NEW PRIVACY MENACE: Your Touch Screen Display is NOW a MICROPHONE!

originally posted by: AdmireTheDistance

originally posted by: StargateSG7
8) Custom HACKED software takes that UNFILTERED
electrical noise and digitally converts it to weak but
discernable impulses of numeric values that represent
amplitude and frequency too!

9) That series of numbers is compared to a secret onboard
database of human voice samples of COMMON phonemes
(male and female) ad pronunications of common vowels
and consonants.

Oh really? I reverse engineer and modify proprietary mobile device firmware (and write custom firmware/kernels) for a living, so where does this "secret onboard database" reside. Bear in mind, I have full access to the entire partition table on modern devices, including bootloaders and radios, so I want specifics; Which block(s) and in which partition(s)?

10) the VOCODER then re-creates a mimicry of what
the voice PROBABLY sounded like in it's original form.
That VOCODER data is then converted to easily compressible
text which is then transferred to VARIOUS locations using
common network tarnsmission techniques (TCP/IP or USB)
OR stored secretly on the connected computing system
for LATER PHYSICAL retrieval!

And how do such relatively resource intensive processes manage to completely mask themselves from the CPU's raw output and from the mobile radio?

---

===

For which TYPE of connected device?

If it's Windows Mobile, probably any part
of the NTFS-like file system that isn't checked
off in the bit-oriented block allocation table
usually placed in the middle of the physical
sectors-space of a storage system) which
can allocate 512 bytes, 1024 bytes, 2048 bytes,
4k up to 64k blocks at a time. Just place the data
NEAREST or FARTHEST away from the bit-oriented
block allocation tables which tends to use a
buiilt-in algorithm to place user files away from it.

With Android being Linux-based and probably
EXT3/4 file system based can store in any sector
(512 bytes in a physical sector) and any free unused
area within a logical 4k block. With smaller files
(i.e. log files) less than the logical block size on
BOTH ETX3/4 and NTFS you can used the free space
AFTER the end-of-file. i.e. if one kb is used
in a logical 4k block then spread your database
amongst the free space in both partially filled
logical blocks at the end of files OR set the
the CRC32 checksums or any digital signatures
to reflect ADDITIONS to the END of normally
unchanging system *.DAT files or common
little-changed data storage files.

---

On Windows PC's NTFS the 63 physical sectors
after the first one ON ANY PHYSICAL DRIVE
which stores older-style 4-partition location
tables or on the newer UEFI tables not typically
used in any meaningful way can be set to store a boot sector
loader and jump table that indicates DAMAGED sectors which
will be locked out by the OS. The threshold is usually
0.5% to 1.0% for a damaged sector list to be acted
upon by the OS code as being considered a damaged
and unusable hard drive. For larger disks, 1% is a very
large amount of free space to store a hidden database!

For SSD Flash drives and even USB sticks there is a
hidden micro-OS and sector table which can lock out
sectors from any upper-level partition management
software BUT can be accessed directly by a kernel
level driver. On a 16 gigabyte flash drive I could
have up to 500 megs freed up before the OS
hiccups -- lotsa room to store database!
===

The mobile radio (BASEBAND SDR) would almost
never do this type of work but if you make some
custom microcode you COULD INCLUDE DSP microcode
into the software-defined radio functions. These new
baseband processors are sometimes very powerful
32 bits wide with some pretty hefty caches to do
DSP work in!

That runs BELOW the main operating system
of the phone and is undetectable at the upper
level Android, iOS, WIndows Mobile OS CPU
monitoring systems since the information
exchange between the BASEBAND processor
and the upper level ARM-based or ATOM-based
CPU is fully asynchronous!

I've been able to even get the NETWORK processor
on my network card to do DSP --- Those have more
horsepower power than any old 286 or 386 or MC68000 cpu!
That LOSTSA power available just on the networking chip!

a reply to: StargateSG7

Why oh why then, is there no app that makes your screen a mic?

originally posted by: combatmaster
a reply to: StargateSG7

Why oh why then, is there no app that makes your screen a mic?

