Cancelling Only the Noise from Afterburners

[Bret Cahill]

Instead of the headphones which cancel all outside sound navy jets could deploy a microphone under the jet just after takeoff and then radio the signal back to troops on the ground who wanted to attenuate the jet sound but not those trying to talk or hear other sounds. A dedicated receiver headphone would be necessary.

The signal could also be sent to the long suffering civilian residents of San Diego and Va. Beach.


Bret Cahill

 

[Jasen Betts]

On 2014-01-16, Bret Cahill <bret.e.cahill@gmail.com> wrote:

Instead of the headphones which cancel all outside sound navy jets could deploy a microphone under the jet just after takeoff and then radio the signal back to troops on the ground who wanted to attenuate the jet sound but not those trying to talk or hear other sounds. A dedicated receiver headphone would be necessary.

The signal could also be sent to the long suffering civilian residents of San Diego and Va. Beach.

Why not just send them a CD of the sound?


That'd work be just as badly for sound cancelling, without
providing a signal for homing AA missiles.

--
For a good time: install ntp

--- news://freenews.netfront.net/ - complaints: news@netfront.net ---

 

[Daniel Pitts]

On 1/16/14 9:47 AM, Bret Cahill wrote:

Instead of the headphones which cancel all outside sound navy jets could deploy a microphone under the jet just after takeoff and then radio the signal back to troops on the ground who wanted to attenuate the jet sound but not those trying to talk or hear other sounds. A dedicated receiver headphone would be necessary.

The signal could also be sent to the long suffering civilian residents of San Diego and Va. Beach.


Bret Cahill

One problem with that is that the radio signal travels much faster than
the sound signal, and the sound arrives at different locations at
different times. It will also have been distorted by reflections by the
time it reaches most people, so even if you could solve the time-delay
issue, you'll end up causing more noise, not less.

 

[Bob Masta]

On Thu, 16 Jan 2014 12:35:33 -0800, Daniel Pitts
<newsgroup.nospam@virtualinfinity.net> wrote:

On 1/16/14 9:47 AM, Bret Cahill wrote:
Instead of the headphones which cancel all outside sound navy jets could deploy a microphone under the jet just after takeoff and then radio the signal back to troops on the ground who wanted to attenuate the jet sound but not those trying to talk or hear other sounds. A dedicated receiver headphone would be necessary.

The signal could also be sent to the long suffering civilian residents of San Diego and Va. Beach.


Bret Cahill


One problem with that is that the radio signal travels much faster than
the sound signal, and the sound arrives at different locations at
different times. It will also have been distorted by reflections by the
time it reaches most people, so even if you could solve the time-delay
issue, you'll end up causing more noise, not less.

In addition, the idea of picking up the sound from a
microphone under the jet just after takeoff is a
non-starter. What makes it "noise" (as opposed to a "buzz"
or something more tonal) is that it is highly random. The
sound at one instant in time (like just after takeoff) would
not correlate (eg be cancellable) with the sound at any
other time.

Also, in addition to the reflections Daniel notes above, the
sound at any given listener will depend upon his orientation
relative to the source. It's not a point source, so a
single mic will not work unless it is near the listener.
Near the source you'd nead a large array to catch all the
separate emitters. Just like an array of speakers, the
sound at any given angle and distance is a vector sum of all
the little sources cancelling and reinforcing differently
according to their instantaneous frequencies.

Best regards,



Bob Masta

DAQARTA v7.50
Data AcQuisition And Real-Time Analysis
www.daqarta.com
Scope, Spectrum, Spectrogram, Sound Level Meter
Frequency Counter, Pitch Track, Pitch-to-MIDI
FREE Signal Generator, DaqMusiq generator
Science with your sound card!

 

[Bret Cahill]

Instead of the headphones which cancel all outside sound navy jets could deploy a microphone under the jet just after takeoff and then radio the signal back to troops on the ground who wanted to attenuate the jet sound but not those trying to talk or hear other sounds. A dedicated receiver headphone would be necessary.



The signal could also be sent to the long suffering civilian residents of San Diego and Va. Beach.

One problem with that is that the radio signal travels much faster than

the sound signal, and the sound arrives at different locations at

different times.

Match filter the noise signal. When the noise arrives and you have a match you know you have a few nano seconds to start cancelling a m sec of signal..

It will also have been distorted by reflections by the

time it reaches most people, so even if you could solve the time-delay

issue, you'll end up causing more noise, not less.

The reflections are, of course, specific to the area.

These training flights aren't really interesting. Depending on the wind direction the jets fly the same path every time.

It should be possible to develop a reflection adjustment for each path and incorporate the reflections into the noise cancellation.


Bret Cahill

 

[Bret Cahill]

Instead of the headphones which cancel all outside sound navy jets could deploy a microphone under the jet just after takeoff and then radio the signal back to troops on the ground who wanted to attenuate the jet sound but not those trying to talk or hear other sounds. A dedicated receiver headphone would be necessary.



The signal could also be sent to the long suffering civilian residents of San Diego and Va. Beach.

One problem with that is that the radio signal travels much faster than

the sound signal, and the sound arrives at different locations at

different times. It will also have been distorted by reflections by the

time it reaches most people, so even if you could solve the time-delay

issue, you'll end up causing more noise, not less.



