@lanodan @volpeon you're wrong. Fuck it, I'll throw you a bone.
First, the reason why CD Audio used 44.1kHz instead of plain 40kHz is because low-pass filters fucking eat shit. You just can't cut out <20kHz frequencies in analogue, so there's some space to work with.
Second, aliasing is actually a thing. E.g. when you sample sound at 44.1kHz rate (or 48kHz if you're classy), higher frequencies might alias into the audible range and you will get shit instead of the sound you wanted to record. More on Wikipedia (
https://en.wikipedia.org/wiki/Aliasing).
Third, oversampling leads to higher sound quality because it allows dithering to remove sampling noise (aka quantisation error). There are several tricks that allow pushing most of that noise into the upper inaudible frequencies, leaving stuff below 20kHz more or less noise-free. Also, more on Wikipedia (
https://en.wikipedia.org/wiki/Dither).
Fourth, going back to DACs, higher resolution DACs work basically with the same principle but in reverse, allowing to reproduce signals closer to the original. Btw they aren't even 384kHz PCM. Most ADCs and DACs today are DSM (like in DSD format) and operate at 2.8224MHz or a multiple of that with 1-bit sample size. Again, more on damn stupid Wikipedia (
https://en.wikipedia.org/wiki/Delta-sigma_modulation).
TL;DR you probably should think about hi-res audio not as a means to listen to bat sonar or dolphin orgasm screams or whatever but as a form of anti-aliasing. Sorta similar to MSAA for graphics.
There, you made me do this. I hope you're happy now.