help in design for sound
Archiver
Posts: 46,084
I need help on a simple concept. sound. two mics on either side moves
to towards the sound source. How should I start in the design. The
mics would pickup sounds like claps etc. and then the stamp would
tell the motors to move to the direction of the sound..simple but i
need some direction! Any information would be sincerely appreciated!!!
Keoki
to towards the sound source. How should I start in the design. The
mics would pickup sounds like claps etc. and then the stamp would
tell the motors to move to the direction of the sound..simple but i
need some direction! Any information would be sincerely appreciated!!!
Keoki
Comments
audio sources what the directional nature of the sound is. I knew some
people who were doing doctorate-level VR research on generation /
detection of 3d sound ; you might want to see if you can pull up those
abstracts.
On Sat, 9 Dec 2000, Keoki wrote:
> I need help on a simple concept. sound. two mics on either side moves
> to towards the sound source. How should I start in the design. The
> mics would pickup sounds like claps etc. and then the stamp would
> tell the motors to move to the direction of the sound..simple but i
> need some direction! Any information would be sincerely appreciated!!!
>
> Keoki
>
>
>
>
>
Sean T. Lamont, CTO / Chief NetNerd, Abstract Software, Inc. (ServNet)
Seattle - Bellingham - Vancouver - Portland - Everett - Tacoma - Bremerton
email: lamont@a... WWW: http://www.serv.net
"...There's no moral, it's just a lot of stuff that happens". - H. Simpson
then servo movement no distinction between any thing other than just sound
and having a motor moving..for instance :
Right mic detects souund then motor turns right,,,left mic detects sound
then mototr turn left..this is esentially what i am looking for and just
need direction on this thanks
Keoki
>From: "Sean T. Lamont .lost." <lamont@a...>
>Reply-To: basicstamps@egroups.com
>To: basicstamps@egroups.com
>Subject: Re: [noparse][[/noparse]basicstamps] help in design for sound
>Date: Sat, 9 Dec 2000 14:59:50 -0800 (PST)
>
>
>Actually this is pretty difficult, since it's not very clear from two
>audio sources what the directional nature of the sound is. I knew some
>people who were doing doctorate-level VR research on generation /
>detection of 3d sound ; you might want to see if you can pull up those
>abstracts.
>
>On Sat, 9 Dec 2000, Keoki wrote:
>
> > I need help on a simple concept. sound. two mics on either side moves
> > to towards the sound source. How should I start in the design. The
> > mics would pickup sounds like claps etc. and then the stamp would
> > tell the motors to move to the direction of the sound..simple but i
> > need some direction! Any information would be sincerely appreciated!!!
> >
> > Keoki
> >
> >
> >
> >
> >
>
>Sean T. Lamont, CTO / Chief NetNerd, Abstract Software, Inc. (ServNet)
>Seattle - Bellingham - Vancouver - Portland - Everett - Tacoma - Bremerton
>email: lamont@a... WWW: http://www.serv.net
>"...There's no moral, it's just a lot of stuff that happens". - H. Simpson
>
>
>
>
________________________________________________________________________________\
_____
Get more from the Web. FREE MSN Explorer download : http://explorer.msn.com
a good start by taking two identical mics and housing
[noparse][[/noparse]as close as possible!] and put the outputs to a
differential amplifier [noparse][[/noparse]op amp connected differentially
with same input impedance per side - or 3 amplifier
instrumentation amp configuration].
You need a preamplifier for each microphone before you
go to the differential amplifier...
Now the diff amp output will be large when the two mics
are not the same input sound... gives you "off center"
signal.
Which side is less? Feed the outputs of the two microphone
preamplifiers also to two rectifiers and look for
which signal is larger [noparse][[/noparse]another diff amp, or comparator]
The audio differential amplifier could also give this
information on it's output but my intuition is that it
would be harder to discern...
Considering the large likely dynamic range of audio input
you may need automatic gain controls etc.... but I'd
give it a try with the preamps, rectifiers, comparator
and differential amp.
Why not just the preamps and rectifiers? Well, the diff
amp could give you a degree of off center reading harder
to extract with a comparator. Remember the two channels
must be as identical as possible and then put in a balance
pot somewhere to trim things out... [noparse][[/noparse]change one preamp
gain should work...]
Hope this is a thought provoking start for you - I apologise
it isn't a simple complete solution.
Good luck - let us know how it works out... premena [noparse][[/noparse]ARS AJ0J]
PS Don't forget to isolate robot motor sounds from your
microphones or locate the microphones symmetrically
with the mics ... :-) You may need to restrict the mic
bandwidth of frequencies you listen for with some bandpass
filtering [noparse][[/noparse]identical per channel].
<><>
Date: Sat, 09 Dec 2000 20:33:15
From: "Keoki christopher" <horangiman@h...>
actually it is simple..because all i need is a simple sound detection and
then servo movement no distinction between any thing other than just
sound
and having a motor moving..for instance :
Right mic detects souund then motor turns right,,,left mic detects sound
then mototr turn left..this is esentially what i am looking for and just
need direction on this thanks
<><>
________________________________________________________________
GET INTERNET ACCESS FROM JUNO!
