[naturerecordists] writes:
> David, now is early in the morning and I would be (we would?) very intere=
sted in the explanation of the theory of the collar! :-)
You asked, so here's my take on it:
The collar is a short acoustic tube in front of the mic. As an open tube it=
has an "end effect" which effectively lengthens the tube by diam X 0.4.
Consider first if the was a long tube which would have a positive reflectio=
n
at the mic end and a negative reflection at the open end. This would
resonate at half a wavelength. With a very short tube, the resonance would=
be dampened but the effect would still be there. For a 10mm effective tube,=
the peak resonance would be about 17KHz.
The slope of the curve up to this peak would give a rising HF response.
The next question is why a collar makes a mic more directional. This effect=
is used in gunmics to get a very directiona response from a multi-input
perforated tube, but a single opening short tube would still have a
directional effect as direct sounds would reach the capsure almost
unaltered, but oblique sounds would have to diffuse into the opening.
Oblique sounds would "see" a reduced opening area.
The plain mic is designed to have a nominally flat response from any angle=
but still loses some extreme HF from the rear. With a collar, rear sounds
have to travel past the short tube then diffuse backwards towards the
diaphragm. Again this is a partial effect increasing at high frequencies,
but audible.
Sounds diffusing from the side or back into the short tube, or any tube for=
that matter, have to pass a change of acoustic impedance at the open end. A=
s
they do so, there are different path lengths from the close rim to the far=
rim, and these can interfere at high frequencies but I'm dubious how much
additional effect that would have, but on testing there is a "side effect"=
on all mics which is frequency dependant.
> Btw, I'm using this Felmicamp configuration taking advantage of the quali=
ty of these small omnidirectional EM 172 capsules, as if they were "pseudo =
cardioids"!
Without a rear null they would be only partially cardioid.
That's my take on the "why", but in real life recording, we have to rely on=
experiment.
David Brinicombe
|