"kony" wrote in message
...
On Sun, 15 Oct 2006 09:32:39 -0500, "Ken Maltby"
wrote:


You made a suggestion that was not resolvable to a
difference in operation of an MP3 player. With a constant
current and constant bitrate output, you'd essentially be
suggesting that from a distance you can discriminate which
bits are flowing on the bus to the memory, in what is likely
a shielded case. I find this highly unlikely.


I was suggesting no such thing. I find your idea that an
ungrounded MP3 recorder has any significant shielding,
very unlikely.

Define significant. Many have grounded copper foil in them.
It's not as though this is a high powered device to begin
with, though, and would commonly have to be detected at a
distance.

Still consumer electronics do not have very good shielding.
Hence it would be a very minor task to detect the sampling clock of the
recorder in question. And most of the times the sampling rate is specified
by the MFG.


The recorder to be a threat and to respond
to sound must let sound waves through, even if it is a
contact microphone/sensor/transducer, and they require
significant amplification in their operation.

No, you are thinking of older devices. There needs be no
amplification prior to the digitization chip which can run
at constant current, very low voltage and no easily
detectable response to room noise from a distance.

You still have the sampling rate, which requires a clock at that rate, so at
a minimum that clock can be detected.
And most designs would include an amplification stage prior to digitization,
as the levels from most mics will not be sufficient, and also to add
isolation between the input stages.

On Sun, 15 Oct 2006 19:21:08 -0800, "Dana"
wrote:


"kony" wrote in message
.. .
On Sun, 15 Oct 2006 09:32:39 -0500, "Ken Maltby"
wrote:


You made a suggestion that was not resolvable to a
difference in operation of an MP3 player. With a constant
current and constant bitrate output, you'd essentially be
suggesting that from a distance you can discriminate which
bits are flowing on the bus to the memory, in what is likely
a shielded case. I find this highly unlikely.


I was suggesting no such thing. I find your idea that an
ungrounded MP3 recorder has any significant shielding,
very unlikely.

Define significant. Many have grounded copper foil in them.
It's not as though this is a high powered device to begin
with, though, and would commonly have to be detected at a
distance.

Still consumer electronics do not have very good shielding.

Doesn't have to be *very good*, only has to further reduce
emissions which likely weren't at a level high enough to
discriminate recording mode even without the shield.

Hence it would be a very minor task to detect the sampling clock of the
recorder in question.

That does not indicate it is an MP3 player, let alone
recording. There is no one "sample clock" common to all MP3
players.

most of the times the sampling rate is specified
by the MFG.


Manufacturer of the chip, yes, not the MP3 player, and
"spec" really means, hardware support as it can't be
selected at random like with most computers running soft
codecs. Even so, this rate is not usually a separate
oscillator, the chip itself has a clock that can also vary
per chip. It is certianly not something that remains
constant over all MP3 players, and not a signal that appears
only when set to recording mode.



The recorder to be a threat and to respond
to sound must let sound waves through, even if it is a
contact microphone/sensor/transducer, and they require
significant amplification in their operation.

No, you are thinking of older devices. There needs be no
amplification prior to the digitization chip which can run
at constant current, very low voltage and no easily
detectable response to room noise from a distance.

You still have the sampling rate, which requires a clock at that rate,

No, it does not. Clock rates are divisible or multiplied
these days, and these rates are often common to process
sizes, or current targets, not a specific functional
requirement. In other words, it's a safe bet you cannot
detect a recording MP3 player with a universal "sampling
rate" detection scheme, even before considering they won't
all necessaril record at the same rate, further lacking
consideration for any possiblity of variable rate or spread
spectrum.

so at
a minimum that clock can be detected.
And most designs would include an amplification stage prior to digitization,
as the levels from most mics will not be sufficient,

Sufficient for hearing through earbuds, no, that'd be amp'd.

Sufficient for a microchip DESIGNED to use a mic input to
digitize MP3? It would be an incredibly poorly designed
chip if it had to have a preamp tacked on after the mic.

and also to add
isolation between the input stages.

You are thinking old-school multi-stage, possibly even
discrete audio designs. All-integrated single chip MP3
players (recording) isn't directly applicable.

