Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

> The camera is engineered so that it can’t activate without the camera indicator light also turning on.

Has anyone torn down the hardware and verified this lately?

It should not be a complicated feature at all. Just ensure you have the same voltage applied to your LED circuit as your camera circuit, and enforce that by having them on the same wire...



Wait: that's not the same as having a camera cover.

A camera cover guarantees that the camera can only work when the camera is uncovered.

A camera light alerts you once the camera is already on.


Important distinction. If your microphone is compromised, your web cam might be triggered to turn on just long enough to take a photo, or the camera could turn on only when the mic is quiet and there are indications the person is not engaged with the computer.


Note that this is not only about security. Sometimes a user simply makes a mistake, or have an app configured the wrong way, and accidentally turn on their cameras without intending to. Nothing has been compromised, everything is secure, but they still suffered an unexpected exposure. A camera cover will prevent many of these situations.


Should be a tiny mechanical device in addition to the light and it should only unlock with touch id. Same for microphone. Honestly, I want the same thing for phones too. The ease by which we are monitored is too damn high.


That's a great point. Maybe there should be a touch-sensitive button up by the camera to turn it on.


Yeah, light can turn on when one is looking away from the computer or sleeping and could potentially be spied on. A cover is a non spying guarantee, at least the visual part, spies can listen on the mic though


A software fridge light!


Yep, and the cover can be toggled when the machine is locked or powered off.


Even if true, do I have time to verify for all devices? A piece of tape is a piece of tape.


Also, pretty unlikely, but what if the LED dies? :)


I’m not an electrical engineer so forgive me if this is a dumb question. Is it possible to wire it in such a way where the power actually travels through the LED, so if the LED ever stopped functioning or lost connection the webcam would literally receive no power?


Yes it's possible, as LEDs are indeed diodes, but done naively it would be a very bright LED, and it would tend to flicker as the current draw from the camera varies. It would be a hassle in general.

The LED could still always fail short, and then you're back to having a broken indicator again.


Very very unlikely. Low power leds rarely malfunction.

Add more than one if this is a concern, and it can still be compact. This is already done today in "RGB" leds.


It's not that simple, what voltage and current need to held on that wire to light the LED and the camera? It's not uncommmon to find a situation where a chip doesn't technically have the right power on it's power rails, but is instead drawing power on the input ports. It's non-trivial to prove this is or isn't happening.


No, I think it is that simple. The camera module likely needs either 3.3v or 5v. The LED can also be powered at either 3.3 or 5v by simply using a series resistor. You just hook the camera and LED to the supply in parallel.

> a chip doesn't technically have the right power on it's power rails, but is instead drawing power on the input ports.

What does this mean? The data lines of camera modules are generally differential pair, and it is highly unlikely that significant power is being drawn from them.


All I'm saying is that I have literally worked with systems that have drawn power over differential pairs and it's fucked with our power measurements. So don't rule it out.




Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: