Foil the paparazzi

September 19, 2005 8:05 AM

Foil the paparazziGeorgia Tech researchers come up with a system that senses nearby digital imaging devices, and fires a beam of light at 'em, foiling attempts to take pictures of 'ya. More high-tech (but less entertaining) than having Sean Penn smash the paparazzi cameras.
posted by RonZ (29 comments total)

Cool idea, but the paparazzi will always come up with a solution.
posted by cleverusername at 8:39 AM on September 19, 2005

also good for suppression of the press in other more importaint areas... but useless gainst good ol' fashion film?
posted by edgeways at 8:39 AM on September 19, 2005

The article is a little garbled-sounding on the mechanism involved. In some parts, it sounds as though the system works by detecting coated optics; in other parts, it sounds as though the system works by detecting a reflection from the CCD/CMOS digital sensor behind the lens.

I think it's probably the sensor reflection system, since that would limit its applicability to digital cameras, as the article implies. They mention that "Fast shutter speeds might also present some challenges"; that would certainly be true in the case of a sensor-reflection detector trying to detect a digital SLR, where the sensor is only exposed for the split-second that the picture is taken.

In any case, the ~10m range implies that even a moderate zoom lens will circumvent this thing pretty easily.
posted by Western Infidels at 8:55 AM on September 19, 2005

Anything that stops them stalking me when I'm out with Halle Berry is fine by me.
posted by Pericles at 8:58 AM on September 19, 2005

This is the same technology behind the 'Spy Cam Finder' (boingboing had a link re: these a while ago) but with the added layer of active detection and focused (?) targeting. Frickin' lasers.

The eventual outcome is probably a dome-shaped unit for greater coverage, but it would have to be carried around the person or object of interest, 24/7, and that seems like serious overkill.
posted by bhance at 9:00 AM on September 19, 2005

While a little cumbersome-sounding, it's a lot better than the alternative that relies on cameras that "turn off when they receive a bluetooth request to do so."

Because if there's any reason *I* need to upgrade, it's because my new camera will have software on it that cripples it. Wa-hoo.
posted by Imperfect at 9:03 AM on September 19, 2005

It seems like it can only "protect" you if your standing next to it and the photographer is right in front. It could be defeated by the photographer moving to the side. What's the point of that?
posted by cillit bang at 9:06 AM on September 19, 2005

I don't know, I kind of like watching Sean Penn deck photographers.

And any technology that makes it harder to get I-forgot-my-underwear photos of celebrity idiots is bad technology.
posted by fenriq at 9:24 AM on September 19, 2005

Western Infidels: I don't think so. It specifically referred to using IR, which makes me think it's using the lens coating. Silicon photovoltaics like CCDs and CMOSs have a really high response to infrared light, and so most cameras have a filter on the lens that blocks IR from hitting the sensor. And when I say "blocks," you can read "reflects." I believe this is also true of film cameras, but I don't know the response of silver halide in the IR region offhand, so I'm not positive.
posted by solotoro at 9:26 AM on September 19, 2005

This strikes me as a really dumb way to spend research money. And the article makes it sound like it detects lens coatings, which would mean this is is no way limited to digital. But I doubt the author of the article has any clue how this works, so you'd probably have to talk to the researcher to know for sure what this thing does, as the talk of sensor types and shutter speeds seem to invalidate that.

If this could pass a real world test, I'd be surprised. I can take a picture with an exposure of 1/8000 of a second, and I can push up to 3 stops if I shoot raw. I don't have anything like the lenses a Paparazi has, but with my long lens and a fast shutter speed, I doubt this technology would have any chance. It also sounds like putting a polarizer on my lens would completely foil it.

But thank god we've made the world safe from the scourge of unwanted photography.
posted by teece at 9:30 AM on September 19, 2005

This strikes me as a really dumb way to spend research money.

I see studies released all the time that are really dumb. Doesn't stop people from needing to research even more.
posted by NationalKato at 9:34 AM on September 19, 2005

solotoro: maybe they put that coating on the lenses of cheap point-and-shoot digicams, but they don't on DSLRs. They block IR by putting a filter right over the sensor inside the camera on some models, I'm not sure what others do to deal with IR. The average SLR lens most certainly does NOT block IR light, though, as many folks use them for IR photography.