In addition, the idea of picking up the sound from a

microphone under the jet just after takeoff is a

non-starter. What makes it "noise" (as opposed to a "buzz"

or something more tonal) is that it is highly random. The

sound at one instant in time (like just after takeoff) would

not correlate (eg be cancellable) with the sound at any

other time.

But the changes aren't random. They can be predicted.

The happy fact is we can know _everything_ that is going to happen to that noise wave over Seatack or Pt. Loma.

Also, in addition to the reflections Daniel notes above, the

sound at any given listener will depend upon his orientation

relative to the source. It's not a point source, so a

single mic will not work unless it is near the listener.

Near the source you'd nead a large array to catch all the

separate emitters. Just like an array of speakers, the

sound at any given angle and distance is a vector sum of all

the little sources cancelling and reinforcing differently

according to their instantaneous frequencies.

The navy might be able to afford a few mics/$80 million aircraft.

"They put up a stop sign. I take it down."

-- Satchmo

DAQARTA v7.50

Data AcQuisition And Real-Time Analysis

www.daqarta.com

Scope, Spectrum, Spectrogram, Sound Level Meter

Frequency Counter, Pitch Track, Pitch-to-MIDI

FREE Signal Generator, DaqMusiq generator

Science with your sound card!

 

[John Larkin]

On Thu, 16 Jan 2014 12:35:33 -0800, Daniel Pitts
<newsgroup.nospam@virtualinfinity.net> wrote:

On 1/16/14 9:47 AM, Bret Cahill wrote:
Instead of the headphones which cancel all outside sound navy jets could deploy a microphone under the jet just after takeoff and then radio the signal back to troops on the ground who wanted to attenuate the jet sound but not those trying to talk or hear other sounds. A dedicated receiver headphone would be necessary.

The signal could also be sent to the long suffering civilian residents of San Diego and Va. Beach.


Bret Cahill


One problem with that is that the radio signal travels much faster than
the sound signal, and the sound arrives at different locations at
different times. It will also have been distorted by reflections by the
time it reaches most people, so even if you could solve the time-delay
issue, you'll end up causing more noise, not less.

Don't expose Bret to science. It confuses him.


--

John Larkin Highland Technology, Inc

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[Bret Cahill]

Instead of the headphones which cancel all outside sound navy jets cou=

ld deploy a microphone under the jet just after takeoff and then radio the =

signal back to troops on the ground who wanted to attenuate the jet sound b=

ut not those trying to talk or hear other sounds. A dedicated receiver hea=

dphone would be necessary.

=20



=20

The signal could also be sent to the long suffering civilian residents=

of San Diego and Va. Beach.



One problem with that is that the radio signal travels much faster than=

=20



the sound signal, and the sound arrives at different locations at=20



different times. It will also have been distorted by reflections by the=

=20

=20

time it reaches most people, so even if you could solve the time-delay=

=20

=20

issue, you'll end up causing more noise, not less.

=20

=20

=20

In addition, the idea of picking up the sound from a

=20

microphone under the jet just after takeoff is a

=20

non-starter. What makes it "noise" (as opposed to a "buzz"

=20

or something more tonal) is that it is highly random. The

=20

sound at one instant in time (like just after takeoff) would

=20

not correlate (eg be cancellable) with the sound at any

=20

other time. =20



But the changes aren't random. They can be predicted.



The happy fact is we can know _everything_ that is going to happen to that =

noise wave over Seatack or Pt. Loma.



The changes I'm talking about are in the source, due to the

way a jet engine works. It's pretty much a pure random

noise generator, due to turbulence in the engine. It's not

likely this will *ever* be predictable, with any amount of

computing power.

The noise from the engine isn't going to predicted. It's going to be recorded in short increments and radioed back to the ground.

Also, in addition to the reflections Daniel notes above, the

=20

sound at any given listener will depend upon his orientation

=20

relative to the source. It's not a point source, so a

=20

single mic will not work unless it is near the listener.

=20

Near the source you'd nead a large array to catch all the

=20

separate emitters. Just like an array of speakers, the

=20

sound at any given angle and distance is a vector sum of all

=20

the little sources cancelling and reinforcing differently

=20

according to their instantaneous frequencies.



The navy might be able to afford a few mics/$80 million aircraft.



So what are you proposing? Suppose each engine of each jet

drags a big ring of mics behind it, to catch sound from

multiple angles. Somehow, they will not interfere with

ariflow or create drag, and will be able to withstand the

temperatures.



Then what? A listener at an arbitrary (and constantly

changing) distance and angle would need to get a continuous

feed from each mic, along with accurate instantaneous

position measurements.

GPS is faster than sound.

Obviously you don't want to broadcast the jet's location to a sophisticated foe but we hardly ever fight anyone with any tech other than RPGs.

And for civilian areas which like to be in the same real estate market as the navy, i.e., ocean front property, broadcasting GPS data would not be an issue at all.

Then it could compute the

instantaneous vector sum for that angle at the source. Then

it would need to compute the arriving waveform at the

listening location, using its own built-in weather radar to

compensate for layers of different windspeeds, temperature,

humidity, pressure, etc... all of which affect propagation

delay.

The OP which was deliberately shortened to save time. It's not trivial. But it isn't impossible or even prohibitively difficult either.

