Quantifying What You Hear...

[David Love]

I read lots of naysaying in the responses to your inquiry.  I hope you're
not discouraged.  I think there are plenty of things to quantify.  Assigning
that data to a qualitative description will be no less important a task but
can be treated separately after, or while in the process of collecting data.


Establishing a baseline for comparison with the soundboards themselves will
not be easy, especially with compression style soundboards whose stiffness
in various parts of the board can vary considerably.  Measuring deflection
characteristics under different loads seems like a logical place to start.
Panel thickness, grain angle and density, rib radius and construction will
all influence the data you are collecting.  Just exactly what it is that you
are measuring may be difficult to determine.  That very big problem aside,
it seems like you should measure at least the following:

1.  For "brightness" perhaps a spectrum analysis of balance and strength of
partials in various parts of the piano.  The analysis should be done with
different types of hammers of different weights and densities (maybe choose
three types: soft, medium, hard).   

2.  For measuring "power" you might consider an analysis of the amplitudinal
relationship (as well as the absolute numbers) between the attack envelope
(the initial spike on the oscilloscope) and the ensuing sustain envelope
that follows.  

3.  Sustain can be measured in terms of both the rate of decay (shape of the
curve itself) as well as the duration.  There are different types of
sustain.  In some cases, following the initial attack, the level drops off
quickly but lasts a fairly long time.  In other cases, the sustain drops at
a more constant rate, etc...  First establish the different types of sustain
quantitatively and then derive a qualitative description.  

Overall, I would first determine what you can measure and after you have
measured it do a qualitative analysis be consensus to figure out how what
you are measuring translates to what you are hearing.  The easiest way to
start would be with a single piano and various types of hammers of different
consistency, weight, etc.  That way, you are only changing one variable.
Experimenting with soundboards complicates the control of variables
considerably but it certainly can be done one piano at a time.  

Undoubtedly there's more but that's what comes to mind immediately.  

A terrific project! 
 
David Love
davidlovepianos at comcast.net 
www.davidlovepianos.com

-----Original Message-----
From: pianotech-bounces at ptg.org [mailto:pianotech-bounces at ptg.org] On Behalf
Of Matt Borland
Sent: Friday, December 29, 2006 4:37 PM
To: pianotech at ptg.org
Subject: Quantifying What You Hear...

Hello,

I'm new to this list, but I was wondering if I could get some 
help/opinions/ideas about the way sound in musical instruments is 
described. Currently I'm doing a masters working with Stephen Birkett 
at the University of Waterloo and one of the big problems we have found 
is the inability to discuss musical acoustics in any quantifiable way 
that has meaning to both musicians and people using a scientific 
approach (not to say that people can't fall into both of those groups 
at the same time). My work is going to involve piano soundboards, but 
before I start on that I want to think about and define some ways to 
describe the sound/tone of the soundboards and pianos I will be 
measuring for vibrational and acoustic properties.  I think we've all 
used words like bright, muddy, crisp, sharp, round, dark, etc. to 
describe the sound of an instrument, but these are highly subjective 
words that are difficult to draw any concrete conclusions from. So the 
question is, are there any other parameters you feel would be useful to 
quantify? Maybe there is a way to measure how "bright" something 
sounds...If you have a concept and some sort of definition to go along 
with it I'd love to hear from you.

Some obvious ones (if these are poorly defined, feel free to redefine 
them) are:

decay time - the time it takes for sound level to decay by a defined 
amount (ie 60dB, or whatever, I'm thinking of the RT60 definition for 
reverberation time from acoustics) linked to sustain
impedance - a measure of opposition to motion of a structure subjected 
to a force
bloom - change in tone over time
response time - is the time a system or functional unit takes to react 
to a given input

Basically I would like to make the link between the frequency and modal 
analysis techniques available with the language that musical 
instruments are discussed in terms of. I really doubt anyone has talked 
about how nice the mode shapes of their pianos are, but maybe if things 
like this are connected to the way they sound by some common definition 
or understanding then they could be used to measure the properties of 
an instrument. I would also like to make the link between these 
techniques and the way we perceive sound and pitch, something that I 
feel has been forgotten in a lot of research work.

Any help or ideas would be greatly appreciated,

Matt Borland