More on acoustics

From: Riley Lee (riley@rileylee.net)
Date: Tue Sep 16 2003 - 04:56:32 PDT


Greetings.

I sent the whole discussion on acoustics to Joe Wolfe. Here is his
reply (see below).

As for me, it's probably a bit like anything really, say horses. Horse
people like to talk about horse food, horse stables, horses themselves,
horse shoes, saddles, horse riding techniques, breeds, genealogies, and
so forth. In general terms, in technical terms, in mystical and in
mythical terms. Whatever. So long as it has to do with horses, it's
interesting.
More than interesting.
It's fascinating.

Best regards, Riley

Sound of Bamboo
PO Box 939
Manly 1655 NSW Australia
Tel: +61 (0)2 9976 6904
Fax: +61 (0)2 9976 6905
mobile: +61 (0)414 626 453
www.rileylee.net

Dear Shakuhachi list,

the Acoustics Lab at the University of New South Wales has done some
work
on impedance spectra of shakuhachi and other wind instruments. Our site
at
http://www.phys.unsw.edu.au/music has a section on shakuhachi
acoustics.
We are working on a few problems related to your discussion.

As scientists working in this area, obviously we think that science can
make a contribution. We think of the instrument as a tool for making
music,
and that in some cases we hope to improve that tool. This attitude may
not
seem sufficiently respectful to shakuhachi tradition, so I apologise:
it is
an attitude that we have developed from working on occidental
instruments.
Whatever you may think of such an attitude, here are a few relevant
observations.

Making the shakuhachi with the 'optimum' impedance spectrum Z(f) is, as
you
have mentioned, an awkward problem. Not so much technically: the
shakuhachi
is one of the few wind instruments in which makers can not only take
material out of the bore, but also put it in. And the physics needed to
calculate the bore shape is mainly messy, not very difficult. Having a
small number of tone-holes simplifies the calculation.

For instance, one property that one might think desirable is to have the
first two minima in Z(f) in the ratio 2:1. (Perhaps higher resonances
might
have harmonic ratios as well, but this is in general very difficult.) In
principle, one would expect this to give not-only well tuned octaves
but a
more stable and stronger lower octave, because the first two (and
sometimes
higher) resonances of the instrument 'cooperate' with the air jet to
produce notes in the lower register.

The complication comes because the Z(f) in question should include the
effect of the player's chin and lips, and also related properties of the
jet. These are quite variable --indeed that is one of the chief features
and attractions of the instrument. They are also frequency dependent,
in a
somewhat different way from the instrument. So one needs to know what
the
Z(f) of the player's embouchure is. (Awkward, but not impossible: we did
some measurements of the Z(f) of the embouchure of players (including
Riley
Lee) for several notes.) The trouble comes in deciding which notes, and
for
how much 'meri', if I may put it that way, one would want to optimise
such
a property.

For two reasons, we decided to tackle this problem first on the
transverse
(occidental) flute. One is that it is somewhat easier to make decisions
about what ought to be optimised. Another is that a local maker of
occidental flutes (Terry McGee in Canberra) was keen to collaborate with
us, and the government is keen on universities collaborating with local
makers. This is now the doctoral project of a student in the lab (Paul
Dickens).

However, we are still quite puzzled about the shakuhachi and want to
return
to that problem in the future (unless the posters on this list have
solved
the problems first).

Anyone who is interested in how such calculations may be done could see
Botros et al (2002), which downloads from
http://www.phys.unsw.edu.au/music/publications.html
This reports doing the calculations in the 'forwards' direction (bore to
Z(f), rather than the inverse), but the inversion problem (making the
backwards direction calculation) is essentially making a 'book' of the
forwards calculation, then looking it up, finding useful areas of
parameter
space, making a new book, looking it up.... fortunately computers don't
mind doing such mindless work.

Good luck and good music
Joe Wolfe
Acoustics Lab, UNSW, Sydney http://www.phys.unsw.edu.au/music

List subscription information is at:
 http://communication.ucsd.edu/shaku/listsub.html



This archive was generated by hypermail 2b29 : Tue Jan 06 2004 - 14:09:34 PST