There are many issues/corrections that can be factored in with every day weather data, flight data, terrain data like overpasses, piers, etc.

It's a fun engineering problem that may yield other completely unrelated inventions.

But let's say we knew this *exactly* for the point in space

where the listener's head is. Oh, but wait a second... he

has *two ears*, facing different directions and separated by

a significant distance (in terms of the wavelengths we want

to cancel). No problem, we just double the amount of

equipment the guy has to lug. Ahh, let's see... how much

does *one* IBM "Deep Blue" weigh, anyhow? (Not to mention

the portable weather radar station...)

Conventional noise cancellation headphones do not eliminate _all_ the noise, yet Walmart has Sonys on sale for $50.

Why is the standard perfection here?

"Don't make the perfect the enemy of the possible."

Bob Masta



DAQARTA v7.50

Data AcQuisition And Real-Time Analysis

www.daqarta.com

Scope, Spectrum, Spectrogram, Sound Level Meter

Frequency Counter, Pitch Track, Pitch-to-MIDI

FREE Signal Generator, DaqMusiq generator

Science with your sound card!

 

[Bob Masta]

On Fri, 17 Jan 2014 20:28:08 -0800 (PST), Bret Cahill
<bret.e.cahill@gmail.com> wrote:

Instead of the headphones which cancel all outside sound navy jets cou=
ld deploy a microphone under the jet just after takeoff and then radio the =
signal back to troops on the ground who wanted to attenuate the jet sound b=
ut not those trying to talk or hear other sounds. A dedicated receiver hea=
dphone would be necessary.
=20

=20
The signal could also be sent to the long suffering civilian residents=
of San Diego and Va. Beach.

One problem with that is that the radio signal travels much faster than=
=20

the sound signal, and the sound arrives at different locations at=20

different times. It will also have been distorted by reflections by the=
=20
=20
time it reaches most people, so even if you could solve the time-delay=
=20
=20
issue, you'll end up causing more noise, not less.
=20
=20
=20
In addition, the idea of picking up the sound from a
=20
microphone under the jet just after takeoff is a
=20
non-starter. What makes it "noise" (as opposed to a "buzz"
=20
or something more tonal) is that it is highly random. The
=20
sound at one instant in time (like just after takeoff) would
=20
not correlate (eg be cancellable) with the sound at any
=20
other time. =20

But the changes aren't random. They can be predicted.

The happy fact is we can know _everything_ that is going to happen to that =
noise wave over Seatack or Pt. Loma.

The changes I'm talking about are in the source, due to the
way a jet engine works. It's pretty much a pure random
noise generator, due to turbulence in the engine. It's not
likely this will *ever* be predictable, with any amount of
computing power.

Also, in addition to the reflections Daniel notes above, the
=20
sound at any given listener will depend upon his orientation
=20
relative to the source. It's not a point source, so a
=20
single mic will not work unless it is near the listener.
=20
Near the source you'd nead a large array to catch all the
=20
separate emitters. Just like an array of speakers, the
=20
sound at any given angle and distance is a vector sum of all
=20
the little sources cancelling and reinforcing differently
=20
according to their instantaneous frequencies.

The navy might be able to afford a few mics/$80 million aircraft.

So what are you proposing? Suppose each engine of each jet
drags a big ring of mics behind it, to catch sound from
multiple angles. Somehow, they will not interfere with
ariflow or create drag, and will be able to withstand the
temperatures.

Then what? A listener at an arbitrary (and constantly
changing) distance and angle would need to get a continuous
feed from each mic, along with accurate instantaneous
position measurements. Then it could compute the
instantaneous vector sum for that angle at the source. Then
it would need to compute the arriving waveform at the
listening location, using its own built-in weather radar to
compensate for layers of different windspeeds, temperature,
humidity, pressure, etc... all of which affect propagation
delay.

But let's say we knew this *exactly* for the point in space
where the listener's head is. Oh, but wait a second... he
has *two ears*, facing different directions and separated by
a significant distance (in terms of the wavelengths we want
to cancel). No problem, we just double the amount of
equipment the guy has to lug. Ahh, let's see... how much
does *one* IBM "Deep Blue" weigh, anyhow? (Not to mention
the portable weather radar station...)

Best regards,


Bob Masta

DAQARTA v7.50
Data AcQuisition And Real-Time Analysis
www.daqarta.com
Scope, Spectrum, Spectrogram, Sound Level Meter
Frequency Counter, Pitch Track, Pitch-to-MIDI
FREE Signal Generator, DaqMusiq generator
Science with your sound card!

 

[Bret Cahill]

Instead of the headphones which cancel all outside sound navy jets cou


ld deploy a microphone under the jet just after takeoff and then radio the


signal back to troops on the ground who wanted to attenuate the jet sound b


ut not those trying to talk or hear other sounds. A dedicated receiver hea


dphone would be necessary.



=20







=20



The signal could also be sent to the long suffering civilian residents


of San Diego and Va. Beach.







One problem with that is that the radio signal travels much faster than


=20







the sound signal, and the sound arrives at different locations at=20







different times. It will also have been distorted by reflections by the


=20



=20



time it reaches most people, so even if you could solve the time-delay


=20



=20



issue, you'll end up causing more noise, not less.