"kony" wrote in message
...
On Sun, 15 Oct 2006 19:21:08 -0800, "Dana"
wrote:


"kony" wrote in message
.. .
On Sun, 15 Oct 2006 09:32:39 -0500, "Ken Maltby"
wrote:


You made a suggestion that was not resolvable to a
difference in operation of an MP3 player. With a constant
current and constant bitrate output, you'd essentially be
suggesting that from a distance you can discriminate which
bits are flowing on the bus to the memory, in what is likely
a shielded case. I find this highly unlikely.


I was suggesting no such thing. I find your idea that an
ungrounded MP3 recorder has any significant shielding,
very unlikely.

Define significant. Many have grounded copper foil in them.
It's not as though this is a high powered device to begin
with, though, and would commonly have to be detected at a
distance.

Still consumer electronics do not have very good shielding.

Doesn't have to be *very good*, only has to further reduce
emissions which likely weren't at a level high enough to
discriminate recording mode even without the shield.

And most consumer electronics are not very well shielded, hence it is a snap
to pick up their emissions with off the shelf test equipment.


Hence it would be a very minor task to detect the sampling clock of the
recorder in question.

That does not indicate it is an MP3 player,

So what. It still indicates the presence of a device that can record the
persons converstaion, and that is what is required. It can be a dictation
device some other kind of recorder, it would still be detected.


most of the times the sampling rate is specified
by the MFG.


Manufacturer of the chip, yes, not the MP3 player

All you need is the chip, and usually the OEM will list what the chip MFG
states anyway.

and
"spec" really means, hardware support as it can't be
selected at random like with most computers running soft
codecs. Even so, this rate is not usually a separate
oscillator,

Usually you have an external clock needed to feed the codec. That clock can
be detected as well.




The recorder to be a threat and to respond
to sound must let sound waves through, even if it is a
contact microphone/sensor/transducer, and they require
significant amplification in their operation.

No, you are thinking of older devices. There needs be no
amplification prior to the digitization chip which can run
at constant current, very low voltage and no easily
detectable response to room noise from a distance.

You still have the sampling rate, which requires a clock at that rate,

No, it does not.

Without a sampling rate, there will be no conversion of analog to digital.
You have to take so many samples of the analog signal.

so at
a minimum that clock can be detected.
And most designs would include an amplification stage prior to
digitization,
as the levels from most mics will not be sufficient,

and also to add
isolation between the input stages.

On Sun, 15 Oct 2006 20:31:16 -0800, "Dana"
wrote:


Doesn't have to be *very good*, only has to further reduce
emissions which likely weren't at a level high enough to
discriminate recording mode even without the shield.

And most consumer electronics are not very well shielded, hence it is a snap
to pick up their emissions with off the shelf test equipment.


"Most" don't have any shield at all. MP3 players, commonly
do. Further, "most" consumer devices have an order or two
of magnitude, more active parts in them and use far more
power, stronger emissions.

Further, detecting a very faint signal is not the same thing
as having a strong enough detection and valid discrimination
method between recording MP3 players and all other consumer
electronics. Remember that we are not just trying to detect
that some "thing" using electricity is present, it has to be
identifed in function and is not just one device buy a
multitude of different MP3 player (or other digital
recorders too if you want to consider all types) recorders.

You will have to find a specific commonality, not just a
vague generalization, to discriminate them. Even this much
is premature- that commonality would have to exist which has
not in itself been established.




Hence it would be a very minor task to detect the sampling clock of the
recorder in question.

That does not indicate it is an MP3 player,

So what. It still indicates the presence of a device that can record the
persons converstaion,

No it does not. Did you think nothing but MP3 players have
clocks, or that all MP3 players have the same clock rate?
Neither is true.

and that is what is required. It can be a dictation
device some other kind of recorder, it would still be detected.

No, in some cases you might detect some devices, but it'd be
random, you'd far more often detect non-recording or devices
completely incapable of recording and wouldn't detect some
actually recording. In other words, random and useless.




most of the times the sampling rate is specified
by the MFG.


Manufacturer of the chip, yes, not the MP3 player

All you need is the chip, and usually the OEM will list what the chip MFG
states anyway.

