Attack transients in clarinet models with different complexity - a comparative view

Abstract : Recent works on simplified clarinet models using results from dynamic bifurcation theory have allowed to predict the evolution of the amplitude of sound (the amplitude envelope) for a gradual increase of the blowing pressure. The unrealistic model that predicted the amplitudes to attain very small values, far below the precision of a computer, was later corrected by the addition of stochastic noise to the model. The two models are useful in explaining and understanding why the oscillations appear with a delay relative to the threshold of oscillation that is predicted by purely steady-state models. Both the model of the instrument and that of the noise are extremely simplistic, raising the question of its applicability to real instruments. These models can however be made gradually more complex by introducing more realistic details in the reed or in the resonator, and applying parameter profiles with more complex shapes or noise amplitudes. This presentation shows the differences encountered in the time-evolution of the acoustic wave simulated using two models of different complexity, one with an instantaneous reflection function, another with dispersion. The article explores to which extent can the dynamic predictive model be used to describe the time evolution of more realistic models, and hopefully that of the real instrument.
Document type :
Conference papers
Complete list of metadatas

Cited literature [10 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01320503
Contributor : Andre Almeida <>
Submitted on : Thursday, June 20, 2019 - 7:15:26 AM
Last modification on : Saturday, June 22, 2019 - 1:40:54 AM

File

000068.pdf
Publisher files allowed on an open archive

Identifiers

  • HAL Id : hal-01320503, version 1

Citation

André Almeida, Baptiste Bergeot, C Vergez. Attack transients in clarinet models with different complexity - a comparative view. International Symposium on Music Acoustics, Jul 2014, Le Mans, France. ⟨hal-01320503⟩

Share

Metrics

Record views

60

Files downloads

6