Download Modal-Type Synthesis Techniques for Nonlinear Strings with an Energy Conservation Property There has recently been increased interest in the modelling of string vibration under large amplitude conditions, for sound synthesis purposes. A simple nonlinear model is given by the KirchhoffCarrier equation, which can be thought of as a generalization of the wave equation to the case for which the string tension is “modulated” by variations in the length of the string under deformation. Finite difference schemes are one means of approach for the simulation of nonlinear PDE systems; in this case, however, as the nonlinearity is spatially invariant, the solution may be broken down into sinusoidal components, much as in the linear case. More importantly, if time discretization is carried out in a particular way, it is possible to obtain a conserved energy in the numerical scheme, leading to a useful numerical stability guarantee, which can be difficult to obtain for strongly nonlinear systems. Numerical results are presented.