naturerecordists
[Top] [All Lists]

Re: Noise Cancelling (was Pollution)

Subject: Re: Noise Cancelling (was Pollution)
From: "richpeet" <>
Date: Tue, 12 Feb 2002 00:15:10 -0000
I tried the hardest test I could come up with.
Here is sparrows recorded at ~150 feet with a highway directly in
back of the sparrows, an interstate 1/2 mile behind, in a 10 mph wind.
two identical mics, me 62. one in a 32" parabolic, one directly over
the first by 24" outside the parabolic. The channel outside the
parabolic was amplified 50%, inverted, and combined to mono. First
half of the recording is the parabolic only second is the combined
result of the two channels. What you can not see is which birds were
directly in the parabolic path and which were not.

a large 1.5 meg for the wave file
a small 39K spectragraph for the jpg file

see
http://people.mn.mediaone.net/richpeet/nc.wav
and
http://people.mn.mediaone.net/richpeet/nc.jpg

files will disappear by the end of the month when the isp goes down.
they are then avail upon request once i change my e-mail

What ya think?




--- In  Lang Elliott <> wrote:
> I just did a second test in which I tried to greatly reduce or
eliminate
> "white noise" coming from behind the parabola (from a point source
speaker).
> I immediately realized that I get a much better result if the
second,
> polarity reversed mike is placed somewhere on the "inside" of the
reflector,
> so that it is receiving the same sound level (from the white noise)
as the
> main mike. I got good results by scrunching the head the the second
mike
> down to the bottom of the parabola, along the axis, well below the
focal
> point (=3D at the bottom of the dish).
>
> I also tested this arrangement using the on-axis playback of my
bird song
> and found the same 7db reduction in signal level when using the
noise
> cancelling mike.
>
> Encouraging! If this kind of setup allows for, let's conjecture, a
15-20 db
> reduction in background hiss due to wind or stream noise as well as
a nearly
> complete elimination of distant rumble, while only reducing the on-
axis bird
> sound (or other animal sound) around 7 db, then this has definite
> application, at least for parabolic-type recording.
>
> Lang
> > Well, my first test is done and the results are encouraging.
> >
> > I did the test in a room by playing a high pitched bird song over
a single
> > speaker and aiming my parabola directly at the sound source,
focusing it
> > carefully before taking measurements. The room is a conventional
one with
> > hard walls and hence some reverberation, which might interfere
slightly with
> > the results.
> >
> > I used the meter on my DAT recorder and adjusted the input level
peak of the
> > repeated bird song to -2 db when using the Telinga parabola with
a single
> > MKH 20 microphone mounted at the focal point of the parabola.
> >
> > As I mentioned earlier, I mounted a second microphone on the edge
of my
> > parabola, pointing it upward and equipping it with a short cable
that I
> > wired to reverse the polarity.
> >
> > When I added the second mike to the recorder input using a "Y"
cable
> > connector, I immediately noticed two things:
> >
> > 1) background rumble from traffic outside completely disappeared
> >
> > 2) the bird song itself was reduced in intensity around 7 db
> >
> > This is encouraging. The 7 db reduction in the bird song seems
minimal in
> > comparison to the extreme reduction of rumble. In non-reverberant
outdoor
> > situations, this reduction might be even less, perhaps only 5 db
or so.
> >
> > As Marty just mentions, this is to be expected for sounds at the
low end,
> > which you could filter away later in the studio. However, will
this work
> > with more broadband sounds such as hiss from wind or stream noise.
> >
> > I propose that it might very well allow for drastic reduction of
things like
> > stream noise, as long as you orient yourself so that the stream
is directly
> > behind you. Thus, if the noise-cancelling mike is mounted along
the edge of
> > the parabola, in the same plane as the mike mounted in the
parabola, then
> > arrival time will be virtually the same and the annoying stream
noise could
> > be almost completely eliminated. Sure, there would be a lot of
complicated
> > things happening as it concerns other off-axis nature sounds
(such as
> > another bird singing thirty degrees off-axis), but this might not
matter
> > unless such off-axis sounds are loud. It seems it would be far
more
> > important to render your main subject cleanly while completely
eliminating
> > both rumble and broadband sound such as a distant stream or
waterfall placed
> > directly behind you.
> >
> >
> > Lang
> >
> >> Thanks that answers my questions on how far the mics can be apart
> >> from the source of the wave. I am hoping if my cancelling mic is