Kinda depends on your screen...if it's not a touch screen
you're outa luck UNLESS someone used a nearby smartphone
or OTHER computing system with access to a sound card
and SOMEHOW was able to create ultra high frequency single tones
to bounce off of the screen surface to see how much it vibrated
from the LOWER FREQUENCIES of the human voice, you could
obtain a useful reconstruction of a waveform.

It would depend upon the QUALITY of the tweeter
(i.e. an audio speaker that reproduces sound at 13 KHZ and above!)
if it can produce sound out the range of human hearing AND the
receiving microphone if it has enough frequency response to
obtain and differentiate the micro-vibrations caused during
acoustic reflections off hard and soft surfaces.

Reflectometry is the technical term for this!

So basically you use HIGH FREQUENCY SOUND (13 KHZ+) to
obtain and digitize the reflected waveforms of the
300 HZ to 3400 KHZ range of human speech that
bounce off as VIBRATIONS from the hard display surface.
Again you're NOT sampling the actual speech, you are
sampling the VIBRATIONS within the display surface
which is THEN reconstructed as a digital waveform
which is THEN converted back into audible data or
converted to ASCII or UNICODE text using
speech-to-text technology.

THAT IS VERY DIFFICULT digital signal processing to do!
There are hardware systems in the $100,000+ range used
for monitoring and researching engine vibration and
acoustic noise signatures that use this type of
reflectometry technology.

GM, FORD, BMW, Mercedes, Toyota all use
this tech to make low-noise engines!

It would be time-consuming to do a consumer
APP but like I said earlier this is work a GOVERNMENT
would do for economic espionage work and almost
ALL the major powers do this!

I know this sounds like a rather roundabout way
of recording voices, but sometimes you need
something MORE HIDDEN from normal anti-espionage
technology and this would be one way to do it!

originally posted by: StargateSG7
For which TYPE of connected device?

Let's go with Android, just cause it's easier to dump and reverse everything. Again, my question is: Where exactly is your "secret onboard database" located, and where are the (presumably numerous) calls to it (and the data passed with those calls)?

Hell, where is the "custom hacked software" that encodes the information in the first place? Such a thing would be far too large to be contained in the low-level display drivers. I know for a fact that nothing like that resides at the kernel level, since I've written large parts of, modified, and built numerous kernels. I also know for a fact that no such code exists in the frameworks, libcore, or any of the native libraries, and to suggest that such a thing occurs at a higher level than that is just laughable. So where is it?

originally posted by: AdmireTheDistance

originally posted by: StargateSG7
For which TYPE of connected device?

Let's go with Android, just cause it's easier to dump and reverse everything. Again, my question is: Where exactly is your "secret onboard database" located, and where are the (presumably numerous) calls to it (and the data passed with those calls)?

Hell, where is the "custom hacked software" that encodes the information in the first place? Such a thing would be far too large to be contained in the low-level display drivers. I know for a fact that nothing like that resides at the kernel level, since I've written large parts of, modified, and built numerous kernels. I also know for a fact that no such code exists in the frameworks, libcore, or any of the native libraries, and to suggest that such a thing occurs at a higher level than that is just laughable. So where is it?

===

"...So where is it? ..."

I've got no idea where such code would be since
in any public repository or Linux distribution it
WOULD NOT BE EXPOSED there...I am suggesting
that it would be uploaded secretly since HOW
many people DO YOU KNOW would disassemble
a linux/android/windows mobile/iOS distro!

They wouldn't and ANY powerful corporation or
government could make their own distro and
FAKE ITS ORIGIN!

".... Such a thing would be far too large to
be contained in the low-level display drivers...."

DEFINE LARGE! AMD supplies 1 gigabyte worth
of drivers for their cards (on desktops) and even
many kernels have over 300k in display driver code!
Depends upon the skill of the microcode writer
since the DSP microcode itself is small but the
phoneme and waveform DATABASE is NOT!

How fast they can access and compare an array of
phonemes is what will be the deciding factor.
Microcode writers would usually store as much
as they can in L1, L2, L3 caches as possible.

I know many L1/L2 caches are 4 kilobytes or less
in size on many ARM chips but L3 cache can be as
large as MEGABYTES in size on certain smartphone
processors which is enough for fast database comparisons.

with SSD flash drives they are so fast,
that storing and accessing the phoneme
database completely from SSD is now possible!

originally posted by: StargateSG7
I've got no idea where such code would be since
in any public repository or Linux distribution it
WOULD NOT BE EXPOSED there...I am suggesting
that it would be uploaded secretly...