Juno offers FREE or PREMIUM Internet access for less!
Join Juno today! For your FREE software, visit:
http://dl.www.juno.com/get/tagj.
>Actually this is pretty difficult, since it's not very clear from two
>audio sources what the directional nature of the sound is. I knew some
>people who were doing doctorate-level VR research on generation /
>detection of 3d sound ; you might want to see if you can pull up those
>abstracts.
>
>On Sat, 9 Dec 2000, Keoki wrote:
>
> > I need help on a simple concept. sound. two mics on either side moves
> > to towards the sound source. How should I start in the design. The
> > mics would pickup sounds like claps etc. and then the stamp would
> > tell the motors to move to the direction of the sound..simple but i
> > need some direction! Any information would be sincerely appreciated!!!
> >
> > Keoki
> >
> >
> >
> >
> >
I think you will need at least 3 mics (easier with 4) to determine the sound
direction.
Take for instance the case of two mics...
EACH mic would have it's own pre-amp, and then possibly a peak detector fed
into
an ADC (Analog to Digital converter that could be read by the stamp)
What you are "looking" for is symmetry between the two opposite mic signals.
With two mics it is easy to determine left and right but when the signal is
straight ahead or straight behind it becomes difficult to distinguish with
just two mics, hence a third or a fourth mic located adjacent to the two
original
mics.
1)
Lmic <
> Rmic
-good for Left and Right, but not for front and back
2)
Cmic
|
|
Lmic <
o
> Rmic
-good for left, right, front, and back, but requires more
software overhead to combine the Lmic and Rmic to act like
a back mic to make front back detection
3)
Fmic
|
|
Lmic <
o
> Rmic
|
|
Bmic
-good for left, right, front, and back detection if used in
two separate comparison pairs... L & R and F & B
with minimal software overhead.
Beau Schwabe IC Mask Designer
National Semiconductor Network Products Division
500 Pinnacle Court, Suite 525 Mail Stop GA1 Norcross, GA 30071
if you can hook up 4, 5 or even 10 mic in a circular configuration.
using directional mics, you can easily build a program to tell which
receiving angle is getting the most sound.
Of course you need so circuitry to amplify/condition the signal from
the mics to the BS (Beau Schwabe suggestion for example).
I have done something similar with a solar telescope, where 8
photo-cells can look at the image of the sun on a screen and keep
checking untill the image is centered and all cells are receiving the
same signal.
Programming wise is rather simple provided that you have some clean
signals and good definition in the difference between the mics. On
complication here is possible translating polar coordinates of the
incoming signal to the Cartesian coordinate of the robot movement.
good luck...
Al Najjar
--- In basicstamps@egroups.com, Beau Schwabe <bschwabe@a...> wrote:
> At 02:59 PM 12/9/00 -0800, you wrote:
>
> >Actually this is pretty difficult, since it's not very clear from
two
> >audio sources what the directional nature of the sound is. I knew
some
> >people who were doing doctorate-level VR research on generation /
> >detection of 3d sound ; you might want to see if you can pull up
those
> >abstracts.
> >
> >On Sat, 9 Dec 2000, Keoki wrote:
> >
> > > I need help on a simple concept. sound. two mics on either side
moves
> > > to towards the sound source. How should I start in the design.
The
> > > mics would pickup sounds like claps etc. and then the stamp
would
> > > tell the motors to move to the direction of the sound..simple
but i
> > > need some direction! Any information would be sincerely
appreciated!!!
> > >
> > > Keoki
> > >
> > >
> > >
> > >
> > >
>
> I think you will need at least 3 mics (easier with 4) to determine
the sound
> direction.
>
> Take for instance the case of two mics...
>
> EACH mic would have it's own pre-amp, and then possibly a peak
detector fed
> into
> an ADC (Analog to Digital converter that could be read by the stamp)
>
> What you are "looking" for is symmetry between the two opposite mic
signals.
> With two mics it is easy to determine left and right but when the
signal is
> straight ahead or straight behind it becomes difficult to
distinguish with
> just two mics, hence a third or a fourth mic located adjacent to the
two
> original
> mics.
>
> 1)
>
> Lmic <
> Rmic
>
>
> -good for Left and Right, but not for front and back
>
>
>
> 2)
> Cmic
> |
> |
> Lmic <
o
> Rmic
>
> -good for left, right, front, and back, but requires more
> software overhead to combine the Lmic and Rmic to act like
> a back mic to make front back detection
>
>
> 3)
> Fmic
> |
> |
> Lmic <
o
> Rmic
> |
> |
> Bmic
>
> -good for left, right, front, and back detection if used in
> two separate comparison pairs... L & R and F & B
> with minimal software overhead.
