Re: Modelling instruments

[Manuel Carro Li~nares <boris@clip.dia.fi.upm.es>]

Jackie Long writes:

>> I am just putting together a proposal for a research project which aims to
>> set up computer models of musical instruments in such a way as to
>> represent their physical parameters.  It is intended that these parameters
>> can be changed by the user in such a way that different physical aspects
>> of a musical instrument can be represented.  As a result, the significance
>> of various physical parameters and their contribution to the sound
>> produced by the instrument would be determined analytically. 
>> 
>> [...] At some stage I should also like to incorporate the effect of
>> various type of wood, brass finishes, cleaning, staining, oiling,
>> preserving, etc.
>> 
>> I should be very pleased to hear from anyone who might be interested in
>> the principles of such a project and any other suggestions or comments you
>> might have from your own point of interest. 

    Well, though I am not an expert on the matter, there has been some
research carried out in the last few years. I cannot provide pointers
to technical papers, but the so-called "morphic synthesis" (pardon if
I this is the wrong spelling) has proven to be successful in
simulating real instruments. Moreover, there are a couple of
synthesizers in the market which use this very advanced technique.

    This technique has been applied mostly to wind instruments, where
sampling and other more classic techniques cannot give the highest
quality. A wind instrument with closed fingering is virtually
different for each note. The idea of "morphic synthesis" applied to
wind instruments is precisely to model the physical phenomena (wind
pressure, etc) in each point inside the instrument. You can change
physical parameters, add holes with different size and shape, etc., to
simulate real and non-real instruments.  This simulation is done in
real time, using algorithms that approximate the dynamic behavior of
the air flow. I believe that most of these algorithms have been
developed at Stanford.

    Checking rec.music.synth would very likely give some more
information.

    Hope this helps,

        Manuel Carro


                  ||  Manuel Carro -- CLIP Group  ||
                  ||   e-mail: mcarro@fi.upm.es   ||
                  ||     Phone: +34 1 336-7448    ||
                  ||       FAX: +34 1 352-4819    ||