=20



=20



=20



In addition, the idea of picking up the sound from a



=20



microphone under the jet just after takeoff is a



=20



non-starter. What makes it "noise" (as opposed to a "buzz"



=20



or something more tonal) is that it is highly random. The



=20



sound at one instant in time (like just after takeoff) would



=20



not correlate (eg be cancellable) with the sound at any



=20



other time. =20







But the changes aren't random. They can be predicted.







The happy fact is we can know _everything_ that is going to happen to that


noise wave over Seatack or Pt. Loma.







The changes I'm talking about are in the source, due to the



way a jet engine works. It's pretty much a pure random



noise generator, due to turbulence in the engine. It's not



likely this will *ever* be predictable, with any amount of



computing power.



The noise from the engine isn't going to predicted. It's going to be recorded in short increments and radioed back to the ground.



I think you still aren't "getting it" here. There isn't a

single "noise" coming from the jet. It's the sum of all the

little emitting regions in the turbulent flow. That means

that the waveform (the thing you want to cancel) is

different at every angle from the engine, due to

constructive and destructive interference from all those

millions of emitters .

The sound from an emitter on one side of the noise generation volume will obviously take longer than that from an emitter on the other side to reach a mic on the other side. It will also be somewhat fainter. Both these factors will certainly give a different wave form for each direction just as stereo has 2 different wave forms.

One difference with stereo is the time lag from different emitters isn't changed. Here we adjust the lag according to direction of the person on the ground.

If you hang a single mic somewhere

to try to record "the noise", it will only tell you about

the waveform at that one position... totally worthless in

predicting what listeners on the ground will get.

The navy ought to be good for a 3 - 4 mics / $60 million aircraft. Then it's possible to back calculate numbers for everything about every emitter.

But we don't need numbers for every emitter here any more than in stereo recordings.

And what they get is continually changing with the relative

position of the passing jet.

The navy should know the speed of the aircraft well enough to correct for Doppler.

Perhaps you have heard the

phenomenon called "jetsounds" or "flanging" in music, or

noted the phenomenon they are named after at an airport. As

a jet moves relative to the listener, he hears a hollow

"sweeping" effect that isn't present when the jet is

stationary. That's due to his ears getting the sum of the

direct sound, plus delayed versions reflected off the

surroundings (runway, buildings, etc).

We know every last detail about the surface, the exact location of every hill, bridge, jetty, street and house.

A more difficult problem might be reflections off of fast moving clouds but the navy should have real time information on that as well.

If there is a single

reflection plus the source, the resultant spectrum will have

a series of dips and peaks called "comb filtering", due to

the interference at wavelengths that are multiples of the

path difference. (See <http://www.daqarta.com/dw_0aff.htm

for a more detailed explanation, and a link to using the

free Daqarta generator to demonstrate it.) One curious

thing about this is that you only notice this while there is

motion, and the dip-and-peak pattern is changing... you

don't notice the dips and peaks when they are static.



Anyway, that effect changes the waveform,

And we know exactly how it changes every waveform.

and it's that

waveform that you need if you want to cancel anything.



I suspect you are thinking of a tonal source, especially

when you mention "recodred in short increments". Yes, if

the jet engine sound was (say) a few strong tones and

harmonics, and they changed only slowly with time, then

there might be a glimmer of hope for your proposal. The

ground listener's equipment could get a list of frequencies

and amplitudes by radio, and set up a matched filter to

cancel them, and only them. This would not use the same

principle as noise cancellation, which requires the exact

waveform and simply subtracts a gain-adjusted version from

the local sound.



But a matched filter can't work with a broadband noise

source like a jet engine, where there is energy at all

frequencies, and everything is constantly changing. You

have to use subtractive cancellation like the

noise-cancelling headphones use, and that requires that you

know the exact waveform of the incoming sound *at the

instant it arrives*. There is no way to get this sent from

the jet... it doesn't have "the waveform" to send. There is

no way to compute it from one or even quite a few mics at

the jet,

Are a lot of mics and calculations necessary for the diva to sound like she's in center stage in a stereo recording?

even if you had a perfectly-known air column and no

reflections. Even if you knew the exact distance, etc, etc.



So, why attack the impossible?

This one is good for more than noise abatement. It may lead to other problems to solve.

If the need is to cancel the

jet noise for troops on the ground, a standard

noise-cancelling approach will do the best job, with mics on

helmets to pick up the ambient noise. If the troops need to

talk among themselves, they can use lip mics or even throat

mics. This is a tried-and-tested approach.



If you *really* want to hear ambient sounds on the ground,

the only hope would be to have an upward-pointing mic as

input to the canceller, and hope that it didn't pick up too

much of the ground sounds. This would only have a chance if

the jet was way overhead, so that the arriving sound

appeared to be coming from a distant point, and not too many

ground reflections. But then you probably don't need it!



Best regards,







Bob Masta



DAQARTA v7.50

Data AcQuisition And Real-Time Analysis

www.daqarta.com

Scope, Spectrum, Spectrogram, Sound Level Meter

Frequency Counter, Pitch Track, Pitch-to-MIDI

FREE Signal Generator, DaqMusiq generator

Science with your sound card!