>
>
>
>
> Beau Schwabe IC Mask Designer
> National Semiconductor Network Products Division
> 500 Pinnacle Court, Suite 525 Mail Stop GA1 Norcross, GA 30071
omni-directional mikes. But borrowing from nature, you could get by with
fewer mikes, if they were directional. The pinna (outer ear) of animals
adds front-to-back directionality. Animals can also turn their heads toward
the sound to find a null position for equal loudness. On a robot, a novel
approach would be to mount a directional mike on a motor that constantly
scans through +/- 180 degrees, similar to air traffic control radar.
Back-and-forth scanning instead of continuous rotation would be useful to
avoid using slip rings for electrical contacts.
Dennis
Original Message
From: Beau Schwabe <bschwabe@a...>
To: basicstamps@egroups.com <basicstamps@egroups.com>
Date: Monday, December 11, 2000 7:35 AM
Subject: Re: [noparse][[/noparse]basicstamps] help in design for sound
>At 02:59 PM 12/9/00 -0800, you wrote:
>
>>Actually this is pretty difficult, since it's not very clear from two
>>audio sources what the directional nature of the sound is. I knew some
>>people who were doing doctorate-level VR research on generation /
>>detection of 3d sound ; you might want to see if you can pull up those
>>abstracts.
>>
>>On Sat, 9 Dec 2000, Keoki wrote:
>>
>> > I need help on a simple concept. sound. two mics on either side moves
>> > to towards the sound source. How should I start in the design. The
>> > mics would pickup sounds like claps etc. and then the stamp would
>> > tell the motors to move to the direction of the sound..simple but i
>> > need some direction! Any information would be sincerely appreciated!!!
>> >
>> > Keoki
>> >
>> >
>> >
>> >
>> >
>
>I think you will need at least 3 mics (easier with 4) to determine the
sound
>direction.
>
>Take for instance the case of two mics...
>
>EACH mic would have it's own pre-amp, and then possibly a peak detector fed
>into
>an ADC (Analog to Digital converter that could be read by the stamp)
>
>What you are "looking" for is symmetry between the two opposite mic
signals.
>With two mics it is easy to determine left and right but when the signal is
>straight ahead or straight behind it becomes difficult to distinguish with
>just two mics, hence a third or a fourth mic located adjacent to the two
>original
>mics.
>
>1)
>
>Lmic <
> Rmic
>
>
>-good for Left and Right, but not for front and back
>
>
>
>2)
> Cmic
> |
> |
>Lmic <
o
> Rmic
>
>-good for left, right, front, and back, but requires more
> software overhead to combine the Lmic and Rmic to act like
> a back mic to make front back detection
>
>
>3)
> Fmic
> |
> |
>Lmic <
o
> Rmic
> |
> |
> Bmic
>
>-good for left, right, front, and back detection if used in
> two separate comparison pairs... L & R and F & B
> with minimal software overhead.
>
>
>
>
>Beau Schwabe IC Mask Designer
>National Semiconductor Network Products Division
>500 Pinnacle Court, Suite 525 Mail Stop GA1 Norcross, GA 30071
>
>
>
>
>
>
>
>
>
>
An excellent application for a rotating transformer Dennis! I believe
there was an article in Nuts-n-Volts about 5 or so years ago that did
this with sonar to 'map' objects, by displaying the results on a PC.
Doing the same thing in the audible spectrum should prove interesting
mapping the 'sound sources' rather than the objects in a room.
>Beau's suggestion of 3 or 4 mikes is excellent for the case of stationary,
>omni-directional mikes. But borrowing from nature, you could get by with
>fewer mikes, if they were directional. The pinna (outer ear) of animals
>adds front-to-back directionality. Animals can also turn their heads toward
>the sound to find a null position for equal loudness. On a robot, a novel
>approach would be to mount a directional mike on a motor that constantly
>scans through +/- 180 degrees, similar to air traffic control radar.
>Back-and-forth scanning instead of continuous rotation would be useful to
>avoid using slip rings for electrical contacts.
>Dennis
>
>
> >At 02:59 PM 12/9/00 -0800, you wrote:
> >
> >>Actually this is pretty difficult, since it's not very clear from two
> >>audio sources what the directional nature of the sound is. I knew some
> >>people who were doing doctorate-level VR research on generation /
> >>detection of 3d sound ; you might want to see if you can pull up those
> >>abstracts.
> >>
> >>On Sat, 9 Dec 2000, Keoki wrote:
> >>
> >> > I need help on a simple concept. sound. two mics on either side moves
> >> > to towards the sound source. How should I start in the design. The
> >> > mics would pickup sounds like claps etc. and then the stamp would
> >> > tell the motors to move to the direction of the sound..simple but i
> >> > need some direction! Any information would be sincerely appreciated!!!
> >> >
> >> > Keoki
> >> >
> >> >
> >> >
> >> >
> >> >
Beau Schwabe IC Mask Designer
National Semiconductor Network Products Division
500 Pinnacle Court, Suite 525 Mail Stop GA1 Norcross, GA 30071