> >> directly above the parabolic mic and I have a line of site noise
> >> source the waves will be similar (ie the barking dogs). But I
don't
> >> know that yet.
> >>
> >> I will confine myself to post production cancelling in cool edit
so
> >> that I have the gain of the unwanted noise matched before
cancelling.
> >> Not to mention determination of what else was cancelled or
added. I
> >> worry that by having it real time that you can actually inject
more
> >> noise and reduce the dynamic range of the desired source.
Possibly
> >> because the cancelling mic heard sources that the parabolic does
not,
> >> as well as the target which it did.
> >>
> >> I too will play.
> >>
> >>
> >> --- In  Marty Michener <> wrote:
> >>> Dear Lang:
> >>>
> >>> The usual problem with automatic acoustical sound cancellation
is
> >> the
> >>> distance between the two mic diaphragms, in wavelengths, and the
> >> effect
> >>> this has on the phase arrivals.  To put it mildly, it destroys
any
> >>> cancellation, per se.
> >>>
> >>> If the mics are 12 inches apart, the results will be not at all
the
> >>> expected ones. This distance is one full wavelength at about 1
> >> kHz.  The
> >>> cancellation effects will only be apparent for such frequencies
as
> >> the
> >>> travel distance DIFFERENCE, to the two mics, is much less than a
> >>> wavelength.  At about 100 hz, for example, wavelength 10 ft.,
for
> >> sounds
> >>> hitting one mic one foot before the other mic, will generate a
> >> phase
> >>> difference of 360/10 =3D 36 degrees.  Soldering the mic inputs
> >> together in
> >>> reverse polarity, (180 degrees) the effective cancellation will
> >> still be
> >>> spoiled by having them from 144 to 216 degrees apart, and all in
> >> between,
> >>> exacerbated by having inexact level matches as well as phase
> >> variation.
> >>>
> >>> For tones lower than 50 Hz, you may achieve some useful
> >> attenuation, and we
> >>> all await your results.  But these are hardly the frequency
domain
> >> handled
> >>> by any parabola anyway, so typically are edited out afterward to
> >> isolate
> >>> bird sounds.  In the range from 500 Hz upward, the phase shifts
and
> >>> cancellations will be highly unpredictable and possibly even
nasty
> >>> sounding, but I do confess more than a meagre curiosity, and do
not
> >> let my
> >>> grousing ruin the tests.
> >>>
> >>>
> >>> my very best,
> >>>
> >>> Marty Michener
> >>> MIST Software Associates
> >>> 75 Hannah Drive, Hollis, NH 03049
> >>>
> >>> coming soon : EnjoyBirds bird identification software.
> >>>
> >>>
> >>> At 09:43 AM 2/11/02 -0500, Lang wrote:
> >>>> Klas and others:
> >>>>
> >>>> I am preparing to test the dual-mike, reversed polarity
technique
> >> using my
> >>>> Telinga parabola and two MKH 20 mikes. I've mounted a second
MKH
> >> twenty on
> >>>> the edge of the parabola, aiming upward. This should receive
most
> >> all
> >>>> background sound. I'm in the process of soldering a special
short
> >> cable with
> >>>> two female XLR plugs, to reverse the polarity coming from the
edge
> >>>> microphone. I'll then connect both mikes with using a "Y"
> >> connector and
> >>>> route the mixed signal to my recorder.
> >>>>
> >>>> This should be very interesting. I hadn't thought of using this
> >> technique,
> >>>> but it should drastically reduce background rumble and it
possibly
> >> could
> >>>> have minimal effect on the song of a bird, as long as the bird
is
> >>>> well-focused using the parabola. Basically, the only
attenuation
> >> of a
> >>>> "focused" bird song should be whatever the level would be at
the
> >> mike
> >>>> without using the parabola.
> >>>>
> >>>> Klas, have you actually tried using this technique in the
field?
> >> If you
> >>>> have, then what were your results?
> >>>>
> >>>> Lang
> >>>>
> >>
> >>
> >>
> >> To unsubscribe from this group, send an email to:
> >> 
> >>
> >>
> >>
> >> Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >>
> >>
> >>
> >
> >
> >
> > To unsubscribe from this group, send an email to:
> > 
> >
> >
> >
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >
> >
> >



________________________________________________________________________
________________________________________________________________________

<Prev in Thread] Current Thread [Next in Thread>
Admin

The University of NSW School of Computer and Engineering takes no responsibility for the contents of this archive. It is purely a compilation of material sent by many people to the naturerecordists mailing list. It has not been checked for accuracy nor its content verified in any way. If you wish to get material removed from the archive or have other queries about the archive e-mail Andrew Taylor at this address: andrewt@cse.unsw.EDU.AU