Whilst the suggestion is clearly not serious, having pre-recorded or even repeated random noise is akin to re-using a one time pad.
Continued ‘listening’ - even long after the relevant conversation - could first be used to extract the fake noise and then from this the signal.
I'm going to put on my pop-science-fiction hat on and revise the Privacy Bulb to include a geiger–müller vibrating tube and a dollop of uranium for random vibrations. I was a fan of spy vs. spy as a child and I am full of bad ideas right now.
Yeah. You can't just produce random white sound[1] to cover the signal either, because speech has enough redundancy to still communicate over a static-y soundscape. The term here is that white noise lacks "diffusion"--analogous to the ECB vulnerability (see the picture of the encrypted Linux penguin image on this page[2]). I can't think of a good way to provide diffusion for sound--a better cover might be to provide alternate signals, i.e. a few layers of random human sounds spoken in a deepfake-d version of the voice of the person speaking.
I'm skeptical even this wouldn't have vulnerabilities however, as history has shown us that cryptography which has not been formally proven is often broken.
[1] I'm using the word "sound" here to differentiate the noise in the auditory sense from noise in the "signal versus noise" sense.
Would it be possible to use the tech in noise-cancelling headphones and put a tiny speaker that phase-cancels audio it hears to solve this issue? I'm asking as someone who knows very little about this.
Well, it's definitely possible to do that but I'll make no claim that it would actually provide security. There's definitely some bleed-through with noise canceling (i.e. the beginning of sounds don't get canceled because the noise-canceling has a "reaction time"). It's probably better than nothing, but I would guess that a sophisticated-enough attacker could get around it.
Again, I have no proof of concept here, so take with a grain of salt.
Then you can hijack the noise canceling device and obtain the original signal. Also the canceling effect is likely never perfect, so there would still remain part of it going through.
Since you mention it, I got to thinking about whether we can have a microphone equivalent to the physical covers for webcams -- that is, a physical means that ensures the device can't detect anything usable, even if an attacker was able to turn it on without you knowing.
Something like, a vibrating module that you don't really notice but which clouds out any sounds the microphone would pick up (or, to go your route, injects fake audio). You turn it on when you want to be sure the microphone isn't listening, just like you can cover the webcam.
Some offices employ white noise privacy machines, you can find them on amazon. And the Japanese have the Sound Princess for flush noise privacy at the toilet. I have yet to find a product where you don't notice that it's on though, would be a great invention if possible.
It does work, but you can have filters that can remove that noise electronically enough to get some actual voice signal. Faucets along with whispering directly in the ear would work. But a better choice is to simply write on a piece of paper instead of talking.
You mean something like this [0]? You could use this to trick your computer/phone into using the audio jack for its microphone, but since the mic-lock passes no usable signal, you'll be better protected. It doesn't otherwise incapacitate the internal microphone, though.
It would need to block the microphone itself that picks up ambient noise -- I don't think there's anything on e.g. a MacBook Air that you could plug that into and block sound.
One way you can do this without another light source is to bounce a laser beam off a window or other things that are reflective.
In fact, the same principle can be achieved with radio. The Russians once planted a device in a wooden seal they gave to a US ambassador as a gift which, when a focused radio beam was aimed at it, would reflect and oscillate that beam due to vibrations in the air.
I remember in the 1990s I was doing contracting work in SoCal and got a job at one of the big defense contractors near LAX. (Trying to remember which one... Hughes?) I lasted only one week as the building was totally secure - no windows at all. Life is waaaaaay to short to work in a hole like that.
Maybe someone with more info can fill in the details, but if I remember correctly, when Nokia changed their Silicon Valley offices to downtown Sunnyvale almost a decade ago, they were worried because Apple had offices in the adjacent building. If I remember correctly Nokia put in a bunch of security glass to prevent spying from vibrations in the windows. (There's some irony there, I know...)
Yup. It's fairly standard. For classified work the rooms and entry/exit protocols have to meet certain security requirements based on the classification level. No windows and no electronics are standard. No walls that is exposed to the exterior. I've been in room within a room setup when it was warranted.