 

[RobertMacy]

On Sat, 18 Jan 2014 06:37:40 -0700, Bob Masta <N0Spam@daqarta.com> wrote:

...snip...
...snip to keep Aioe happy

But let's say we knew this *exactly* for the point in space
where the listener's head is. Oh, but wait a second... he
has *two ears*, facing different directions and separated by
a significant distance (in terms of the wavelengths we want
to cancel). No problem, we just double the amount of
equipment the guy has to lug. Ahh, let's see... how much
does *one* IBM "Deep Blue" weigh, anyhow? (Not to mention
the portable weather radar station...)

Best regards,


Bob Masta
DAQARTA v7.50
Data AcQuisition And Real-Time Analysis
www.daqarta.com
Scope, Spectrum, Spectrogram, Sound Level Meter
Frequency Counter, Pitch Track, Pitch-to-MIDI
FREE Signal Generator, DaqMusiq generator
Science with your sound card!

You're spot on saying that 'active' form of cancellation is a bit complex.
[hi, Bob!] The microphone functions can be replaced with other ways to
measure 'acoustic disturbance' without the drag, etc, but not very
important unless you can figure out what to do with the information. As
you point out, even with 'total' information cancellation for ALL is a bit
daunting compared to cancellation for ONE. A lot to do with the dispersive
nature of atmosphere and the placement of all those audial 'receivers'

I proposed a method a few years ago for mitigating landing and takeoff
noises around commercial/military airports. It requires some humongous
electrical power, but can drop noise for EVERYONE outside by over 16-30
dB. Doesn't 'sound' like much, but sound is a bit insidious, once made
difficult to get rid of, and usually your only hope is blocking/reflecting.

So, what other ways are used? From memory I believe there are some
techniques done right at, or near, the source, very imaginative solutions
too, but the details are still classified. And don't know what, if any, is
the overhead cost to flight performance is.

 

[Bob Masta]

On Sat, 18 Jan 2014 08:40:46 -0800 (PST), Bret Cahill
<bret.e.cahill@gmail.com> wrote:

Instead of the headphones which cancel all outside sound navy jets cou=

ld deploy a microphone under the jet just after takeoff and then radio the =

signal back to troops on the ground who wanted to attenuate the jet sound b=

ut not those trying to talk or hear other sounds. A dedicated receiver hea=

dphone would be necessary.

=20



=20

The signal could also be sent to the long suffering civilian residents=

of San Diego and Va. Beach.



One problem with that is that the radio signal travels much faster than=

=20



the sound signal, and the sound arrives at different locations at=20



different times. It will also have been distorted by reflections by the=

=20

=20

time it reaches most people, so even if you could solve the time-delay=

=20

=20

issue, you'll end up causing more noise, not less.

=20

=20

=20

In addition, the idea of picking up the sound from a

=20

microphone under the jet just after takeoff is a

=20

non-starter. What makes it "noise" (as opposed to a "buzz"

=20

or something more tonal) is that it is highly random. The

=20

sound at one instant in time (like just after takeoff) would

=20

not correlate (eg be cancellable) with the sound at any

=20

other time. =20



But the changes aren't random. They can be predicted.



The happy fact is we can know _everything_ that is going to happen to that =

noise wave over Seatack or Pt. Loma.



The changes I'm talking about are in the source, due to the

way a jet engine works. It's pretty much a pure random

noise generator, due to turbulence in the engine. It's not

likely this will *ever* be predictable, with any amount of

computing power.

The noise from the engine isn't going to predicted. It's going to be recorded in short increments and radioed back to the ground.

I think you still aren't "getting it" here. There isn't a
single "noise" coming from the jet. It's the sum of all the
little emitting regions in the turbulent flow. That means
that the waveform (the thing you want to cancel) is
different at every angle from the engine, due to
constructive and destructive interference from all those
millions of emitters . If you hang a single mic somewhere
to try to record "the noise", it will only tell you about
the waveform at that one position... totally worthless in
predicting what listeners on the ground will get.

And what they get is continually changing with the relative
position of the passing jet. Perhaps you have heard the
phenomenon called "jetsounds" or "flanging" in music, or
noted the phenomenon they are named after at an airport. As
a jet moves relative to the listener, he hears a hollow
"sweeping" effect that isn't present when the jet is
stationary. That's due to his ears getting the sum of the
direct sound, plus delayed versions reflected off the
surroundings (runway, buildings, etc). If there is a single
reflection plus the source, the resultant spectrum will have
a series of dips and peaks called "comb filtering", due to
the interference at wavelengths that are multiples of the
path difference. (See <http://www.daqarta.com/dw_0aff.htm>
for a more detailed explanation, and a link to using the
free Daqarta generator to demonstrate it.) One curious
thing about this is that you only notice this while there is
motion, and the dip-and-peak pattern is changing... you
don't notice the dips and peaks when they are static.

Anyway, that effect changes the waveform, and it's that
waveform that you need if you want to cancel anything.

I suspect you are thinking of a tonal source, especially
when you mention "recodred in short increments". Yes, if
the jet engine sound was (say) a few strong tones and
harmonics, and they changed only slowly with time, then
there might be a glimmer of hope for your proposal. The
ground listener's equipment could get a list of frequencies
and amplitudes by radio, and set up a matched filter to
cancel them, and only them. This would not use the same
principle as noise cancellation, which requires the exact
waveform and simply subtracts a gain-adjusted version from
the local sound.

