A "in a hurry" hack might be to run it through a bad phone line? Still understandable, but you can probably guess that the phone provider is going to crush the spectrum and significantly reduce the bit depth.
Sort of like the content recognition systems in use:
https://en.wikipedia.org/wiki/Automatic_content_recognition
on the other hand, I found this interesting:
In January 2018, a YouTube uploader who created a white noise generator received copyright notices about a video he uploaded which was created using this tool and therefore contained only white noise.
Maybe something like mu-law, which only has 3 kHz bandwidth centered on human speech (as opposed to mp3 with usually 44.1 kHz) would perform best.
I also noticed that zoom corrects for lag in my connection by speeding up and slowing down speech. (I noticed that when my counterpart was running a metronome.)
Zoom could artificially introduce micro-lags to encode data. Eg as a human I can't really tell whether some sound starts on an even or an odd milli-second.
Not really. And if a user does, they will hear rather mundane codec artifacts, and or noise that is basically like other noise.
Given a redundant mark, and the system I linked has multiple, concurrent redundancy, the chance of the mark surviving grows with the amount of audio.
They can literally drop marks during speech the entire time.
POTS is a few KHz. So less, but without looking into Nielsen's fingerprinting methods I wouldn't say you could assume that the fingerprint wouldn't be preserved running it through a phone line first.
You lose the "voice" in the process, but in some situations it might actually be fine