Back in the 90's I found an old cardboard box full of radioshack electronics components and a paper schematic of how to put it together. I asked my dad what it was. He had been into an electronics hobby in the 1980s and told me that it was a amplifier to receiving laser light bounced off of a window for spy stuff. I couldn't believe it. The kit became my first significant soldering project. I got it all together and put my beloved laser pointer on a tripod and tried bouncing it in. I never did get it to the point that I could hear voices but boy could I ever hear cool vibration noises when I tapped the window.
It was extremely fun and I've been soldering things here and there ever since.
Thanks! It was a double-pane sliding glass door actually so maybe that did have something to do with it. Though, I also tried from the same side as the laser for most of my trials (i.e. laser and voice both coming from the inside).
Vendor qualification isn’t easy but it’s also not especially difficult as long as you’re an American citizen with no criminal history. Figuring out the paperwork is half the battle.
Past that, the hoops you'll need to jump through depend a lot on which department is handling procurement. If you look at defense-related contracts, you'll also have to go through personal clearance (which can be difficult by itself at higher levels).
When I worked at Boeing, we even had them on the windows of offices that weren't being used for government projects. The company's commercial business was also a target for spying so they were very protective in general.
To this day, the company's laptops don't have cameras and they have a whole process that you have to go through to use a camera anywhere on company property.
Basically drowns out the room vibrations upon the window as it is a in direct contact. You can make your own with piezo discs and small circuit to drive them. That and high frequency will not cause added noise noticeable to humans.
In any building spies had access to there was a game of cat and mouse as (for example) spies would see people with radio equipment scouring rooms through the window as they heard their bugs go pop one by one.
Peter Wright (who wrote spycatcher) was the one who first got to look at The Thing; He actually broke it and had to make a new membrane for the microphone. He also invented (within MI5 at least) their technique for detecting superhet receivers (RAFTER), at one point they apparently were parked next to the soviet receiver.
I've always wondered -- if I used this on the board room window to listen into public company earnings meetings, is that inside information? Technically, the information is "public" in that anyone could do what I'm doing...
I'm sure that one way or the other, that question has been dealt with exhaustively. My guess, the information you obtain is considered definitely not public, even though it's possible for a member of the public to obtain.
Along similar lines, I've wondered about how often restaurant and bar owners have bugged their own premises.
There's probably a lot of valuable information discussed in such establishments.. especially, say, around Wall St.
Regarding the legal aspects, I'm not sure how much of a legal expectation of privacy one has in a "public" place such as a restaurant or bar, and we've pretty much already accepted ubiquitous video surveillance in such places (even if such videos are often without sound).
Someone could potentially infer speech just by lipreading soundless videos, but if someone did decide to use audio surveillance on their own property in a bar, restaurant, or other place of business, is there any law against that?
Quite apart from the legal and ethical objections that might arise against doing this, I'm sure some business owners are not above giving in to the temptation to spy on their patrons, especially if there's a big profit motive (like getting access to inside information by spying on Wall St execs gossiping over drinks or dinner).
That's not to mention dirt that might be revealed in conversations considered to be "private", which could be used to blackmail people or for other nefarious purposes.
The technology to perform such spying has been around for a long time, and in a bar or restaurant the owner (or rogue employee, or customer even) wouldn't have to resort to exotic techniques such as this light bulb trick. A simple microphone would suffice.
> Along similar lines, I've wondered about how often restaurant and bar owners have bugged their own premises.
In a few select Houston strip clubs.. all the time, or at least ten years ago some did, the ones owned by certain organizations. I haven’t been around that scene for a decade, so I can’t speak to now. I know of a few other places that seemed to magically never have trouble with city officials or permitting. Another such place, an after hours club, was frequented by the mayor’s “party-oriented” daughter, never had trouble with police raids, or fire marshals. The venue survived unscathed until the next mayor took over and such leverage became unavailable. The Texas Alcoholic Beverage Commission was the only real challenge, but often the local cops would be knowledgeable about pending raids and would graciously provide some advance notice. It certainly helped that many of those cops were paid as off duty private security by the venue. DEA was another frequent adversary, but those folks aren’t as undercover as they thought they were. I might suggest that the DEA was (or maybe still is,) one of the more sketchy law enforcement agencies in the government.
