Acoustic nulls

[radam...@gmail.com]

Suppose I play a low- frequency sine tone (100hz) through 2 spatially separated loudspeakers, and at some point in the room complete cancelation takes place. Now suppose I move the microphone away from the null point by a small fraction of the wavelength. I would expect to still get a pretty good null. However I find that if the null is near a room boundary or some other physical structure then when I measure nearby I don't get a very good null anymore. Does this make any sense?

Bob

[Scott Dorsey]

<radam...@gmail.com> wrote:
>Suppose I play a low- frequency sine tone (100hz) through 2 spatially separ=
>ated loudspeakers, and at some point in the room complete cancelation takes=
> place. Now suppose I move the microphone away from the null point by a sma=
>ll fraction of the wavelength. I would expect to still get a pretty good nu=
>ll. However I find that if the null is near a room boundary or some other p=
>hysical structure then when I measure nearby I don't get a very good null a=

>nymore. Does this make any sense?

Yes, because the sound is reflecting off that structure, so now you have
an interference pattern with three sources: the two speakers and the
reflection.

Now, it might be possible to still find a good null, but it won't be in
the same place where you'd expect, and it will be surrounded by lots of
little local minima and maxima which will make it hard to find.

I went looking for a photo of a ripple tank showing this effect but I
couldn't find one. Still, you can fill up the sink and drop a couple
balls in at the same time and watch.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[radam...@gmail.com]

Thanks.

I guess where I get confused is how you get a big pressure difference over a distance much smaller than the wavelength, even in the presence of a nearby boundary.

Bob

[robert bristow-johnson]

On 9/17/14 6:21 PM, radam...@gmail.com wrote:
>
> I guess where I get confused is how you get a big pressure difference over a distance much smaller than the wavelength, even in the presence of a nearby boundary.
>

i think, Bob, that at least for very small displacements, that this
"displacement response" (gain vs. displacement location) becomes very
similar to frequency response. keeping frequency constant and changing
displacement a little is about the same as keeping the displacement
constant and adjusting the frequency or wavelength a little.

and, perhaps, you can treat a flat wall (or floor) of hard material as a
near-perfect reflection of the sources, as if it were a mirror and the
sources were luminous rather than acoustic. so then your 2 loudspeakers
become 4. then what are the equations and path length differences that
get you a null? becomes a sorta 4th-order equation now and, perhaps,
the notch in the displacement response (that is like a notch in
frequency response for a 4-tap system, unequal spacing) is sharper than
for 2 loudspeakers.


--

r b-j r...@audioimagination.com

"Imagination is more important than knowledge."

[Scott Dorsey]

<radam...@gmail.com> wrote:
>Thanks.
>
>I guess where I get confused is how you get a big pressure difference over a distance much smaller than the wavelength, even in the presence of a nearby boundary.

With just two sources in an anechoic chamber (or out in a grassy field), the
cancellation point is much smaller than a wavelength. If you move even a
small bit, they aren't in phase. Remember the difference between a null and
a peak is only going to be a wavelength, and only a small fraction of that
difference is going to be audible.

Note that if there are no reflections and the sources are "out of phase"
meaning opposite in polarity, the positions of the null are
independent of frequency. The distance between the two sources have to
be exactly the same, but any frequency will work (although as it goes up
the point of the null gets smaller and smaller).

But, if there are reflections, the phase of the reflection at any different
point is a function of the path distance and the frequency both, so if you
have more than one frequency from the source (as you invariably do if there
is any distortion in your transducer) you will find the null positions are
different at different frequency and you may not be able to find a common
null.

I have frequently set up wedge monitors in the recording studio out
of phase, with a microphone directly between the wedges. Direct sound is
effectively nulled out, and while ambient sound isn't nulled out at all
in a dead room it might be low enough that nobody cares.

[radam...@gmail.com]

Oops, I realized I left this thread hanging , too much going on!

I guess what I was trying to understand is whether or not a pressure microphone tells the entire story. As an EE I probably mistakenly think that sound pressure is analogous to voltage, velocity is analogous to current, and acoustic impedance is analogous to impedance. In distributed RLC electrical circuit, you can have low voltage but still have high current at a particular node, and at some other "downstream" point you could have high voltage due to a change in impedance. So I assume the same could happen in acoustics. And I wonder if boundaries make a "local" change in the impedance.
Wish i had a degree in acoustics!
Bob

[Glen Walpert]

Your electrical analogy is good, and your observation that a pressure
microphone reveals nothing of impedance is correct. Boundaries do make a
local change in impedance, which, like impedance changes in an electrical
transmission line, cause reflections. In acoustics, a horn shaped
structure, with gradual change in cross section, is used for impedance
matching between driver and open air for example; without the horn
structure the driver would couple very poorly to open air.

In addition to pressure microphones there exist velocity or pressure
gradient microphones in both the very old ribbon microphone type and
modern MEMS versions, which can for example distinguish between sound
reflecting from a surface and sound originating from vibration of the
surface where a pressure microphone cannot. There was an article on this
in Sound and Vibration Magazine a while back which I cannot find now, but
I recommend the (free) magazine to anyone interested in acoustics:
http://www.sandv.com

As to the degree in acoustics, you could get the equivalent of two
semesters of grad level acoustics by reading Physical Acoustics by
Blackstock, a former acoustics prof who put about 40 years of acoustics
teaching experience into this book, which focuses on the physics of sound
more than engineering applications like Acoustical Engineering by Olson,
and unlike Olson presumes a basic understanding of partial differential
equations. As far as I know Blackstock is the only acoustics book author
who completely derives the wave equation for sound in an ideal gas from
basic principles of conservation and continuity, with all assumptions
fully justified.

Having now extolled the virtues of Blackstock's book several times now I
suppose I should admit to having worked as a 'student observer' in his
acoustics lab at U of Rochester in the summer of '68, collecting data on
several entertaining shock wave and non-linear propagation experiments.

Glen