But a matched filter can't work with a broadband noise
source like a jet engine, where there is energy at all
frequencies, and everything is constantly changing. You
have to use subtractive cancellation like the
noise-cancelling headphones use, and that requires that you
know the exact waveform of the incoming sound *at the
instant it arrives*. There is no way to get this sent from
the jet... it doesn't have "the waveform" to send. There is
no way to compute it from one or even quite a few mics at
the jet, even if you had a perfectly-known air column and no
reflections. Even if you knew the exact distance, etc, etc.

So, why attack the impossible? If the need is to cancel the
jet noise for troops on the ground, a standard
noise-cancelling approach will do the best job, with mics on
helmets to pick up the ambient noise. If the troops need to
talk among themselves, they can use lip mics or even throat
mics. This is a tried-and-tested approach.

If you *really* want to hear ambient sounds on the ground,
the only hope would be to have an upward-pointing mic as
input to the canceller, and hope that it didn't pick up too
much of the ground sounds. This would only have a chance if
the jet was way overhead, so that the arriving sound
appeared to be coming from a distant point, and not too many
ground reflections. But then you probably don't need it!

Best regards,



Bob Masta

DAQARTA v7.50
Data AcQuisition And Real-Time Analysis
www.daqarta.com
Scope, Spectrum, Spectrogram, Sound Level Meter
Frequency Counter, Pitch Track, Pitch-to-MIDI
FREE Signal Generator, DaqMusiq generator
Science with your sound card!

 

[Bob Masta]

On Sun, 19 Jan 2014 09:07:41 -0800 (PST), Bret Cahill
<bret.e.cahill@gmail.com> wrote:

<snipped originals that had a bazillion added carrriage
returns>

The sound from an emitter on one side of the noise generation volume will o=
bviously take longer than that from an emitter on the other side to reach a=
mic on the other side. It will also be somewhat fainter. Both these fact=
ors will certainly give a different wave form for each direction just as st=
ereo has 2 different wave forms.

One difference with stereo is the time lag from different emitters isn't ch=
anged. Here we adjust the lag according to direction of the person on the =
ground.

If you hang a single mic somewhere
=20
to try to record "the noise", it will only tell you about
=20
the waveform at that one position... totally worthless in
=20
predicting what listeners on the ground will get.

The navy ought to be good for a 3 - 4 mics / $60 million aircraft. Then it=
's possible to back calculate numbers for everything about every emitter.

But we don't need numbers for every emitter here any more than in stereo re=
cordings.

Sorry, but if you were trying to cancel a stereo recording
*at the listener* using only information from the *source*,
you *would* need the waveforms of each emitter (each
speaker), *and* their angles and distances, as well as all
relevant info about intervening transmission/propagation
effects. That's very different from cancelling using mics
at the listener.

And what they get is continually changing with the relative
=20
position of the passing jet. =20

The navy should know the speed of the aircraft well enough to correct for D=
oppler.

We haven't even gotten around to Doppler effects yet. We're
just talking about relative phase changes due to angle and
distance. (Not even including the non-trivial effects of
atmospheric phenomena on transmission.)

Perhaps you have heard the
=20
phenomenon called "jetsounds" or "flanging" in music, or
=20
noted the phenomenon they are named after at an airport. As
=20
a jet moves relative to the listener, he hears a hollow
=20
"sweeping" effect that isn't present when the jet is
=20
stationary. That's due to his ears getting the sum of the
=20
direct sound, plus delayed versions reflected off the
=20
surroundings (runway, buildings, etc). =20

We know every last detail about the surface, the exact location of every hi=
ll, bridge, jetty, street and house.

Huh? Dunno about you, but not even the NSA has a fraction
of the "details" needed for this sort of thing. These are
constantly-changing details, dependent on every surface and
the relative angles of source and listener. It's way more
involved than the seemingly-similar computer-generated
graphics problem, where you "only" have to know millions of
angles of incidence and reflection and a few intensities...
those guys don't have to worry about keeping track of
waveforms. And you may have heard about the huge computing
power and time they use.

A more difficult problem might be reflections off of fast moving clouds but=
the navy should have real time information on that as well.

If there is a single
=20
reflection plus the source, the resultant spectrum will have
=20
a series of dips and peaks called "comb filtering", due to
=20
the interference at wavelengths that are multiples of the
=20
path difference. (See <http://www.daqarta.com/dw_0aff.htm
=20
for a more detailed explanation, and a link to using the
=20
free Daqarta generator to demonstrate it.) One curious
=20
thing about this is that you only notice this while there is
=20
motion, and the dip-and-peak pattern is changing... you
=20
don't notice the dips and peaks when they are static.
=20
=20
=20
Anyway, that effect changes the waveform,=20

And we know exactly how it changes every waveform.