As I've already said, the idea of something such as what you're talking about running as a high-level process is laughable, at best, and any lower-level code, even if "uploaded secretly" somehow, wouldn't do anything, as it would require recompiling the entire operating system with the "secret" code in place.

...since HOW many people DO YOU KNOW would disassemble a linux/android/windows mobile/iOS distro!

Several, including myself.

Hey bedlam,

If you wanted to thwart laser mics on the old window trick would taping a speaker cone to a window pane and playing white noise be sufficient. Or would there be a software code that filters through the white noise. Like they can do with phones and bugs. If the volume on the speaker is high enough do you think it would work if your a poor dude on a budget like me. I got 4 10 inch subwoofers and 400 watts I can use. Would it work in your opinion.

originally posted by: AdmireTheDistance

originally posted by: StargateSG7
I've got no idea where such code would be since
in any public repository or Linux distribution it
WOULD NOT BE EXPOSED there...I am suggesting
that it would be uploaded secretly...

As I've already said, the idea of something such as what you're talking about running as a high-level process is laughable, at best, and any lower-level code, even if "uploaded secretly" somehow, wouldn't do anything, as it would require recompiling the entire operating system with the "secret" code in place.

...since HOW many people DO YOU KNOW would disassemble a linux/android/windows mobile/iOS distro!

Several, including myself.

===
Other than someone like me and you, very FEW would
disassemble an OS and debug it. Consumers just want something
that works and if your smartphone is intercepted during shipment
and some chips and microcode swapped out by a large organization,
(i.e. NSA/CIA/DIA/MI6/CSE/GRU, etc) you'd never know
as a consumer what was happening.

From a technical point of view MANY AGENCIES recompile code
for their own purposes. I will use the example of the Aircraft Carrier
USS Ronald Reagan (CVN-76) which used (uses?) a re-compiled version
of Windows 2000 (YES! Windows 2000) for many of its on-board systems.

They put extra cryptographic functions into it and
some ship systems hardware-specific code in it!
In fact, I think some parts of (CVN-76) STILL run Win2000!

So it is NOT INCONCEIVABLE that agencies compile their
OWN OS'es for whatever purposes they desire and since
android is almost fully open source, it would be a piece
of cake to make a new version with DEEPLY HIDDEN
functionality.

In any case the system I am espousing would HARDLY be
regarded as a high-level task. It would be hidden deeply
and darkly into the Linux OS code / Windows System.
It simply WOULD NOT SHOW UP in any task manager
or CPU logging function.

You would have to do a hardware debug to find out
what instructions are being fed to which chips.
On smartphones if it was BASEBAND Radio Modem
microcode, Android/iOS/WinMobile would NEVER
SEE the code AT ALL since baseband OS'es operate
SEPARATE FROM AND BELOW the primary operating
system in an asynchronous manner!

a reply to: BASSPLYR

Just turn on a radio to static and turn it up loud enough to drown out the conversation you're having. Or, record yourself talking to someone else about mundane things, play it on loop. It would confuse voice recognition software. Better would be to do both at the same time.

Gotta think with yer noggin' ...

originally posted by: StargateSG7


See where I am getting at?

1) You are speaking near a capacitive touchscreen.

2) The pressure waves produced by your vocal chords
follow a specific pattern of rise-times and fall times.

3) Those impulses HIT the screen bouncing back off
the front-side acoustically reflective surface.

4) Those bouncing waveforms disturb the microdust
and water vapour within the air close to the screens
surface.

Sound waves don't cause net motion of aerosols. The energy in speech audio is insanely low.

5) That moving dust and water vapour (i.e. humidity)
changes the CONDUCTANCE of that 3D-XYZ space.
That change in conductance INDUCES a spurious
introduction of ELECTRICAL NOISE into the
capacitive touch screen circuits.

