Advertising and Web Hosting on Woodwind.Org!

Klarinet Archive - Posting 000036.txt from 1999/01

From: Martin Pergler <>
Subj: [kl] Re: Inharmonic partials, was RE: [kl] re:Intonation training
Date: Sat, 2 Jan 1999 11:22:02 -0500

Continuing from the last message, here are specific cases of
inharmonicity to see at home, or wherever you have a microphone and
a spectrum analyser. For those with IBM PC's and a SoundBlaster-type
sound card, I have a very old public-domain sofware spectrum
analyser named FREQ on my home page (address below, go under Music
then Oscilloscope Programs), though you may have to play with your
set-up to make it work since it runs under DOS.

Ex. 1. Nice note

Play a nice, resonant, beautiful focussed note in the chalumeau
register. Look at the spectrum. You will see strong peaks at odd-#
multiples of the base freq and weaker peaks at even-# multiples.
(The strength of the even # seems for to vary on set-up and distance
and mpiece; curious).

WHITE LIE 1: Clarinet tone has only odd partials.
TRUTH: In the chalumeau register, odd partials generally strongly
predominate. It is true that you can only "overblow" on odd
partials, this is related via the goemtry of the "pipe" but not the
same issue....continue...

Ex. 2. Nice clarion note

Play a beautiful note in the clarion register. You will see strong
peaks at both even and odd multiples of a base freq. (Play around
also with overblowing on the 12th and watch the dynamism of the

WHITE LIE 2: Clarion clarinet tone is not even mostly odd partials.
TRUTH: By opening the register key (itself a bugbear), the geometry
which gives rise to the "odd only" approximation is changed.

Ex. 3. Bugbear note

Play a throat tone, eg. Bb and don't try to make it sound "nice" or
use resonant fingerings. Eg for Bb, I notice (hopefully you too)
that the freq peak at 415 Hz has lots of noise around it (presence
of freqs not at integer multiples of 415 !!!). Also notice a
secondary peak at about 1100 Hz under 3x415@-----. This is a
(albeit weak) inharmonic partial.

WHITE LIE 3: All partials in wind instruments are harmonically
TRUTH: They generally are, and when not the additional partials had
better be weak so we still hear a nice note. I believe (though
haven't seen a reference) that the strength of these additional
inharmonic partials, the noise around the peaks, and any
imprecisions in harmonic relationships, are directly related to the
"diffuseness" and out of tune-ness of the note.

Ex. 4. Multiphonic

Play a nasty multiphonic. I tried fingering a low F, lifting 1st
finger, and blowing hard. Very "ugly" and *at least* 2 perceived
pitches. Look at the spectrum; quite messy and not harmonically
related. I haven't tried doing so, but theory predicts that the
peaks present should be approximately integer multiples of either of
the 2 *perceived* pitches. Explanation: our brain couldn't find one
harmonic pattern, but could more or less find two, ergo multiphonic.

WHITE LIE 4: Wind instruments can only produce only
almost-harmonically related partials.
TRUTH (tongue in cheek). Makers expended great effort to make sure
this is almost true for "sensible" fingering patterns. But if you do
strange things, you pass what the makers optimized for and create
nonharmonic patterns. Dastardly contemporary composers with nothing
better to do think this is fun and want us to do it deliberately!


Martin Pergler
Grad student, Mathematics
Univ. of Chicago

Unsubscribe from Klarinet, e-mail:
Subscribe to the Digest:
Additional commands:
Other problems:

     Copyright © Woodwind.Org, Inc. All Rights Reserved    Privacy Policy    Contact