and it's that
=20
waveform that you need if you want to cancel anything.
=20
=20
=20
I suspect you are thinking of a tonal source, especially
=20
when you mention "recodred in short increments". Yes, if
=20
the jet engine sound was (say) a few strong tones and
=20
harmonics, and they changed only slowly with time, then
=20
there might be a glimmer of hope for your proposal. The
=20
ground listener's equipment could get a list of frequencies
=20
and amplitudes by radio, and set up a matched filter to
=20
cancel them, and only them. This would not use the same
=20
principle as noise cancellation, which requires the exact
=20
waveform and simply subtracts a gain-adjusted version from
=20
the local sound.
=20
=20
=20
But a matched filter can't work with a broadband noise
=20
source like a jet engine, where there is energy at all
=20
frequencies, and everything is constantly changing. You
=20
have to use subtractive cancellation like the
=20
noise-cancelling headphones use, and that requires that you
=20
know the exact waveform of the incoming sound *at the
=20
instant it arrives*. There is no way to get this sent from
=20
the jet... it doesn't have "the waveform" to send. There is
=20
no way to compute it from one or even quite a few mics at
=20
the jet,=20

Are a lot of mics and calculations necessary for the diva to sound like she=
's in center stage in a stereo recording?

They would be if you were proposing to have her swinging
thousands of feet overhead at hundreds of miles per hour,
with an unknown atmosphere in between, and unknown
reflection sources on the ground, and arbitrary listener
positions. (Actually, the stereo effect in most recorded
music is generated strictly by relative amplitude
differences at the ears, without regard to timing/phase
differences. That's how recording engineers can change
perceived position so easily... they just fade one channel
up and the other down. That won't work for cancellation.)

even if you had a perfectly-known air column and no
=20
reflections. Even if you knew the exact distance, etc, etc.
=20
=20
=20
So, why attack the impossible? =20

This one is good for more than noise abatement. It may lead to other probl=
ems to solve.

You mean problems like understanding sound propagation,
wavefront interference, and random sources? <g>

Best regards,



Bob Masta

DAQARTA v7.50
Data AcQuisition And Real-Time Analysis
www.daqarta.com
Scope, Spectrum, Spectrogram, Sound Level Meter
Frequency Counter, Pitch Track, Pitch-to-MIDI
FREE Signal Generator, DaqMusiq generator
Science with your sound card!

 

[RobertMacy]

On Mon, 20 Jan 2014 07:46:54 -0700, Bob Masta <N0Spam@daqarta.com> wrote:

...snip to keep Aioe happy
They would be if you were proposing to have her swinging
thousands of feet overhead at hundreds of miles per hour,
with an unknown atmosphere in between, and unknown
reflection sources on the ground, and arbitrary listener
positions. (Actually, the stereo effect in most recorded
music is generated strictly by relative amplitude
differences at the ears, without regard to timing/phase
differences. That's how recording engineers can change
perceived position so easily... they just fade one channel
up and the other down. That won't work for cancellation.)

that's why I hate recorded music! well, sloppily recorded music!

Thanks. Didn't know what was causing the 'shallowness' and lack of
dimension in a recording.

For example, I once listened to a Direct Recording [which means
unadulterated and not homogenized by some sound engineer] of an old guitar
player. The flash of insight into the performance was incredible. (You
can't do this purposely has to ome like when you're daydreaming while
listening), but you could envision the whole room, size of room, where he
was sitting, even openings, doors, etc Absolutely, amazing! Only after
hearing the recording and experiencing that 'visualization' did I get a
chance to read about the history of the recording. The text pretty much
described what one 'heard'

After that experience, it's really sadly thwarting to hear the way music
is butchered, compressed, mixed, etc etc

Must explain why rap gained such a large audience. At least with speech,
it was more difficult to remove the 'character' coming through.

 

[Bret Cahill]

snipped originals that had a bazillion added carrriage

returns



The sound from an emitter on one side of the noise generation volume will o=

bviously take longer than that from an emitter on the other side to reach a=

mic on the other side. It will also be somewhat fainter. Both these fact=

ors will certainly give a different wave form for each direction just as st=

ereo has 2 different wave forms.



One difference with stereo is the time lag from different emitters isn't ch=

anged. Here we adjust the lag according to direction of the person on the =

ground.



If you hang a single mic somewhere

=20

to try to record "the noise", it will only tell you about

=20

the waveform at that one position... totally worthless in

=20

predicting what listeners on the ground will get.



The navy ought to be good for a 3 - 4 mics / $60 million aircraft. Then it=

's possible to back calculate numbers for everything about every emitter.



But we don't need numbers for every emitter here any more than in stereo re=

cordings.



Sorry, but if you were trying to cancel a stereo recording

*at the listener* using only information from the *source*,

you *would* need the waveforms of each emitter (each

speaker), *and* their angles and distances, as well as all

relevant info about intervening transmission/propagation

effects. That's very different from cancelling using mics

at the listener.



And what they get is continually changing with the relative

=20

position of the passing jet. =20



The navy should know the speed of the aircraft well enough to correct for D=

oppler.



We haven't even gotten around to Doppler effects yet. We're

just talking about relative phase changes due to angle and

distance. (Not even including the non-trivial effects of

atmospheric phenomena on transmission.)



Perhaps you have heard the

=20

phenomenon called "jetsounds" or "flanging" in music, or

=20

noted the phenomenon they are named after at an airport. As

=20

a jet moves relative to the listener, he hears a hollow

=20

"sweeping" effect that isn't present when the jet is

=20

stationary. That's due to his ears getting the sum of the

=20

direct sound, plus delayed versions reflected off the

=20

surroundings (runway, buildings, etc). =20



We know every last detail about the surface, the exact location of every hi=

ll, bridge, jetty, street and house.



