文档介绍：COMPUTER MODELLING OF SOUND
FOR
TRANSFORMATION AND SYNTHESIS
OF
MUSICAL SIGNALS
Paul Masri
December 1996
A thesis submitted to the University of Bristol in accordance with the requirements
for the degree of Doctor of Philosophy in the Faculty of Engineering, Department
of Electrical and Electronic Engineering.
©1997 Paul Masri
PDF downloaded from
ABSTRACT
The purpose of this thesis is to develop a sound model that can be used as a creative tool by
professional musicians. Post-production editing suites are used piling and arranging
music tracks, and for creating soundtracks and voice-overs for the radio, television and film
industries. A sound model would bring a new dimension of flexibility to these systems,
allowing the user to stretch and mould sounds as they please.
Sound models already exist but they are limited both in their usability and in their scope for
representation. All aspects of the model in this thesis use designer-preset global variables
which are transparent to the user. Within this restriction and preserving manipulation
flexibility, the aim of the thesis is to improve the range of sounds that can be modelled and the
accuracy of modelling. These are dependent on the choice of model elements and the accuracy
of the analysis-resynthesis system (which translates between the playable time domain
waveform and the controllable model feature domain, making the model usable).
The basis of the model of this thesis is a deterministic-stochastic classification; the partials of
the harmonic structure of pitched sounds are individually represented in the deterministic
aspect, whilst the stochastic aspect models the remainder as broadband noise. Three studies
were carried out to improve aspects of the analysis-resynthesis system. These focus on:
• the time-frequency representation, by which the analyser ‘sees’ detail in the sound;
• frame linking, which converts the instantaneous partial estimates into co