It does not. Capacitive touch screens depend on the relatively large change in mutual coupling between drive and sense lines in the screen caused by the presence of a finger/stylus. You do not get a measurable change from sound. Hint: go look up a table of the change of dielectric constant vs air pressure. Now, find out how much air pressure change is associated with conversational level audio. Insanely small multiplied by insanely small = forget it. This is what you claim is being sensed with this method. Ain't happening.

6) That noise travels to the ADAC (analog to digital
converter software/hardware) which NORMALLY filters
it out because it is UNWANTED ELECTRICAL NOISE.

The front end of this system will filter it out. It can't be programmed away. And the noise is far far far far below a single bit change. Below the inherent noise in the ADC. Below the Johnson noise in the resistors in the front end. Below the transistor noise in the preamps. Below the switching noise that's caused by muxing the sense and drive lines. You might as well try to use the thing to detect electrical noise caused by the background RF of the universe.

7) That unwanted electrical noise follows the
SAME PATTERN as the acoustic waveform that
actually hit the screen as pressure waves.

The unwanted electrical noise is less than the other noise in the system by a few dozen orders of magnitude, and is lost.

8) Custom HACKED software takes that UNFILTERED
electrical noise and digitally converts it to weak but
discernable impulses of numeric values that represent
amplitude and frequency too!

All the King's hackers and all the King's software geeks can't put together something that isn't there.

9) That series of numbers is compared to a secret onboard
database of human voice samples of COMMON phonemes
(male and female) ad pronunications of common vowels
and consonants.

Right. Stored where? Compared by what?

SO! CAN YOU READ THE ABOVE AND DO YOU UNDERSTAND IT?

If you can, DO YOU AGREE with the BASE engineering
and mathematics of such technology! If you CANNOT, then
I can't help you and you need to consult an expert level MScEE
who KNOWS something about Digital Signal Processing,
vocoder technology, and noise filtering.

Raises hand. Got one right here. And it's just wrong.

You left out point 10: Or, you know, just use the microphone that's on the phone. Problem solved.

originally posted by: BASSPLYR
Hey bedlam,

If you wanted to thwart laser mics on the old window trick would taping a speaker cone to a window pane and playing white noise be sufficient. Or would there be a software code that filters through the white noise. Like they can do with phones and bugs. If the volume on the speaker is high enough do you think it would work if your a poor dude on a budget like me. I got 4 10 inch subwoofers and 400 watts I can use. Would it work in your opinion.

Sure. It doesn't take much. Just picking it up with the AC going is tough. Also you have a lot of issues with noise coming from OUTSIDE, like traffic noise.

a reply to: Bedlam

Your additional point 10 has been pointed out before. Look back and he says that Occam's razor isn't always the case because (pick any 3 letter agency) wants to make it more secret and have money to burn.

:eye roll

originally posted by: TerryDon79
a reply to: Bedlam

Your additional point 10 has been pointed out before. Look back and he says that Occam's razor isn't always the case because (pick any 3 letter agency) wants to make it more secret and have money to burn.

:eye roll

A thing he doesn't want to consider, apparently, is that there's a difference between "Can you do such a thing once, for a specialized reason, using insane amounts of money thrown at the problem" which is "probably", vs "Can you then do this en masse for production of a price-driven commodity, without affecting the price at all, for no particular reason", which is a resounding "hell no".

eta: I've piddled around with something like this in the past, the goal being "can you design a sound pickup that doesn't involve a traditional membrane plus a strain gauge or moving coil/magnet or the like", possibly by Schlieren optics or direct sensing of air density based on capacitance changes, and it's insanely difficult. I wanted the World's Best Microphone that you could mass produce. The reason you don't see one already is, it's monumentally tough.

a reply to: Bedlam

Exactly.

I made comment earlier in this thread to stuxnet.
Let's call the original stuxnet v1.

After it has been used (and subsequently detected) it is no longer a viable attack as there's counter measures for it.

It then gets changed so much it is now something new.
Let's call it stuxnet v2.

After v2 was used it is no longer viable.

And so on and so on.

It would be (in theory) the same for this hardware software combo. It is the most unreliable way of getting intelligence (especially from air locked systems as they can't broadcast anything).

originally posted by: StargateSG7

originally posted by: AdmireTheDistance

originally posted by: StargateSG7
I've got no idea where such code would be since
in any public repository or Linux distribution it
WOULD NOT BE EXPOSED there...I am suggesting
that it would be uploaded secretly...

