The reason for 32bit resolution is that during signal processing, which includes all kinds of effects and editing, overflow is very common with short word. Typical effective sampling range is about 24bit, hence with 32bit word processing depth, overflow will not occur. When converting back to 16bit for the final product, dither will be used to ensure good quality. Obviously you don't know what protools, or DAW is, it is not surprise, since those are not consumer product, unless you work with pro audio, not "listening" but actually production for music or movie, you won't come across these things.