"Isaac Wingfield" wrote in message



Paul Klipsch used to do a doppler distortion comparison between some
arbitrary 12" direct radiator and one of his big horns.

Modulation Distortion in Loudspeakers
Author(s): Klipsch, Paul W.
Publication: Preprint 562; Convention 34; April 1968

Abstract: When comparing 2 loudspeakers, one with direct radiator bass
system and the other with horn loaded bass, a subjective judgment was that
the one with the horn loaded bass is ·cleaner.· Both speakers were by the
same manufacturer. Various tests were applied and by process of elimination
it appears the difference in listening quality is due to frequency
modulation distortion. Beers and Belar analyzed this form of distortion in
1943, but since that time the effect has been almost ignored. Now, with
amplifiers and source material reaching new lows in distortion, differences
between good loudspeakers begin to appear significant. The mathematical
analysis has been reviewed, and measurements have been made using a spectrum
analyzer. These have been correlated with listening tests by preparing tapes
of oscillator tones and music with and without a low frequency source to
produce frequency modulation distortion. The spectrum analyses corroborate
the mathematical analysis and the listening tests offer a subjective
evaluation. The conclusion is that frequency modulation in loudspeakers
accounts in large measure for the masking of ·inner voices.· As Beers and
Belar put it, ·The sound is just not clean.· Reduction of diaphragm
excursions at lower frequencies reduces FM distortion. Horn loading,
properly applied, offers the greatest reduction, while simultaneously
improving bass power output capability. Tentatively it is wondered if FM
distortion in loudspeakers may be the last frontier in loudspeaker
improvement.


Even when the
difference in amplitudes was 10dB (the K-Horn being louder), the
difference in sideband amplitude was significant (the horn being a
much lower percentage). He was careful to keep the higher tone low
enough in frequency so that both tones were emitted by the woofer.

As things evolve, this makes it harder to prove that the modulation
distortion at hand is FM, mot AM

There was an obvious audible difference between the two, with the
direct radiator sounding "rougher", even when 10dB lower in amplitude.

Direct radiator drivers have improved considerably since then. For example,
the spec Xmax was introduced some decades later.

As I remember, he wanted to find some way to determine the relative AM
to FM contributions, but couldn't figure out how to do it with the
technology of the times (late '60s to early '70's, AFAIR).

The paper I cited was published in 1968. Ironically, the FFT-based
measurement technology we enjoy today was just becoming well-known at that
time.

I think that the triple tone test and modern spectrum analyzer technology
provides valuable insights into this area. I think that I've established
that when there are two upper-frequency probe tones, FM distortion will
produce sidebands with a higher amplitude with the highest frequency tone,
all other things being equal. This finding can be, and probably should be
applied to investigations relating to both Doppler distortion and jitter.