As I've already said, the idea of something such as what you're talking about running as a high-level process is laughable, at best, and any lower-level code, even if "uploaded secretly" somehow, wouldn't do anything, as it would require recompiling the entire operating system with the "secret" code in place.

...since HOW many people DO YOU KNOW would disassemble a linux/android/windows mobile/iOS distro!

Several, including myself.

===
Other than someone like me and you, very FEW would
disassemble an OS and debug it. Consumers just want something
that works and if your smartphone is intercepted during shipment
and some chips and microcode swapped out by a large organization,
(i.e. NSA/CIA/DIA/MI6/CSE/GRU, etc) you'd never know
as a consumer what was happening.

Perhaps. I am not an average consumer, however. This does nothing to answer my question.

From a technical point of view MANY AGENCIES recompile code
for their own purposes. I will use the example of the Aircraft Carrier
USS Ronald Reagan (CVN-76) which used (uses?) a re-compiled version
of Windows 2000 (YES! Windows 2000) for many of its on-board systems.

They put extra cryptographic functions into it and
some ship systems hardware-specific code in it!
In fact, I think some parts of (CVN-76) STILL run Win2000!

I know nothing of the U.S.S. Ronald Reagan or what operating system it's on-board systems are built on, nor do I care. This is completely irrelevant and again, does nothing to answer my question.

So it is NOT INCONCEIVABLE that agencies compile their
OWN OS'es for whatever purposes they desire and since
android is almost fully open source, it would be a piece
of cake to make a new version with DEEPLY HIDDEN
functionality.

Of course various government entities use proprietary operating system systems. To suggest otherwise would be daft. I also believe I stated that code such as you're talking about would be "DEEPLY HIDDEN" a couple posts ago. This does nothing to answer my question.

In any case the system I am espousing would HARDLY be
regarded as a high-level task. It would be hidden deeply
and darkly into the Linux OS code / Windows System.
It simply WOULD NOT SHOW UP in any task manager
or CPU logging function.

Yep. I've already stated as much, and you've already parroted it. This does nothing to answer my question.

You would have to do a hardware debug to find out
what instructions are being fed to which chips.

I have a plethora of devices at my disposal and am more than capable of doing this. This doesn't answer my question though.

On smartphones if it was BASEBAND Radio Modem
microcode, Android/iOS/WinMobile would NEVER
SEE the code AT ALL since baseband OS'es operate
SEPARATE FROM AND BELOW the primary operating
system in an asynchronous manner!

This is not entirely true, but it is, however, totally irrelevant. It also fails to answer my question.

You seem to be implying that this magical, secret code is in the modem somewhere...So for the third time: Where? And where is the code that handles the raw input data from the screen, and passes it to the magic, secret code in the modem?

I have folders containing the full source code for a handful of different devices open right now. Just waiting for you to answer my question, so I can verify whether or not you're telling the truth. Your move.

originally posted by: Bedlam

originally posted by: StargateSG7


See where I am getting at?

1) You are speaking near a capacitive touchscreen.

2) The pressure waves produced by your vocal chords
follow a specific pattern of rise-times and fall times.

3) Those impulses HIT the screen bouncing back off
the front-side acoustically reflective surface.

4) Those bouncing waveforms disturb the microdust
and water vapour within the air close to the screens
surface.

Sound waves don't cause net motion of aerosols. The energy in speech audio is insanely low.

5) That moving dust and water vapour (i.e. humidity)
changes the CONDUCTANCE of that 3D-XYZ space.
That change in conductance INDUCES a spurious
introduction of ELECTRICAL NOISE into the
capacitive touch screen circuits.

It does not. Capacitive touch screens depend on the relatively large change in mutual coupling between drive and sense lines in the screen caused by the presence of a finger/stylus. You do not get a measurable change from sound. Hint: go look up a table of the change of dielectric constant vs air pressure. Now, find out how much air pressure change is associated with conversational level audio. Insanely small multiplied by insanely small = forget it. This is what you claim is being sensed with this method. Ain't happening.