Huh? Dunno about you, but not even the NSA has a fraction

of the "details" needed for this sort of thing. These are

constantly-changing details, dependent on every surface and

the relative angles of source and listener. It's way more

involved than the seemingly-similar computer-generated

graphics problem, where you "only" have to know millions of

angles of incidence and reflection and a few intensities...

those guys don't have to worry about keeping track of

waveforms. And you may have heard about the huge computing

power and time they use.



A more difficult problem might be reflections off of fast moving clouds but=

the navy should have real time information on that as well.



If there is a single

=20

reflection plus the source, the resultant spectrum will have

=20

a series of dips and peaks called "comb filtering", due to

=20

the interference at wavelengths that are multiples of the

=20

path difference. (See <http://www.daqarta.com/dw_0aff.htm

=20

for a more detailed explanation, and a link to using the

=20

free Daqarta generator to demonstrate it.) One curious

=20

thing about this is that you only notice this while there is

=20

motion, and the dip-and-peak pattern is changing... you

=20

don't notice the dips and peaks when they are static.

=20

=20

=20

Anyway, that effect changes the waveform,=20



And we know exactly how it changes every waveform.



and it's that

=20

waveform that you need if you want to cancel anything.

=20

=20

=20

I suspect you are thinking of a tonal source, especially

=20

when you mention "recodred in short increments". Yes, if

=20

the jet engine sound was (say) a few strong tones and

=20

harmonics, and they changed only slowly with time, then

=20

there might be a glimmer of hope for your proposal. The

=20

ground listener's equipment could get a list of frequencies

=20

and amplitudes by radio, and set up a matched filter to

=20

cancel them, and only them. This would not use the same

=20

principle as noise cancellation, which requires the exact

=20

waveform and simply subtracts a gain-adjusted version from

=20

the local sound.

=20

=20

=20

But a matched filter can't work with a broadband noise

=20

source like a jet engine, where there is energy at all

=20

frequencies, and everything is constantly changing. You

=20

have to use subtractive cancellation like the

=20

noise-cancelling headphones use, and that requires that you

=20

know the exact waveform of the incoming sound *at the

=20

instant it arrives*. There is no way to get this sent from

=20

the jet... it doesn't have "the waveform" to send. There is

=20

no way to compute it from one or even quite a few mics at

=20

the jet,=20



Are a lot of mics and calculations necessary for the diva to sound like she=

's in center stage in a stereo recording?



They would be if you were proposing to have her swinging

thousands of feet overhead at hundreds of miles per hour,

with an unknown atmosphere in between, and unknown

reflection sources on the ground, and arbitrary listener

positions. (Actually, the stereo effect in most recorded

music is generated strictly by relative amplitude

differences at the ears, without regard to timing/phase

differences. That's how recording engineers can change

perceived position so easily... they just fade one channel

up and the other down. That won't work for cancellation.)



even if you had a perfectly-known air column and no

=20

reflections. Even if you knew the exact distance, etc, etc.

=20

=20

=20

So, why attack the impossible? =20



This one is good for more than noise abatement. It may lead to other probl=

ems to solve.



You mean problems like understanding sound propagation,

wavefront interference, and random sources? <g

See the post just above this one complaining about flat stereo recordings.

They haven't given out the details but the navy now has a ship based IR laser weapon.

They must have done a lot of research on the various atmospheric situations and how to monitor those conditions where it'll work.

Best regards,







Bob Masta



DAQARTA v7.50

Data AcQuisition And Real-Time Analysis

www.daqarta.com

Scope, Spectrum, Spectrogram, Sound Level Meter

Frequency Counter, Pitch Track, Pitch-to-MIDI

FREE Signal Generator, DaqMusiq generator

Science with your sound card!

 

[Bret Cahill]

NASA once hired a symphony director to come up with jet noise reduction ideas. He wasn't successful.

Instead of the headphones which cancel all outside sound navy jets could deploy a microphone under the jet just after takeoff and then radio the signal back to troops on the ground who wanted to attenuate the jet sound but not those trying to talk or hear other sounds. A dedicated receiver headphone would be necessary.



The signal could also be sent to the long suffering civilian residents of San Diego and Va. Beach.





Bret Cahill

 

[amdx]

On 1/17/2014 10:28 PM, Bret Cahill wrote:

The happy fact is we can know _everything_ that is going to happen to that noise wave over Seatack or Pt. Loma.

You don't know the phase angle where I'm standing.
No I'm over here now.
Mikek

 

[Bret Cahill]

The happy fact is we can know _everything_ that is going to happen to that noise wave over Seatack or Pt. Loma.



You don't know the phase angle where I'm standing.

I have a plant who works at NSA. Every time you blow your nose it appears on youtube in real time.

Lot's of advertising $.

 

[John Fields]

On Wed, 22 Jan 2014 20:26:47 -0800 (PST), Bret Cahill
<bret.e.cahill@gmail.com> wrote:

The happy fact is we can know _everything_ that is going to happen to that noise wave over Seatack or Pt. Loma.



You don't know the phase angle where I'm standing.

I have a plant who works at NSA.

---
Well, it looks like you've finally taken enough rope...

Congratulations!