And if this system does not work against SLR cameras, it's worthless.
posted by teece at 9:34 AM on September 19, 2005

Won't somebody think of the dressing room voyeurs?!
posted by squirrel at 9:44 AM on September 19, 2005

This strikes me as a really dumb way to spend research money.

Unless you're tired of seeing photos of Dear Leader giving the finger all over the internets. Everyone knows that news photographers are among those who hate America most.

Seems to me that this technology could be a standard feature of GWBs public entourage.
posted by three blind mice at 9:49 AM on September 19, 2005

solotoro: I think teece is correct, the inbuilt IR filter/reflector is generally the last thing light passes through before it hits the sensor - it's not part of the lens. I'm pretty certain, from on-line discussion on the matter, that this is the case with my own (mass-market, major-maker, non-SLR, point-and-shoot-ish) digital camera. Despite the presence of the filter, the camera picks up near-IR (like from a TV remote) pretty well.

IR filters are probably common in digital cameras, but I don't think they're universal. Previous generations of my own camera supposedly didn't have one, and were good candidates for IR photography when fitted with an IR-pass filter. I don't know if that has implications for a camera-disabling system like this or not.
posted by Western Infidels at 9:52 AM on September 19, 2005

Many years ago I saw a demo of sort of a brooch or thing that you could wear that was basically a flash slave. When it detected a flash, it set off one of it's own, which would basically blind the camera pretty good and screw up the exposure. It's down side was that it's recharch time was terrible (obviously, there are going to be multiple paparazzi each with big battery packs, so competition would be quite difficult)
posted by RustyBrooks at 9:58 AM on September 19, 2005

teece: It doesn't matter which side of the lens the coating is on. The light will bounce off it the same. Lenses work in both directions.

On the Canon D30, at least, the IR filter is behind the mirror. So, yeah, you'll only see it while the shutter is open.
posted by aubilenon at 10:28 AM on September 19, 2005

Whoops, my bad. I stand corrected. You'd really think I'd know better than to rely on my own shoddy memory without fact-checking by now.
posted by solotoro at 10:48 AM on September 19, 2005

Sure, aubilenon, it does not matter if the IR coating is on the lens or somewhere else. But I know it's not on my Nikkors.

But I suspect the article was just wrong to talk about lens coatings, the more I think about this. I bet it uses IR because it is not visible, and because so many digital cameras are highly reflective of IR on purpose. That would also explain why it's a digital-only thing: film cameras don't do anything to filter IR, as you can (mostly) control IR sensitivity with the film, without the need for any filters.

But if that is how this works, it's got major problems. IR filters are not universal on digital cameras, so it won't work against all of them. It will have issues with reacting fast enough to short shutter speeds on SLRs, as on those cameras the IR reflectivity is not in play until the shutter is depressed. But lastly, it is totally useless against film, which is a fairly big deal.

Overall, this seems like a technology that is going nowhere, to me. Or, at the most, it will be deployed in places like CIA headquarters where millions can be spent and lives are at stake. But I doubt this will ever be deployed to stop paparazzi.
posted by teece at 11:09 AM on September 19, 2005

It is difficult to get a grip on what exactly this system is detecting from the vague writing of the article, but my impression is it's the reflections from lens coatings. All modern photographic lenses on modern point-and-shoot and SLR cameras (film and digital) are coated (mainly to reduce flare and increase contrast). Coatings reflect light differently than an uncoated glass surface, so I guess this sort of makes sense. Using an IR beam to foil the camera also makes sense, if it's powerful enough, because most digital sensors record a little IR just beyond the visible - even those with IR filters in front of the sensor, although just how much varies with the filter each manufacturer uses.

Practical ways to beat this might be as simple as stretching a bit of very fine cloth over the lens (softens the image slightly, but still gets the shot), using a polarizer (still coated, but reducing reflection a bit) or simply being very quick with getting the shot (no idea how long this system might need to aim and fire).

Another problem they might have is how many beams they can fire at once if there's more than one camera about. What about what this will do to CCTV cameras? They have coated lenses too. As do some lenses in glasses. Or, yes, just use a long lens from further away.

On preview: Looks like I disagree with teece. Maybe he's right and it's not the lens coatings at all. But then this seems like a bad idea for all of his reasons.
posted by normy at 11:17 AM on September 19, 2005

I'm going to wager that it's detecting the autofocus IR stuff being sent -out- of the camera..rather than some kind of reflection off a lens.
posted by odinsdream at 11:25 AM on September 19, 2005

cleverusername writes "but the paparazzi will always come up with a solution."

