[ Contributed by a music fan.  I wish that this respected author
 [ had also addressed Signal-to-Noise Ratio...!   -- Robbie

 - - - - - -

This is from the Letters to the Editor section of the July/August issue
of that respected scientific journal, The Annals of Improbable Research.

"I have made a careful study of bandwidths and their influence on
communications. Using a fixed number of trombones (N=76) in each study,
I have determined that the width of the band is actually a function of
the width of the avenue. This correlation is maintained whether or not
there are 110 cornets right behind.

"There is, on the other hand, an inverse correlation between bandwidth
and bandlength for different values of N, when avenue remains constant.
In avenues of communication, there are no statistically measurable
differences in band-reception from the standpoint of a streetside
audience, except when band quality is so bad it is poorly received.

"Researchers in data communications are cautioned that bandlength can
be more critical than bandwidth. Very long data strings, which I call
"parades," can invoke inattentiveness in the receivers.

"(Signed) J. P. Sousa, Philadelphia, PA"