<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; =
charset=iso-8859-1">
<META content="MSHTML 5.50.4134.100" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT face=Arial size=2>Hi all,</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>I thought more about how to time =
sustain while
taking my morning shower (ordinarily a very productive time for =
me). It
occurs to me that the best way to do it is electronically -- with a =
simple
circuit made just for that purpose. Many of you work at =
universities, and
you could probably find a departmental electronics technician who could =
put
together such a circuit on a breadboard in no more than an hour or =
two.
Some of you may know enough about electronics to do it yourselves. =
As for
myself, I once maintained an electonics test bench and designed/built =
these
sorts of circuits, but I've been downsizing, and most of that stuff is =
either
given away or stored in boxes. I won't be able to do this in the
forseeable future.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Anyway, here's a description of the =
circuit.
This is very elementary stuff, and any electronics tech should know how =
to put
these circuits together out of no more than a few dollars in parts (not =
counting
the breadboard, which you would borrow):</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>There would be two op-amp circuits with =
100-fold
(40 dB) differences in gain. Inputs to both would be capacitively =
coupled
from the same microphone. (An adjustable-gain preamp circuit would =
be nice
here.) Outputs from the two amps would be capacitively =
coupled to two
unity gain op amps ("voltage followers." Outputs from the voltage
followers would then be quarter-wave rectified and low-pass filtered, so =
as to
yield an approximately DC output. (i.e. diode, serial with =
resistor,
feeding to one lead on a capacitor, with the other lead grounded). =
Although this voltage wouldn't be RMS, it would be "similar" and would =
summate
energy not only only from the fundamental but also from the
harmonics. Those who want true RMS can possibly find =
specialized V^2
amplifiers, but that's a bit expensive/weird/exotic. Feed the =
outputs of
those circuits to two comparator circuits that would detect a voltage
threshold. (Assuming the 100:1 gain is accurate, the threshold =
voltage
should be the same for both the high- and low-gain circuits.) =
Finally, feed the outputs of the comparators to logic circuit that would =
output
H when the high-gain circuit's output is H and the low-gain circuit's
output is L. This could be done with an XOR chip, I =
think. Use
this output to trigger a millisecond timer (e.g. an old Hunter, probably =
available for loan from the junk pile of any dept of =
psychology).</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>How to use it:</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>(1) Hook a soundboard mic to the =
circuit's
input. To avoid constructive/destructive interference effects, =
mute two of
the three unisons of the note being tested, both in the main *and* =
duplex
sections.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>(2) Play a note loudly enough to exceed =
the
low-gain circuit's threshold (loudly enough that the timer =
won't run
immediately.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>(3) As the amplitude falls, the output =
signal of
the low-gain circuit will fall below threshold, and the timer will start =
running. The sound may "warble" a bit over the threshold, =
resulting
in the timer starting, stopping, starting, stopping, and then =
running.
That's OK, because the same thing will happen when the timer shuts
off.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>(4) As the amplitude falls further, the =
output signal of the high-gain circuit will also fall below =
threshold,
cutting off the timer -- again probably with some warbling, which is
OK.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>(5) The 40 dB decay time can then be =
read in
ms. Assuming a linear falloff (and I don't know if this is a valid =
assumption), the 6 dB falloff (i.e. amplitude half-time) would the 40 dB =
decay
time divided by 6.67. If decay is nonlinear, then just report the =
40 dB
decay and try to reference it to some initial SPL value. With =
regard to
SPL, 99% of the world has no clue that an SPL reading is utterly =
useless
unless it is referenced to a distance from the sound source. Owing =
to the
"large" sound producing area of the piano and the complex nature of its
acoustics, I would probably put the meter on a soft blanket on the floor =
beneath
the midpoint of the "strip" of soundboard immediately under the string, =
with the
mic pointed straight up and the piano lid closed. That's probably =
as
reproduceable as you're going to get.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>For those who aren't knowledgeable in =
electronics,
don't worry. This isn't neary as comlicated as it sounds. =
It's a
very easy project for some electronics technician. Look in the =
following
departments: physics, any engineering, neurobiology, psychology,
RTF, possibly even music (???).</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Peace,</FONT></DIV>
<DIV><FONT face=Arial size=2>Sarah</FONT></DIV></BODY></HTML>