It is weird, though, that they're different modes and you have to choose on both the device and the software which to use.

Imagine if every individual song, speaker/earbud, and MP3 player had a different implementation of 8/16 bit music, and it was up to you to compare/contrast the 8 variations to decide if 8 or 16 bits was more auditorially satisfying.

You get this with like levels and mixers, but not with such a fundamental quality as definition. Sure, you can chase hi-fi, but I feel like music has less artificial upscaling and usually hi-fi is a different product, not a toggle buried in various menus.

I don't know, it's kind of crazy.

It’s the transport. Imagine if the D/A was in the speaker; you might set params differently then.

We kind of did have an equivalent situation back in the day - you used to have A-law vs μ-law

Super Audio CD used 1 bit audio (at a very high sample rate) and sounded just fine

32 bit audio is higher resolution, it can have effects applied to it in higher detail. like subpixel coordinates might more clearly specify a pixels colours...

anything beyond 16bit (which is streamed to your speakers) is not audible though in the sense of playing it to the speaker. it matters in processing (dsp). (44 or 48khz inguess?)