Audio compression and encoding is such an interesting area, and I'd recommend anyone to read up on the basics. I had a naive idea about compression being something-about-cutting-waveforms or something like that, and had no idea that principles of vocoding was used to encode phone conversations - we're essentially encoding speech by passing artificial air through a software model of an artificial human speech system, and decoding it by blowing the air back out the other way! The compressed data is essentially not even the sound itself, but parameters about the artificial tongue, mouth etc.
I know this is extremely dumbed down, but every once in a while you realise there's a subject out there you know nothing about, that's just so interesting to learn a bit about - and speech encoding definitely checks that box for me!