Just under the surface of “normal,” there is some really fascinating stuff that goes on.
Some Dunkin Donuts stores in the New England area were (decades ago) accused of listening to their customers by having microphones at seats where customers ate. Not sure if that was intentional or part of their security system as they claimed. [0]
The article involves a remote device watching the light bulb but it would also make sense to think about devices in the room as well.
Your device might have its microphone disabled, in hardware even, but I wonder if the ambient light sensor has enough gain to see audio frequency variations in room lighting. We already know hard drives and speakers and whatever else can act like mics for exfiltration...
I was expecting to be dissapointed (I thought this was going to be a test in the same room, with perfectly calibrated equipment) but I was wrong. Very impressive that you can do this at that distance and with affordable ($400) equipment. I wonder how long intelligence agencies have been using this capability.
I knew someone who started building one of the laser based units in the fall of 1990, in high school. So that's 17 or 18?
By the end of the decade or beginning of the next, we had telecom hardware companies filtering the signal to the activity LED so that you couldn't read packet info out of the flickering.
Odds are good that someone combined those ideas a lot earlier than we might like to think.
This technology and technique has been known for a number of decades. Intelligence agency use a variety of counters to avoid this issue. SCIFs don't have windows for a reason. The standoff distance for intel buildings are far, and well monitored for similar reasons. Some facilties use double insulated glass with a randomized noise maker in the frame.
Some corporate board rooms use similar technology and safe guards.
Sorry - but that sounds like BS. Think about it - a hanging light bulb vibrates when you speak... Really? I don't see any movement, and I'm right there.
But yet, that can be picked up 25m away by a telescope? With equipment that costs less than $1000? No way.
Sounds like some phony story to me, that's meant to make us think spies have superpowers, or even that we're all being spied on. (Which we are, but not like this.)
I watched the video. I'm sorry, but its inane. This is the worst sort of science magic. In fact its just trickery.
Do you seriously think that you have been provided evidence in that video? All that happened is that I ended up looking at a leaf and packets of things that weren't moving at all. That's not evidence!
It says that they are picking up vibrations from a leaf or bag of crisps. But they don't show you those vibrations. They just give you a bunch of graphics and fuzzy sounds and tell you that they have done something.
You do see that this is could be very simple video trickery right? As with the initial article - there's a claim, but nothing to really be able to use to verify that claim. Just a graphic, and references to Shazam. Perhaps this is guerrilla marketing for Shazam? Its a more viable thesis anyway, given the evidence!
Firstly, note that it provides no evidence. It has a claim, and shows a fancy graphic. That's it. Its not evidence. Anyone can make that sort of thing up. How can one differentiate this from a story I just made up?
Digging in, it says:
"LED bulbs also offer a signal-to-noise ratio that's about 6.3 times that of an incandescent bulb and 70 times a fluorescent one."
Why? Why are LEDs better for this noise reflection stuff, than incandescent or fluorescent ones? They're all in glass. The heat of the glass is stable. The LED itself surely can't help.
Why does Shazam feature in these tests, at all?!
It also says:
"Researchers have known for years that a laser bounced off a target's window can allow spies to pick up the sounds inside. Another group of researchers showed in 2014 that the gyroscope of a compromised smartphone can pick up sounds even if the malware can't access its microphone."
Right. Are we meant to believe that this is what spies do? I mean, a spy just has to access whatever systems are already at his or her disposal and listen to your calls. They don't need to do any of that!
And:
"Still, Nassi says the researchers are publishing their findings not to enable spies or law enforcement, but to make clear to those on both sides of surveillance what's possible. "We want to raise the awareness of this kind of attack vector," he says. "We’re not in the game of providing tools.""
So, they even say that this article is about raising awareness of this kind of attack vector. They are raising awareness. Thanks spies!
The article is a joke. At best this is a puff piece for a spy agency, or more likely just to increase paranoid awareness amongst those in tech. Be scared, danger everywhere! There are spy agencies listening to you everywhere - they're not using your phone, they use light bulbs and leaves!!
I think we should demand more. We are being presented science right? We should be able to recognise that this article and the videos, etc are just presenting a claim. They are not providing any evidence. This is just a story.
