Re: 2 source interference w/ sound outdoors

[John Denker <jsd@MONMOUTH.COM>]

At 09:26 AM 4/10/00 -0400, GARY HEMMINGER wrote:
> We're planning to try again this year the demo with 2 speakers & a
> sine wave generator in a field.  Last year it didn't work very well,

Can you be more specific about the setup, what you observed, and why you
didn't like it?

> there was a building about 50 yards downrange from the speakers,
> and relefections might have been too great, but we also found that
> whereas a pair of cheap midrange speakers gave a weak
> interference pattern, a pair of expensive two-way speakers (which
> were louder and thus should have been better) didn't work at all.  Is
> it possible that crossover circuits in two or three way speakers
> would somehow muddy the waters?

I doubt the crossovers have much to do with it.  By Floquet's theorem, for
a linear system, if you put sine waves in you will get sine waves out, at
the same frequency.

The possibility of nonlinearity depends not directly on the price of the
speaker, but on how hard you are driving it relative to its power-handling
ability.  It should be straightforward to establish an upper bound on any
possible nonlinearity (harmonic distortion).

Of course the presence of buildings etc. downrange will muddy the
waters.  But it should be straightforward to establish an upper bound on
the magnitude of this effect (experimentally and/or theoretically).

===

You also need to take into account 3D effects.  This is a case where a
physically large multi-cone speaker could be more troublesome than a small
single-cone speaker.  Large multi-cone speakers (especially the "tower"
types) go to some trouble to "aim" the sound roughly parallel to the
ground.  In such a case, small uncontrolled variations in aiming the
speakers and/or small uncontrolled variations in mike height will muddy the
waters, especially at midrange and higher frequencies.

Even a single-cone speaker is likely to be ported, meaning it looks more
like a dipole than a point source at some frequencies.  This will lead to
interference patterns quite different from what you were expecting.

Suggestions:

1) Try using a horn-type speaker (which needs no port) rather than a
cone-type speaker.
2) If you must use a cone-type speaker,
  a) use one with a single cone.
  b) if possible, use an unported "infinite baffle" design.  You may have
to build this.
3) Use the smallest-sized speakers and the lowest frequencies you possibly can.