You'll need ALL of the chips in existence, and you'd find
some are not putting out enough noise to be detected in a
typical scenario. Maybe if you put a scanner up against the
device. Is that really useful? If you had the device out
already, no further scanning is needed at all unless you
have far-fetched idea like if the MP3 recorder were built
into a shoe-heel or a clock, etc. Even then, it's a matter
of scenario. If that scenario doesn't allow getting the
scanner close enough to find the shoe is a source, you'll
never even know it was suspicious there was a noisey shoe.

I've gone off on a tangent though, for our purposes an MP3
player should be considered what is bought off the shelf.
OEMs do not "list what the chip MFG states". Most often you
have to tear open the specific player and examine it
yourself, or rely on reports from someone else who has.



and
"spec" really means, hardware support as it can't be
selected at random like with most computers running soft
codecs. Even so, this rate is not usually a separate
oscillator,

Usually you have an external clock needed to feed the codec. That clock can
be detected as well.

Again you are thinking of older electronics, today's
player/recorders are highly integrated. That doesn't mean
ALL devices will have a different or undetectable, or
indistinuishable clock signal, but it does mean you don't
have a commonality that allows detection as an MP3 player,
let alone one recording.


No, it does not.

Without a sampling rate, there will be no conversion of analog to digital.

The existence of a sampling rate does not suggest it is
always the same rate nor that it is measureable in any
particular scenario.

You have to take so many samples of the analog signal.

Yes, but this does not lead to any of the other conclusions.

It all boils down to a weapon vs armor arms race.

Suppose one vendor produced a device that can detect some device which
samples at 8 kHz. More likely it will detect frequencies that are multiple
of 8 KHz. Then, an MP3 recorder doesn't have to use any external xtal
frequency which is n*8000 - an on-chip PLL is commonplace. A mic is
connected directly to the chip, so it won't give any EMI.

Most reliable signature would be periodic access to the flash serial
interface, though.

Still, if someone wants to record a conversation, undetected, a custom
shield may be manufactured for the recorder (like 1 mm of permalloy/copper
sandwich), and bingo: no detection.

"kony" wrote in message
...
On Sun, 15 Oct 2006 20:31:16 -0800, "Dana"
wrote:


Doesn't have to be *very good*, only has to further reduce
emissions which likely weren't at a level high enough to
discriminate recording mode even without the shield.

And most consumer electronics are not very well shielded, hence it is a
snap
to pick up their emissions with off the shelf test equipment.


"Most" don't have any shield at all. MP3 players, commonly
do. Further, "most" consumer devices have an order or two
of magnitude, more active parts in them and use far more
power, stronger emissions.

Further, detecting a very faint signal is not the same thing
as having a strong enough detection and valid discrimination
method between recording MP3 players and all other consumer
electronics. Remember that we are not just trying to detect
that some "thing" using electricity is present, it has to be
identifed in function and is not just one device buy a
multitude of different MP3 player (or other digital
recorders too if you want to consider all types) recorders.

You will have to find a specific commonality, not just a
vague generalization, to discriminate them. Even this much
is premature- that commonality would have to exist which has
not in itself been established.




Hence it would be a very minor task to detect the sampling clock of the
recorder in question.

That does not indicate it is an MP3 player,

So what. It still indicates the presence of a device that can record the
persons converstaion,

No it does not. Did you think nothing but MP3 players have
clocks, or that all MP3 players have the same clock rate?
Neither is true.

and that is what is required. It can be a dictation
device some other kind of recorder, it would still be detected.

No, in some cases you might detect some devices, but it'd be
random, you'd far more often detect non-recording or devices
completely incapable of recording and wouldn't detect some
actually recording. In other words, random and useless.




most of the times the sampling rate is specified
by the MFG.


Manufacturer of the chip, yes, not the MP3 player

All you need is the chip, and usually the OEM will list what the chip MFG
states anyway.

You'll need ALL of the chips in existence, and you'd find
some are not putting out enough noise to be detected in a
typical scenario. Maybe if you put a scanner up against the
device. Is that really useful? If you had the device out
already, no further scanning is needed at all unless you
have far-fetched idea like if the MP3 recorder were built
into a shoe-heel or a clock, etc. Even then, it's a matter
of scenario. If that scenario doesn't allow getting the
scanner close enough to find the shoe is a source, you'll
never even know it was suspicious there was a noisey shoe.