I'm saying, if someone writes an article and it is published, should we believe it?
You're saying that I should use tools to replicate the experiment. Great - this is what we should do before we accept something as true. Otherwise, we are in the realms of 'belief'.
So, do you believe this story? And did you do due diligence, and confirm with the tools that it is true? Or did you skip that bit and believe it despite the fact that NO evidence is presented at all? Be honest when you answer please!
Stepping back a bit, I'm not saying anything so drastic. I certainly don't see why I'm being downvoted. I'm really talking about applying the scientific method personally. I'm saying don't accept articles, videos by default, without even critically reading what is being presented. I really think its pretty obvious stuff, tbh!
You’ve been provided with links to an article from a credible source, a paper, and a video, and you say you don’t believe any of it. The physics of this effect are both well known and easily demonstrated. What level of evidence is required in your mind?
PS, at what point would alarm bells ring for you, re this story. If they had said that they could pick up voices from a lightbulb with a telescope that was 50m away? 100m? 1000m?!?
I suspect that alarm bells wouldn't ring at all....
And why should some skepticism of an article relating to the purported tools of spycraft be downvoted? Its ok to expect more than a simple schematic, surely?
> Think about it - a hanging light bulb vibrates when you speak... Really? I don't see any movement, and I'm right there.
I don't see any movement on my computer speakers or mic when they're emitting or receiving sound, but they wouldn't function if vibration wasn't happening. You can measure the vibrations from a short distance away with a laser setup that costs about $300 IIRC.
So the basic concept is sound. After that it's just a question of what kind of measurement setup a government agency with absurd amounts of time and money could cook up.
The voice and music recordings they used in their demonstrations were also louder than the average human conversation, with speakers turned to their maximum volume
I can clearly see vibrations in glass near loud speakers with my naked eye, but when someone is speaking I generally can't see any vibrations. Still really cool that they were able to reproduce sounds this way, but it's not like you can spend 1k and be able to actually listen to conversations using this method.
If their chart is correct, the 30cm mirror show a 20db improvement in SNR at 200Hz compared to 20cm, same curve but 20db lower gain.
That is an absolute crazy scaling, I'm not entirely sure how that's possible, though I have a suspicion. If it's an actual scaling law that can be exploited, I would expect you can extend this to entirely practical audio volumes on a shoestring budget.
In college for a project I wanted to encode messages over small brightness variations in light bulbs. It's a pain to deal with AC, so I ended up doing it with a laser pointer. It turned out to be stupidly easy with a laser pointer and photodiode attached almost directly to a pair of PC speakers. All these devices have a +5v DC signal. I could attach my iPod headphones to the laser pointer directly, then the photodiode to the speakers, and boom, music at about 100 yards, with almost no quality decrease, though I didn't do anything quantitative :)
I've always wondered what places communication started getting hidden since we went online and no longer rely on things like numbers stations or drop boxes (though I assume dead drops still exist)
K what? That's amazing! Apparently I need to learn more about lasers.... haha :) I guess it's pretty "fragile" since the slightest misalignment of the source laser & destination sensor would mean a total loss of signal, right?
Given that the laser forms a spot, minor misalignments are very easy to debug (just line up the red spot with the sensor again). Also given a laser with sufficiently wide spread, minor misalignments would manifest themselves as volume loss.
Not great if you don't have easy access, but reliable enough for e.g. a dorm room or a living room.
This was a lab we did in an intro EE class in college and it worked wonders as a "whoa cool" demonstration with very basic EE knowledge required.
Seeing as how Starlink doesn't have this capability, it would probably be better to study any other instance of free-space optics, many of which exist.
The other commenter is right, store bought lasers have a descent spread over a few dozen or hundreds of feet. And the signal is still pretty readable to the receiving photodiode.
Free space optical system *intentionally" spread the laser beam using a lens. The light beam still consists of parallel rays, and when they hit the receiving lens they focussed again on the receiving diode (or whatever sensor). The idea is to accept that the beam will be disperse in the air, or due to variation at the transmitter, but still to catch enough of the transmission. https://en.wikipedia.org/wiki/Free-space_optical_communicati...
