Fwd: How strings age (OT)

Subject: Fwd: How strings age (OT)
From: Kevin Austin (kevin.austin@videotron.ca)
Date: Mon Jan 24 2005 - 04:07:17 EST

For those interested in problems related to physical modeling
(notably of strings), there is the following. It partially answers a
question about why 'real' strings sound different from some
synthesized strings (the boundary conditions at the bridge are
different for the vertical and horizontal vibrational modes,
producing variable beating of the harmonics).


>Date: Mon, 24 Jan 2005 00:00:07 -0500
>From: Automatic digest processor <LISTSERV@LISTS.MCGILL.CA>
>Subject: AUDITORY Digest - 22 Jan 2005 to 23 Jan 2005 (#2005-15)
>Date: Sun, 23 Jan 2005 05:47:56 -0600
>From: Jont Allen <Jontalle@UIUC.EDU>
>Subject: Re: AUDITORY Digest - 21 Jan 2005 to 22 Jan 2005 (#2005-14)
>In the article below (you'll have difficult time getting it I
>suspect) I measured B for new and "old" guitar strings. The
>hypothesis was that as the string ages, B would change.
>What I found was that the damping of the high partials changed and
>that was the real source of the aging effect.
>A second conclusion was that the size of B is perceptually very
>important. [A tiny value of B (numbers are in the paper, and I
>suspect in many other places as well) leads to a huge perceptual
>A third conclusion was that there are two sets of modes, and the
>beating of these harmonics from the two sets of modes, was as
>important as the inharmonisity (sp?) effect (the effect of B). These
>mode sets make the measurement of B a nontrivial exercise in a real
>string sound. The "vertical" and "horizontal" modes are most likely
>(this was the assumption of the model I developed in the paper, at
>least) due to the difference in the boundary condition at the
>bridge. Thus in the paper, the impedance was represented as a
>matrix, with the two modes separated by the bridge matrix boundary
>A forth conclusion was that there was another effect I didd not nail
>down. There is a fast wave that arrives much before the normal wave
>on the string, which is easily seen in the impulse response. I have
>no idea how important this fast wave is perceptually, but I cant
>rule it out as important, other than to say the following. With
>Mindy Garber (my student many years later) we did time domain
>simulations of piano strings, and we used three strings in parallel,
>with vertical and horzizontal waves (for a total of 6 modes at each
>partial, all beating against each other), at we found excellent
>agreement with real sounds. They were really natural sounds. As far
>as I know, Mindy never published this result. She went off to
>Stanford and started working for Charles Steel, and I never (well
>almost never) saw here again.
>author = {Allen, J. B.},
> title = {On the aging of steel guitar strings},
> journal = {Catgut Acoustical Society Newsletter},
> year = {1976}
>If I get more than 10 requests for this, I'll scan a copy and make
>it available on my web site. If you wish to make one of these
>requests, send it to the following email address. Please dont spam
>my regular email address:
>Jont Allen (jba "at" auditorymodels.org)

This archive was generated by hypermail 2b27 : Sat Dec 22 2007 - 01:46:06 EST