home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.electronics
- Path: sparky!uunet!cs.utexas.edu!sdd.hp.com!elroy.jpl.nasa.gov!ames!eos!jbm
- From: jbm@eos.arc.nasa.gov (Jeffrey Mulligan)
- Subject: Re: Filtering wind noise
- Message-ID: <1992Sep2.020657.27738@eos.arc.nasa.gov>
- Organization: NASA Ames Research Center
- References: <dfAW02=71de601@JUTS.ccc.amdahl.com> <24875@castle.ed.ac.uk> <1992Aug28.030426.6210@newsserver.rrzn.uni-hannover.de>
- Distribution: usa
- Date: Wed, 2 Sep 1992 02:06:57 GMT
- Lines: 46
-
- riepe@ifwm11.ifw.uni-hannover.de (Michael Riepe) writes:
-
- >In article <24875@castle.ed.ac.uk>, cam@castle.ed.ac.uk (Chris Malcolm) writes:
- >|> In article <dfAW02=71de601@JUTS.ccc.amdahl.com> drp40@cd.amdahl.com writes:
- >|>
- >|> >I'd like to be able to use two-way radios for skydiving. Problem
- >|> >is, 120mph of wind hitting up against a microphone would probably
- >|> >mask out any voices. Anyone have any ideas of how to get around
- >|> >this?
- >|>
- >|> 1) A throat mike under clothing.
- >|>
- >|> 2) Use combined ear speaker/mike, inside substantial ear muff.
- >|> The voice gets to the ear via head transmission.
-
- >3) Use 2 mikes connected to a differential amplifier.
-
- I don't get it; don't you want to sum the signals from the two mikes?
- Assuming they both pick up the same voice signals, but that the
- noise from each is uncorrelated, this would improve the SNR by sqrt(2).
-
- This reminded me of the solution to the problem of allowing pilots
- to understand voice communications on their headsets in the presence
- of lots of noise from the aircraft. Legend has it that this was
- discovered by chance when someone made a wiring error, but the idea
- is that one ear's speaker is wired up backwards, so that that ear
- gets an inverted signal. The noise, which comes throught the air
- is correlated, between the two ears, while the signal is anti-correlated.
- It is an empirical psychological finding that this improves the
- "perceptual SNR"; this is thought to be related to the mechanisms of
- sound localization.
-
- Maybe this idea could be combined with the two mike idea: one mike
- per ear. The listener's brain might do some of the work of
- separating correlated voice signals from the uncorrelated
- noises from the two mikes, as well as the uncorrelated wind noise
- generated at each ear.
-
- Unfortunately, this would require twice as many radios, or some
- kind of modulator/demodulator to send both signals using one radio...
-
- --
-
- Jeff Mulligan (jbm@eos.arc.nasa.gov)
- NASA/Ames Research Ctr., Mail Stop 262-2, Moffett Field CA, 94035
- (415) 604-3745
-