6) That noise travels to the ADAC (analog to digital
converter software/hardware) which NORMALLY filters
it out because it is UNWANTED ELECTRICAL NOISE.

The front end of this system will filter it out. It can't be programmed away. And the noise is far far far far below a single bit change. Below the inherent noise in the ADC. Below the Johnson noise in the resistors in the front end. Below the transistor noise in the preamps. Below the switching noise that's caused by muxing the sense and drive lines. You might as well try to use the thing to detect electrical noise caused by the background RF of the universe.

7) That unwanted electrical noise follows the
SAME PATTERN as the acoustic waveform that
actually hit the screen as pressure waves.

The unwanted electrical noise is less than the other noise in the system by a few dozen orders of magnitude, and is lost.

8) Custom HACKED software takes that UNFILTERED
electrical noise and digitally converts it to weak but
discernable impulses of numeric values that represent
amplitude and frequency too!

All the King's hackers and all the King's software geeks can't put together something that isn't there.

9) That series of numbers is compared to a secret onboard
database of human voice samples of COMMON phonemes
(male and female) ad pronunications of common vowels
and consonants.

Right. Stored where? Compared by what?

SO! CAN YOU READ THE ABOVE AND DO YOU UNDERSTAND IT?

If you can, DO YOU AGREE with the BASE engineering
and mathematics of such technology! If you CANNOT, then
I can't help you and you need to consult an expert level MScEE
who KNOWS something about Digital Signal Processing,
vocoder technology, and noise filtering.

Raises hand. Got one right here. And it's just wrong.

You left out point 10: Or, you know, just use the microphone that's on the phone. Problem solved.

====

Hmmmm.....on the surface it SEEMS like all your points are valid,
I must beg to differ on A FEW of them....

====

"Sound waves don't cause net motion of aerosols. The energy in speech audio is insanely low. "

This paper should quite expose the fact that speech and breath DO affect aerosol
movement...The report has to do with infectious agents but it is the VIDEO EVIDENCE
and photographic plates that reduces the absolute certainty of your assertion.
YES it is small on an absolute basis but NOT insignificant.

Report:
Qualitative Real-Time Schlieren and Shadowgraph Imaging of
Human Exhaled Airflows: An Aid to Aerosol Infection Control

See link:
www.ncbi.nlm.nih.gov...

====


"....It does not. Capacitive touch screens depend on the relatively
large change in mutual coupling between drive and sense lines in
the screen caused by the presence of a finger/stylus. You do not
get a measurable change from sound. ..."

I have to answer this in TWO PARTS:

First to offer evidencce of the atrocious air quality
and high concentration of aerosols within most
indoor air environments which was quite
shocking for me to read....

Indoor Aerosol Modeling: Basic Principles
and Practical Applications
Tareq Hussein & Markku Kulmala

See link:
blogs.helsinki.fi/airquality/files/.../Water-Air-Soil-Pollut-Focus-2008.pdf

Then here is paper that INFERS that magnetic signatures within indoor air
indicate a high metal content due to general pollution and that there are
changes in electrical properties of those indoor particulate concentrations
that are EASILY MEASUREABLE by instruments and which CAN involve changes
in localized conductance as those particulates move around.

Magnetic signature of indoor air pollution: Household dust study
Beata Górka-Kostrubiec1 / Maria Jeleńska1 / Elżbieta Król1

1Institute of Geophysics, Polish Academy of Sciences, Warsaw, Poland

(I should note that I am NOT an EE so I must defer to Bedlam on this
but the paper is INFERING that the AMOUNT of change in conductance
is NOT AS LOW as you have stated. I'm no math or EE major but if I read
this paper correctly, simply BREATHING introduces electrical property changes
in the localized environment that are SIGNIFICANT ENOUGH to cause induction
of noise into a conductive circuit.

While it seems the above paper DOES NOT PROVE my assertion,
it does NOT entirely dismiss it either. I will keep looking for another
paper that is more relevant to this discussion.

====

"...You might as well try to use the thing to detect electrical noise caused by the background RF of the universe. ..."

Depending upon the bit-depth of the converters (10/12/14 bits+), that might not be
such a bad Idea I would say....A relatively cheap radioscope it just might be... ;-) :-)

===

"....The unwanted electrical noise is less than the other noise
in the system by a few dozen orders of magnitude, and is lost...."