Like that unobtanium tech 35mm.

The article is really unclear but it seems to only effect CMOS and CCD chips, which would make sense if it was using an infrared knock out beam.

It doesn't really matter anyways this tech will be instantly outlawed if truely effective. Why? Speed, red light and security cameras would be effected if it worked for all cameras. So either it would only work in limited special cases, there by not effecting paparazzi at all, or it will work on everything and therefor be illegal practically everywhere.
posted by Mitheral at 12:22 PM on September 19, 2005

Ok I could see this working under very specific limited circumstance like maybe within the confines of a lab against cellphone cams (could even be rigged to automatically electocute the guy pulling his phone out if wanted) but not for the much more marketing-worthy and attention-getting concept of an anti-paparazzi weapon. Too many variables and avenues of counter-attack such as using block-long distances, multiple cams, older non-electronic tech etc.

I'm going to wager that it's detecting the autofocus IR stuff being sent -out- of the camera..rather than some kind of reflection off a lens.

Could be, and all these auto kinds of features can be turned off on a DSLR. It could be as simple as that, just reconfigure to manual mode on the fly.

Older tech would be a very fruitful avenue of counter-attack in general. Still cameras have been around a long time, you can pick up old film bodies that use no electronics at all for next to nothing.

The article is pretty fuzzy but they're really out to lunch if they're depending on modern lens coatings. In the Nikon system there are 35 year old (and older) lenses that are very usable on a brand new digital D70 for example. Also, the coatings on front elements are only refinements anyway, they're not necessary for cameras to work. It is entirely possible to polish them off and still take perfectly usable shots.
posted by scheptech at 12:28 PM on September 19, 2005

We use a rotating color pattern to overwhelm the camera's auto exposure and white-balance algorithms, in addition to over-saturating the CCD to produce the best effect.

scheptech writes "just reconfigure to manual mode on the fly."

Yup. If this is the resulting image when the neutralizer is running but not targetting the camera, I assume that this would also be the result of correctly exposing the scene. Anyone with experience beyond snapshot photography would be able to avoid this problem by locking exposure or switching to manual.
posted by Kiell at 1:38 PM on September 19, 2005

I think it would be trivial to defeat this system. It uses, according to the article, a localized beam of light which is focused on the place where the system believes a camera lens to be. Since this system has no means of detecting when a photo is actually taken it has to keep up this beam for the whole duration it "sees" the lens. Which means (since the system doesn't have an infinite number of constant, tightly focused light beams at its disposal) that you only need to create a number of false positives that's too much for the system - depending on whether this works better if the false targets are close to the actual lens or farther away, you could use rings with coated glass pieces, a necklace, rhinestone-like shinies on the edge of your jacket etc.
Heck, if the stuff is cheap (it just has to pass the reflectiveness test of the system, it does not need to have the actual optical properties of a finely tuned camera lens) just throw a couple dozen coated glass pieces on the ground and watch the system trying to cope with that.
posted by PontifexPrimus at 2:01 PM on September 19, 2005

Which means (since the system doesn't have an infinite number of constant, tightly focused light beams at its disposal) that you only need to create a number of false positives that's too much for the system

If you look at the actual rig, there aren't individual beams but a single beam from the projector. I presume that, instead of a beam, the computer just adds a white spot to the image being sent to the projector, simulating a beam.
posted by cillit bang at 2:39 PM on September 19, 2005

In fairness, the aim isn't to combat DSLR cameras:

Our primary goal in addressing this problem was to design an environment that prevents certain portions of that space from being captured by mobile phones that include a CCD or CMOS camera.

The news.com article is slightly misleading, as is MeFi FPP. I'm not sure of the relevance of the sensor being able to spot "lenses cloaked with infrared filters" as, to my knowledge, mobile phone cameras don't use filters.
posted by Kiell at 2:58 PM on September 19, 2005

An IR filter isn't all that sophisticated. You can buy 16" square inches of Wratten IR gel filter for less than C$100. You could probably get 100 PhoneCam filters out of a single 4" gel.
posted by Mitheral at 12:29 PM on September 20, 2005

It would be so much cheaper to just have your clothes impregnated with xenon strobes. This whole phenomenon is so bizarre... technology, what have thee wrought?
posted by squirrel at 10:46 PM on September 20, 2005