This is called the “nyquist frequency.” Basically, in order to accurate record (and later replay) sounds at a given frequency (in this case around 20 kHz like you mentioned) you needs to “sample” it twice as often as that frequence. Sampling means taking a snap shot of the value at a given point in time.
This is because sound is a wave and you need to require both the peak and the valley to accurately recreate it. And a freuncy of 20kHz means there’s 20k peaks and 20k valleys every second. So that’s 20+20 =40.
As for why we go up to 44 instead of staying at 40 (a lot of people can’t even hear about 17 kHz so even lower might work) I’m not entirely sure.
Hopefully that makes sense, this stuff is a lot easier to explain/understand with pictures.
Latest Answers