Wednesday, 21 October, 2009
KTH, Auditorium F2, Lindstedtsvägen 26
The symposium is free of charge and open to everyone interested. The auditorium is large, but to ensure that seats will be available, please confirm your attendance, by e-mailing to aa-60@csc.kth.se before 16 October.
The talks indicated by [ T ] will be mediated live from across the globe, using teleconferencing systems.
Doors open at 08:45.
Morning Session – Chair: Sten Ternström
09:15 Welcome and Introduction
Sten Ternström, head of the KTH Music Acoustics Group
09:30
Johan Sundberg, Prof. Em. of Music Acoustics, KTH
10:00
Acoustical effects of the player’s vocal tract on wind
instrument performance
Jer Ming Chen, John Smith and [ T ] Joe Wolfe
School of Physics, The University of New South Wales, Sydney, Australia
The reed or lip valve generator of many wind instruments is loaded by the impedance of the bore downstream and by that of the player’s tract upstream. This talk presents acoustical impedance spectra measured inside the mouth, during performance, and uses them to explain how vocal tract resonances can select or affect the pitch and modify the spectral envelope of the tones produced.
10:30 Coffee break
11:00
[ T ] Jim Woodhouse, Cambridge University, United Kingdom
Computer models of violin bowing have developed over many years, but there is still no model which is entirely reliable for predicting transient details following a given bow gesture. The main source of uncertainty lies in the friction behaviour of rosin. Recent work in Cambridge and in Pittsburgh will be summarised, to show the current best guess at a “correct” model, and some measurements which show that this is still not correct!
11:30
Nonlinear delay-free loop filter networks: the case of the voltage-controlled filter
Federico Fontana, University of Verona, Italy
Several nonlinear analog systems containing feedback circuits can be conveniently transformed in discrete time filter networks and computed, even in presence of delay-free loops. An application of this transformation to the analysis, computation, and real time simulation of the voltage-controlled filter, a module that was on board of some analog music synthesisers of the ‘60s, will be presented.
12:00
Antoine Chaigne, ENSTA, Paris, France
Modeling the initial transient of piano tone is essential for obtaining realistic simulations. In this talk, a model is presented that reproduces the nonlinear coupling between transverse and longitudinal motion of the string. First results are shown of time-domain simulations of a nonlinear string coupled to a finite impedance. In addition, recent experiments on string transients performed on a strung upright piano soundboard illustrating the nonlinear effects will be described and discussed.
Afternoon session – Chair: Anders Friberg
14:00
Knut Guettler, Oslo University College of Music, Norway
Multiphonics can occasionally be heard from woodwind instruments, both in jazz and contemporary music. However, a similar effect is obtainable in bowed string instruments through a careful combination of bow position and position of a lightly touching finger―on a single string. Different from in normal “harmonics” (flageolet tones), in multiphonics the fundamental of the open string is not fully suppressed, and there are normally more than one high-pitched frequency ringing on top of it. This presentation will give examples and analyses of a number of tone combinations achieved with this novel technique.
14:30
Performance modeling for instrumental
sound synthesis:
bowing control for violin sound generation
[ T ] Esteban Maestre and Alfonso Pérez, Music Technology Group, Universitat Pompeu Fabra, Barcelona, Spain
A framework for modeling violin sound and performance is presented. First, we introduce sensing techniques for accurate acquisition of relevant timbre-related bowing control parameters. Then, we describe an analysis/synthesis scheme for modeling bowing parameter contours, and how it is used for generating bowing controls from a symbolic score. Next, we show how these bowing controls can be used to predict bridge-pickup signal’s spectral envelopes by means of machine learning techniques. Realistic violin sound can be obtained by integrating these ideas into physical models, spectral models and sample-based synthesis frameworks. We conclude the talk by providing some sound examples and pointing out future directions.
15:00
15:30 Coffee break
16:00
[ T ] Julius O. Smith III, CCRMA, Stanford University, USA.
This talk will provide an overview of recent work at the Center for Computer Research in Music and Acoustics pertaining to virtual musical instruments, including work on virtual acoustic guitar elements and haptic (force feedback) control of virtual string and percussion instruments.
16:30
Vesa Välimäki, Dept of Signal Processing and Acoustics, TKK Helsinki University of Technology
This presentation summarizes our recent studies on the sustain pedal effect in the piano. A digital model for the sustain pedal effect is presented, and an analysis of the partial-pedaling effect is discussed.
17:00 End