[sdiy] Unstability of oscillators and psychoacoustic qualities

[Magnus Danielson]

From: Glen <mclilith at ezwv.com>
Subject: Re: [sdiy] Unstability of oscillators and psychoacoustic qualities
Date: Sun, 22 Sep 2002 00:18:38 -0400

> At 10:27 PM 9/21/02 , Magnus Danielson wrote:
> 
> >I may have an incorrect personal assumption that "warm" oscillators have some
> >sort of phase noise in them. This causes them to wander about the note
> >frequency.
> 
> I think your assumption is probably correct. Rodgers electronic church
> organs (when they were still designed with analog oscillators) were
> designed with circuitry that purposely introduced frequency "jitter" into
> the oscillators. This was an effort to emulate the "warmth" and
> "naturalness" of the sound of air turbulence (and a resulting frequency
> jitter) in a traditional wind-blown organ pipe. 
> 
> They also purposely used a separate oscillator per musical pitch, and they
> avoided divider circuitry like the plague. Larger organ models had multiple
> sets of oscillator banks, including banks of "celeste" oscillators which
> are purposely tuned slightly sharp for a special effect. Of course, the
> celeste oscillators were only used for special "celeste voices", which were
> sort of the pipe-organ equivalent of a twelve string guitar. In other
> words, for each note you played on the instrument, your ear heard two
> slightly different pitches simultaneously. Such voices are considered
> especially "warm" or "lush" in nature by most organ people.

Right. There's loads of similar indications.

If you have a piano, hit one string, and then hit two and then all three for
the note. Quite a different sound. The sound of the piano builds on the
multiple string setup.

> I think that you'll find a "warm" oscillator has some form of frequency
> "jitter" inherent in its nature. What the best form or style of "jitter"
> might be, is something that I couldn't say for sure. You might end up
> making a spectral plot of the frequencies and amplitudes of the various
> deviations from perfect pitch exhibited by a given oscillator, and be able
> to display its inherent warmth or coldness as a 2D image or 3D shape. Such
> "electronic signatures" might be of some use in characterizing elusive
> audible properties of oscillator designs, both old and new.

Indeed. There are specialized frequency counters that can do this and you can
get cute plots of all possible kinds out of these. Postprocessing them can
bring out many cool aspects.

> At least, it's something that will keep you busy for awhile.  :)

Well, I did say I did have some spare time on my hands... ;O)

Cheers,
Magnus