Re: Overtones

Subject: Re: Overtones
From: Jay Smalridge (
Date: Mon Jun 28 2004 - 15:03:31 EDT

that helps!

On Monday, June 28, 2004, at 11:45AM, Richard Wentk <> wrote:

>At 11:16 28/06/2004 -0700, you wrote:
>>I was wondering if you could elaborate for me:
>>When you say:
>>"Instruments with more than one sound generator (e.g. anything with multiple
>> >strings, or any ensemble) produce complex phase cancellation effects in the
>> >overtone structure which also affect the perceived sound."
>>do you mean that if more than one note is playing(on purpose or by
>>accident) then there will be phase cancellation?
>I mean anything with more than one physical noise source (e.g. the
>different strings in a guitar, piano, violin) OR any collection of
>instruments playing in an ensemble.
>For ensemble playing you get a simple chorus effect where minor mistunings
>create movement in the overtone structure.
>With physical strings and some other instruments there's also a feedback
>process via physical resonance which makes the overtone structure more
>complicated than for a single string. The sound energy isn't just created
>by the strings, it's also picked up and bounced around among them. If
>there's a soundboard, that will also contribute to the process. That
>creates some very complex and interesting variations in the overtone spectrum.
>This is why sampled pianos always sound so dead. You can hear it's a piano,
>kind of, but there's none of the movement in the sound you'd get from a
>real piano.
>If you want to hear this to best effect, either make any noise into a piano
>with the damper held down, or find a hammer dulcimer or zither - which has
>tens of strings and no damping of any sort.
>>and also when you say that pitch is important too, did you mean the note
>>being played, as in the frequency of the note?
>Yes. Every note will have a slightly different overtone structure, because
>instrument bodies tend to act like resonant filters, damping some frequency
>ranges and enhancing others.
>In physical modelling synthesis you model this by using one process to
>create an excitation signal, and then another to simulate the resonance of
>a physical body. At the risk of oversimplifying, it's a bit like putting
>the raw sound through a graphic EQ.
>It's very easy to hear this. Go find a music store and try out a selection
>of acoustic guitars. Some will be quite bass heavy, otherwise will favour
>the mid more, some will have more bite in the treble. That's all down to
>the resonant properties of the body, which in turn depend on the wood used,
>the thickness of the soundboard, the finish and also the shape and size.
>The body acts like an acoustic EQ, emphasising some of the overtones from
>the raw string sound more than others.
>Master luthiers also understand how to control the amount of string
>resonance to make guitars that really sing.

This archive was generated by hypermail 2b27 : Sat Dec 22 2007 - 01:46:02 EST