A long time ago back in college our team designed a free space optical gigabit digital link via laser diode and photodiode. We were transmitting I think 1 Gbps easily over maybe 50 meters. We didn't have enough time to polish the system or test faster speed because the sponsoring corporation took our design back to their RD group.
Twenty years ago I worked at a place that had offices in buildings across the street from each other and had a laser network link between them in case the link through our ISP stopped working. I don't recall if we ever had to rely on it while I was working there, but I remember the guy in charge of monitoring it used to complain that it didn't work on foggy mornings. One end of the link was on the 21st floor, and we got fog at that level a lot more often at that height than at ground level.
Is an old project, that I have memorized the bookmark ('Optar', is another project somewhere on that page that puts a multi-kb 2d barcode data on a printed page to be reliably scanned in, even if damaged)
I wish I could, I'll see if I can dig up my old report from class, but that was from 2007.
Very simply, I literally just wired the laser in series with the headphones, like, attached directly to the battery of the laser pointer so the audio +/- 5v would jiggle the laser's brightness around that value (the laser was also +5v so this "just works" if I recall. I might have had a resistor in there, and at some point, current leaked back into my iPod and fried it :)
On the receiving end, same deal, PC speaker audio jack attached to a photodiode and power.
It's remarkably analog, but it's also the same reason that things like the now anachronistic CD-player cassette adapter work, throwing 5V AM signals around :)
Li-Fi projects are the best if you want people to get into information transmission. It's simple an cheap and easier to grasp than any radio based transmission. Two diodes (you could even use a LED as sender AND receiver!) and you are ready to go.
Those stories keep fascinating me, ever since I talked about side-channel attacks during my study times.
Re-creating CRT images through walls, listening in on keystrokes through the electric wires of a building or learning about traffic patterns while observing network interface flashing LEDs.
Just a fascinating world when you leave aside what they are aimed towards.
A friend of mine had a side-gig at an electronics factory that used to fit/repair government printers, telegram, and fax machines in the 80s. When he first started he noted that the schematics said to double up the number of capacitors. When he raised this as a potential error in the schematic with his boss, he was told that it was quite deliberate and was there to stop giving away what was being printed by anyone measuring the power fluctuations. Feels kinda obvious now but when I first heard it I had a mini-mind-blown moment.
The shock absorber isn't a great analogy for a filter cap but, a capacitor has inductance and resistance, even though it is primarily a capacitor and we usually ignore the L and R for simplicity's sake.
Since capacitors differ in their high-frequency characteristics (and capacitors with good high-frequency properties are often types with small capacity, while large capacitors usually have worse high-frequency response), decoupling often involves the use of a combination of capacitors. For example in logic circuits, a common arrangement is ~100 nF ceramic per logic IC (multiple ones for complex ICs), combined with electrolytic or tantalum capacitor(s) up to a few hundred ?F per board or board section.
A similar mind-blown moment for me was reading about foreign intelligence agencies listening for radio frequencies emitted by typewriter keys. They were able to map specific frequencies to specific letters, and could reconstruct most outgoing letters. It inspired a side-project using an SDR (software defined radio) to fingerprint specific hardware I had around the house.
TEMPEST, it's as old as the Cold War. In retrospect, side-channel attacks and bugs are always the norm, a cryptographic break is rare. A good introductory article is TEMPEST: A Signal Problem written by the NSA.
I even remember there where schematics for a laser-microphone circulating on the early 2000s internet that could translate the vibrations of windows into sound.
To defeat this, don't you just do what the mobsters in movies do? Play music or make some other loud noise and then whisper or talk softly close to each other?
The best thing to do when you see a comment like that is to click on the "n hours/minutes ago" link next to the username to open the comment on its own page. There you will find a "flag" link as well the downvote button (they do separate things).
If enough users flag a comment it will turn "[flagged] [dead]" automatically.
I'm not sure on the cost, but I know it's common to use lasers at an angle on windows to spy on conversations via a laser microphone. That requires less of a line of sight. Is this a cheaper method?
I'm getting tired of these stories, it's always the same set of researchers doing different variations on the same thing, and feels increasingly like pg's submarine.
reply