"Paul Guy" wrote in message


snip mostly agreeable lead-up comments

This applies to narrow band FM (M less than 0.3). Notice that the
sign of the lower sideband is the only difference. Because IM
distortion is a second or higher order effect, they increase in
amplitude more than FM sidebands as the signal increases.

The practical way to tell the difference is to look at the amplitude
of the microphone signal versus time.

Not much help when the modulation indices are low and the measurement
environment is noisy, as it always seems to be for speakers.

If there is a variation of
amplitude (AM modulation), you have IM distortion. Then take your
signal and "clip it" (square wave), then look for sidebands. These
sidebands will be from the FM or Doppler distortion.

A bit more than that. You have to narrowband bandpass filter it before
clipping in order to avoid intermodulation. After clipping you either filter
it again, or just ignore the out-of band distortion product.

In terms of psychoacoustics I don't know if the ear can tell the
difference between AM and narrow band FM (Doppler) distortion. My
guess is that it can't, since the cochlea cannot determine phase,
unless both frequencies are very close (you'll hear "beating"), there
will be no way that the hearing system can determine the phase
relation.

Do tremolo and vibrato sound different?

What is most relevant to your discussion, is that in the presence of
a loud tone, there is considerable "masking" at frequencies close to
the main tone. Roughly speaking, for frequencies from the main tone to
about 20% lower, and from the main tone and about 50% higher, the
sidebands will not be heard if they are 20 db less than the main tone.
The relationship is complicated, in some cases its much less than the
20 db (after Wegal and Lane, 1924). What this really says, is that
NEARBY sidebands must be more than 10% before they can be heard in
any form. If this is truly the case, Doppler distortion is probably
there, but the ear cannot sense it.

A bit more than that. If nearby sidebands were that hard to hear, we'd never
hear tremolo and vibrato when the modulating frequency is low.

There are two modes of perception of modulation. If the sidebands are close
to the carrier, and they are audible, they are audible as roughness. If they
are distant from the carrier, then they are audible as separate tones,
subject to the kind of spectral masking shown above.

If you check the sensitivity of the ear to multiple tones, you will
find that masking makes many of these forms of distortion into a
non-issue.

See previous comments about low frequency modulation.

snip mostly agreeable other comments