Investigating Clarinet Articulation Using a Physical Model and an Artificial Blowing Machine

Abstract

A time-domain physical model is presented that is capable of simulating a variety of articulation techniques in single-reed woodwind instruments and suitable for real-time sound synthesis. Due to the nonlinear nature of the excitation mechanism, an energy-based approach is adopted for the construction of the numerical scheme in order to ensure algorithm stability. To validate the model, measurements are carried out using an artificial blowing machine. The construction of the machine, including a sensor-equipped reed and mouthpiece as well as an automated artificial tongue and lip, is described in detail. By adjusting the motion of the tongue, the blowing machine can generate audio signals corresponding to portato and staccato articulation. These signals are resynthesised following an inverse modelling approach based on the presented physical model, during which model parameters are estimated. All estimated parameters lie in a physically feasible range and may be used for sound synthesis and sound analysis applications.