About a decade ago, researchers developed a way to recover sound from video footage. Lightweight objects in the video move in response to acoustic waves bouncing off of them, and image processing can identify sub-pixel vibration of said objects and reconstruct those acoustic waves.
The downside is that you need high video frame rates to capture the acoustic frequencies that matter for human beings. Ordinary HD video is about 30 frames per second, so you can only recover frequencies less than 15 Hz; if you want to recover intelligible speech sounds, you need a camera that can record at several thousand frames per second.
There was a whole episode of Mission Impossible (spoilers ahead for a 58 year old show) where they were trying to retrieve a hidden wire spool with secret information recorded on it. The evil guys are searching the whole area for a spool when the wire was hidden in plain sight by being made part of a landscaping fence.
Before digital, delay units (effects that produce echo) were using tape, but there were also BBD (bucket brigade) units that used a set of charged capacitors, Binson Echorec used metal drum or spinning plate with magnetic wire, and oil can delay used oil/emulsion on a drum wich whas electrostatic. Or something like that.
There’s an interesting synth on the market (released last year I think) that uses wheels on motors read optically, rather than using oscillators to produce sound.
About 15 years ago a member of the Home Shop Machinist Forum independently discovered this and I believe was trying to patent it. As best as I can recall he found that leaves on a tree vibrated according to sound that hit them. He was able to convert the vibrations back into sound and recorded trucks driving by, etc.I can’t remember how he captured the vibrations - it might have bee reflections from a LASER.
Using a laser aimed at an exterior window to remotely eavesdrop on conversations in the room by capturing the window vibrations is an old trope of spy movies.
And apparently is an actual threat as well. At least some high security buildings are built with multiple layers of outside windows to attenuate the acoustic signal from the room beyond recovery.
I understand it doesn’t even require a laser now - just high framerate video footage through the window - the minuscule movements(less than a pixel in the video footage) of objects in the room (the leaves of plants, or snack wrappers) can be amplified using a technique similar to motion amplification, to derive the audio from inside the room.
In the early 1980’s a friend of mine was working on a system using the motion of ferroelectric domains to read out the distance between electrodes that were lithographically patterned to encode short voice recordings. To my knowledge, this was never commercialized and was soon eclipsed by inexpensive non-volatile semiconductor memory. This was happening at the same time magnetic bubble memories were a thing, manipulating magnetic domains to store information digitally. By contrast, the ferroelectric device was analog. Magnetic bubbles were commercialized, but soon got squeezed between the rapidly improving semiconductor and magnetic disk memory technologies that soon dominated.