I've gone off on a tangent though, for our purposes an MP3
player should be considered what is bought off the shelf.
OEMs do not "list what the chip MFG states". Most often you
have to tear open the specific player and examine it
yourself, or rely on reports from someone else who has.



and
"spec" really means, hardware support as it can't be
selected at random like with most computers running soft
codecs. Even so, this rate is not usually a separate
oscillator,

Usually you have an external clock needed to feed the codec. That clock
can
be detected as well.

Again you are thinking of older electronics, today's
player/recorders are highly integrated. That doesn't mean
ALL devices will have a different or undetectable, or
indistinuishable clock signal, but it does mean you don't
have a commonality that allows detection as an MP3 player,
let alone one recording.


No, it does not.

Without a sampling rate, there will be no conversion of analog to digital.

The existence of a sampling rate does not suggest it is
always the same rate nor that it is measureable in any
particular scenario.

You have to take so many samples of the analog signal.

Yes, but this does not lead to any of the other conclusions.

"Alexander Grigoriev" wrote in message
ink.net...
It all boils down to a weapon vs armor arms race.

Suppose one vendor produced a device that can detect some device which
samples at 8 kHz. More likely it will detect frequencies that are multiple
of 8 KHz. Then, an MP3 recorder doesn't have to use any external xtal
frequency which is n*8000 - an on-chip PLL is commonplace. A mic is
connected directly to the chip, so it won't give any EMI.

Most reliable signature would be periodic access to the flash serial
interface, though.

Still, if someone wants to record a conversation, undetected, a custom
shield may be manufactured for the recorder (like 1 mm of permalloy/copper
sandwich), and bingo: no detection.


Right it's a secret part of an "arms race", but the same
lab(s) are developing surveillance equipment and counter-
surveillance equipment. With unlimited funding, you can
have your "Undetectable Device" and you can have a
Device to detect the undetectable. It becomes what you
can have at what cost and in what numbers. The more
expensive and rare devices are reserved for the most
sensitive and vital situations.

It is in the interest of those working the counter-surveillance
side, that those thinking of using a surveillance device, not
know the likelihood of their being an effective detection
device in play or not. Things electronic get cheaper all the
time, last year's rare laboratory sensor may well be in next
year's field unit.

Luck;
Ken

Aly wrote:

Joey wrote in message
...
Suppose someone visited your office or home and tried to make a voice
recording using a hidden recorder.


Your only real option is to hold your meetings in the middle of field, and
for everyone to be naked.

Don't forget the full cavity search before and after. 8)

In comp.sys.ibm.pc.hardware.storage chrisv wrote:
Aly wrote:

Joey wrote in message
...
Suppose someone visited your office or home and tried to make a voice
recording using a hidden recorder.


Your only real option is to hold your meetings in the middle of field, and
for everyone to be naked.

Don't forget the full cavity search before and after. 8)

With the size these things are getting today, some Xt-ray imaging will
also be necessary....

Arno

On Tue, 17 Oct 2006 10:19:33 -0500, chrisv
wrote:

Aly wrote:

Joey wrote in message
...
Suppose someone visited your office or home and tried to make a voice
recording using a hidden recorder.


Your only real option is to hold your meetings in the middle of field, and
for everyone to be naked.

Don't forget the full cavity search before and after. 8)

.... and the two layers of sound insulating walls so
directional mics can't pick anything up.

In comp.sys.ibm.pc.hardware.storage wrote:
On 17 Oct 2006 17:35:42 GMT, Arno Wagner wrote:

In comp.sys.ibm.pc.hardware.storage chrisv wrote:
Aly wrote:

Joey wrote in message
...
Suppose someone visited your office or home and tried to make a voice
recording using a hidden recorder.


Your only real option is to hold your meetings in the middle of field, and
for everyone to be naked.

Don't forget the full cavity search before and after. 8)

With the size these things are getting today, some Xt-ray imaging will
also be necessary....

Arno
ah that puts me in mind of incendent for the polit of B% where Sinclar
infrom G'Kar that he just shawlowed a naotech tracer , when in fact no
tracer existed , and then he and garbaldi coment on what his own
people were goignt o do to G'Kar looking for the tracer that isn't
there

Hehehe. That was fun!

Arno