[Trombone-l] Acoustics of brass instruments

[Daniel Pliskin]

>Model your sound with your ears.
>
>Much better than a machine.
>
>Seriously.

Sam and all,

I certainly understand the frustration of those who have to put up with a 
world that wants to slap numbers on everything, rather than trying to hear 
differences.  I basically wound up with an electrical engineering degree, 
while trying to figure out why sound synthesizers don’t sound real.

There are wonderful uses for models.  Let’s say I decide to see whether a 
trombone will sound darker if the slide is longer and the tuning slide is 
appropriately shorter.  I can take a hacksaw to my trombone or I can alter 
my model.  If I like what the model shows me, I’ll still take a hacksaw to 
my trombone, but just maybe my model is correct and I’ll get what I’m after. 
  If my model isn’t correct or if it isn’t accurate enough, I then go back 
and change my model, until it starts being a better predictor of what does 
what to the sound.

At that point, I’m free to check out all kinds of things.  How come a 
drilled-out mouthpiece takes on a whispery tone?  What should I first look 
for in the way of a new leadpipe?  The model directs me to where to can 
start looking.  And every time I use it, the model gets updated to make sure 
that it is as accurate as it can get it.

Even if the model is scarcely a first order approximation of what a trombone 
sounds like, it still directs me towards several solutions.  At that point, 
I know where to look for the real, audible answer.


If I thought I wanted a larger aperture in my mouthpiece but I didn’t want 
to potentially screw up my favorite mouthpiece, I’d start with a model.  I’d 
try drilling out another mouthpiece, to see whether it did, to that 
mouthpiece, what I wanted to do to my favorite mouthpiece.  That would be 
using the second mouthpiece as a model.

But I also sense a thread of anti-electronics, almost anti-technology, here. 
  Let all who are anti-technology, go back to using machine oil for slide 
oil.

DanP