I am another user like you.
I have the old camera module and when you switch it off, the whole usb device disappears.
There was a post from FW saying the new camera behaves differently to the switch, but i don’t know the details of that.
Interesting. I have the 2nd gen camera:
Bus 003 Device 002: ID 32ac:001c Framework Laptop Webcam Module (2nd Gen)
The device stays present when the switch is flipped, but you see this instead of video:
The video freezes instantly when the switch is turned to the off position, and the warning placeholder pops up a second later.
I would still prefer a physical blockage of the camera lens. As it stands, I can be reasonably paranoid that the camera module is still on, and the switch simply indicates through software that the placeholder be displayed. I am forced to trust that the engineers did what the said they did.
I’m curious if this placeholder is the same across different OSes? It would seem like it should be, if the camera module itself displays it.
I find searching their blog posts is the best way to find these things, search?q= Camera site:Frame.work/blog
We’ve made some improvements in the hardware privacy switch circuitry to slightly reduce standby power consumption too. The privacy switch behavior on the camera also uses a new mode where instead of powering off the full camera, it cuts off the image sensor power and generates a placeholder blank image in the camera controller. This results in better behavior in applications and faster privacy mode switching.
I don’t recall FW giving schematics of the camera. But I would think it’s straight forward & not interesting. The image sensor will have power pin(s), VCC, you just cut it. The microphone is an i2c or i2s MEMS microphone if I recall. It will also have a VCC power pin that can be cut.
Their devices are open source (software, firmware), not currently open full schematic or open PCB design. They provide general overview schematics of their mainboards, github.com/FrameworkComputer/Framework-Laptop-13/tree/main/Mainboard, with the full schematics available to repair shops who sign an NDA. They probably would go fully open hardware if they could, but I assume it’s not feasible for them at this time. Last I heard, their hardware is designed in partnership with their hardware manufacturers, and not a lot of other companies are quite as on-board with open source as Framework is.
I tend to be a bit paranoid myself at times, but I feel there is no reason to not trust Framework when they say “it cuts off the image sensor power”. They have nothing to gain from lying, what they say is not technically difficult at all to do (cutting power to the image sensor), and they have never been shown to be lazy or unskilled hacks.
But the camera is not still on. The placeholder you see is the actual video feed coming from the camera controller. It’s not something on windows or linux doing that, it’s the camera itself. Your windows could be completely under the control of a virus & there is just no camera video for it to steal.
1 Like