# Spying on microphones through electrical interference (or something along those lines)

Today I was using my `rtl2832u`-based software defined radio to listen to air
traffic control and such. At some point when I was scanning around in 6m
Ham-Band I noticed two pretty strong signals. I was suprised to even find
anything with an antenna only 50cm of length. However when I
frequency-demodulated any of the two signals I was able to hear myself speak. At
that point I got _slightly_ creeped out and started investigating.

I was able to confirm that the audio was what is being recorded by my laptop's
built-in microphone. My first hypothesis was that this was not actually
broadcast but just interference because the SDR is in the same electrical
circuit as the rest of my laptop.

Later I tried to check if this is a problem in laptops in general by my friend
testing it on their laptop. This did yield an audio signal, however the one of
**my** laptop! (They were positioned about 1m apart.)

When I arrived at home I tried spying on my laptop's mic but wasn't able to
reproduce this behaviour if it was not in the same circuit.

I would really like to know why this is happening. I still can't explain why the
signal is frequency modulated and why it is carried by such a high frequency
(around 50MHz). In case you want to reproduce, here are some details about the
setup:

- Laptop: Acer Nitro AN515-43
- SDR: Nooelec NESDR SMArt v5
- Frequency: 52.125 MHz

In the scenario where this can actually be received from nearby, this is a
somewhat critical security problem since even keystrokes might be recovered from
the audio.

Also if you have any expertise or clues please [contact me](/contact).

![Screenshot of GQRX](assets/2023-07-12-gqrx.png)