ON THIS I AGREE WITH YOU !!!
Our engineer had just such reservations since the bit depth
of the ADC would be the primary means to discern
patterns in noise from signal...basically he asked
how are you going to decide where is the absolute
random-value noise floor and then discern a signal
between random environmental/system noise and
induced but patterned signals below that of the
touch trigger signal level?

I had no good answer other than
to use advanced software analysis!

---

"....Right. Stored where? Compared by what?...."

I think I answered it previously in where I indicated
that someone WILLING ENOUGH to send me a rather
well-presented document espousing such a scheme
would ALSO be willing enough to make the financial
and technical effort to actually go through with it
and secretivly install the horsepower required
to do such signal analysis.

In my PERSONAL sphere, Texas Instruments
has some REALLY SMALL 600 MIPS DSP's that
could be embedded into consumer gear.

(TI C6000TM 7mW of standby power at
1.0V/25 degrees C and 420mW total power - $10 to $20 for one)

So I think it might work just fine for the Spook World !!!!

a reply to: StargateSG7

Exhaled air isn't sound.

The rest of your cites are either non sequiturs or you're making huge assumptive stretches of what they're telling you.

originally posted by: Bedlam
a reply to: StargateSG7

Exhaled air isn't sound.

The rest of your cites are either non sequiturs or you're making huge assumptive stretches of what they're telling you.

Regarding the above papers...they are a bit of a stretch
but have SOME support for my assertions that indoor air
is dirty containing particulate matter that can have it's
electrical properties change is it moves about in a room
which was caused by exhalation (AND SPEECH!) which
can then affect in some small (or large?) way the noise
characteristics of an electronic device caused by induced
or stray current.

----

...So on a technical basis, YES I am making huge assumptive
stretches as to what I READ ON A HARD-COPY DOCUMENT
given to me by a trusted source....

BUT

...I think I understood most of it (except for the long drawn out
equations --- am not a math major!) but I do must say that the
said document was professionally published and had more than
a few people with the initials MScEE at the end of them.

I DOUBT it was Top Secret since it had no such markings
BUT I do believe it to be semi-proprietary material based
upon public research.

The 10-step part I outlined above was a paraphrasing
of the approximate steps described to obtain a useable
waveform from a touch screen interface based upon
induced noise caused by localized changes in
atmospheric electrical properties affected
by incoming acoustic waveforms.

Would it help to say that one of the contributors
to the paper was from Georgia Tech! I think that
might give you a BIG CLUE as to the source!
And you can go from there..... ;-) :-)

originally posted by: StargateSG7

Regarding the above papers...they are a bit of a stretch
but have SOME support for my assertions that indoor air
is dirty containing particulate matter that can have it's
electrical properties change is it moves about in a room
which was caused by exhalation (AND SPEECH!) which
can then affect in some small (or large?) way the noise
characteristics of an electronic device caused by induced
or stray current.

Sound is a rapid series of pressure waves. You don't get net motion of all the aerosols in a room. The waves are pretty high frequency. The aerosols don't have a chance to do much in the way of vibratory motion, because they're way heavier than air molecules, and the air molecules themselves don't move very much. Because the energy in conversational level sound is insanely small.

As the differences between the high and low pressures in sound waves aren't very high, and the characteristics of air's (even dirty air) dielectric constant doesn't give it much in the way of change even over very LARGE pressure changes, and the capacitive touch screen isn't designed to detect the tiny changes of capacitance caused by environmental dielectric constant ANYWAY, you've got infinitesimal pressure changes causing infinitesimal dielectric constant changes causing infinitesimal capacitance changes. Multiplied together, you get bupkes.

No matter how many backsprings and stretches you do, that isn't going to change.

In order to make it even worse, you're trying to do it with hardware that's barely up to doing its intended function, that has to be dirt cheap, and is next to three transmitters.

And the damned thing HAS a microphone.

Besides which, if you're just looking to pick up sound from a room, and you can afford to spend a huge amount of money per system (i.e. NOT built into a cell phone), it's already done. And it doesn't take a laser, or a window.