A band-aid solution

I am creeped out by having a built-in video camera on my laptop. I never use it, so the only thing it’s good for is being taken over by malicious spyware in order to stream video of me pottering about in my bathrobe to malfeasants across the globe. I don’t trust disabling it in software – I’m an OS programmer, and I can think of a dozen ways to turn it back on. I don’t even trust the little "I am recording" light to go on. Even if the programming interface for the device doesn’t define a knob to turn it off, it’s the kind of thing you’d build in to the chip and disable via firmware (or just not mention it in the spec). Go ahead and call me paranoid because, well, I am.

So the only reliable solution is to disable it via hardware – i.e., tape it over. I finally got around to it today, but when I pulled out the electrical tape, I realized it might gum up the camera lens, which would be irritating if I ever wanted to use it again. So I used a band-aid. I took one of those "spot" band-aids, trimmed it a little, and aligned the padded part over the camera lens. Results:

And because I’m *that* paranoid, I took a picture with it:

I release this valuable intellectual property into the public domain.

19 thoughts on “A band-aid solution

  1. OLPC camera

    Tidbit: The OLPC XO camera and microphone lights are (fairly uniquely, as you suggest) directly hooked up to the hardware for precisely this reason. If voltage is applied to either device, the respective warning light goes on.

    Your paranoia is not alone. :-)

    (I have yet to comment on your new hair. I am still recovering.)

    — Jeff (who can’t openid on LJ atm)

  2. paranoid

    I don’t think that’s paranoid, actually. The odds that precisely your camera would be misused is probably low. But the odds that someones camera will at some point be misused by someone is 100% in my mind.

    About the “now-recording”-led, if I designed the hardware the led would be physically connected between ground and the power to the imaging-chip, so that it is physically impossible to provide power to one, without providing power to both.

    But I am certain that atleast some of the laptops out there won’t be constructed that way, there -are- legitimate reasons to want random leds on a laptop to be software-controllable afterall.

  3. Yeah, I always felt a bit odd about that little camera when I had a MacBook Pro. Not quite paranoid enough to do anything about it, though, so kudos to you!

  4. Re: paranoid

    Mm. The camera in the Apples is a dumb USB controller which accepts a firmware upload and then reappears as something that’s almost (but not quite) entirely unlike a USB Video Class device. The firmware is part of the driver file on disk. I strongly suspect that the LED control is just hooked up to a GPIO line on the chip and that an appropriately modified driver would have the undesirable behaviour. I guess it’s possible that they’ve tied it to the voltage line for the sensor, but that starts sounding sort of needlessly complicated…

  5. not entirely blank

    Val, you’re just ASKING for someone to do hollywoody some image processing magic on that “blank” picture, and comment upon the color of your wardrobe.

  6. Don’t forget the microphone!

    Way back when, I used to have fun with peoples’ microphones remotely… If anything, they’re more of a worry. The bandwidth requirement is tiny, so it won’t be noticed. And it can be recorded to disk in batches and dumped all at once.

  7. Re: paranoid

    See, if *I* were designing a camera, I’d make the LED software controllable because I’d want to sell it for all kinds of applications, most of which would not be laptops where the user wants to know that the camera is recording. Keep your options open and all that.

  8. I think *some* laptops have a proper hardware-based killswitch for wireless, others it just sends a soft signal for the OS to deal with. I vaguely recall one of the NetworkManager guys commenting on a setup where the hardware killswitch was completely shutting down not just wireless transmission, but the actual network device itself, causing it to disappear.

  9. My Dell XPS M1330 had a switch that killed the wifi antennas. It worked great.

    Linux just would log a message about ‘kill switch was thrown, we can’t use the wireless’ and stop using the card. I don’t think it was a software option; the hardware was being turned off.

  10. The hardware is being turned off by a line on the PCIe slot being asserted, and the firmware in the card responding to that by powering down the transmitter. Given access to the firmware I suspect that you could still get somewhere in disabling that functionality.

  11. Oooh! I have a Dell XPS M1330 too, and I just found that switch. Nice! (That’s the only thing I like more about it than my Air or my Panasonic R5.)

  12. Re: Don’t forget the microphone!

    Man, this one is harder. I think the mic is behind the grill that also houses the light sensors. But then, I’m not a super-big fan of the automatic display brightness changes in response to ambient light.

  13. Re: Don’t forget the microphone!

    No, no, only one of them has a light sensor, the other one is the mic. I will have to experiment with muffling it.

  14. Re: not entirely blank

    perhaps, but i’d say the bandaid was fairly effective:

    wget -qO- http://valhenson.org/pix/camera_test.jpg | djpeg -dct float | pnmnorm | pamscale 0.0625 | pnmsmooth | cjpeg -dct float -quality 90  | python -c 'import sys; print "data:image/jpeg;base64," + "".join(sys.stdin.read().encode("base64").split());'
    pnmnorm: remapping 7..20 to 0..255


  15. Needs a hardware solution

    My wireless device on my laptop has a hard switch to disable it (and the bluetooth). Why dont these laptops with built in cameras have something similar?

Comments are closed.

%d bloggers like this: