Even if your speakers could produce frequencies higher than 20KHz (which requires 40KHz of bandwidth), the human ear can typically only hear up to 15KHz (which requires 30KHz bandwidth). The only reason that there's any sort of difference between 15KHz and 20KHz is various sample aliasing artifacts and the like; I can see merit in going up to 48KHz of bandwidth (since that's definitely way more than enough, whereas 44.1KHz still has some wiggle-room for claiming that there are audible aliasing issues), 192KHz is certainly overkill - not even dogs or cats can hear frequencies up to 96KHz, even if it were physically possible to have a traditional (i.e. magnetic) driver to vibrate that fast (which it isn't). Very high-quality electrostatic and maybe insanely-thin mag-planar drivers could reproduce frequencies that high, but it certainly wouldn't be audible. Unless you're a sonar dish, anyway.
Not to mention that the equipment which is recording and mastering such high-frequency signals, as well as the amps driving your limited-by-inertia speakers, do have this nice thing known as 'capacitance' which ends up applying a low-pass filter which would make it practically impossible to even record or amplify a 96KHz square wave (the highest-frequency waveform which 192KHz of bandwidth can represent).
Basically, audio bandwidth over 48KHz is definitely useless, and over 40KHz is probably so.
This is all in terms of the final mastered product, though; I can think of a few legitimate situations where you'd want more audio bandwidth in pre-production, but even then, anything over, say, 64KHz is probably useless. Certainly the final master doesn't need more than 48KHz.
Oh, and as far as bitdepth goes, the same pre-production stuff applies - 24- or even 32-bit audio in the production of music makes sense, but the final master certainly doesn't need more than 20-bit audio. 16-bit audio's artifacting is a whole whopping -110dB below reference (this means that even if your speakers were blasting at an SPL of 93dB, which is considered "really damned loud" and is enough to blow most drivers, the artifacting would contribute at most -17dB, which is about as loud as a fly landing on a stick of butter...)
The reason that the added bandwidth and bitdepth are useful in pre-production is that you generally lose a few bits of precision when you're futzing around with the volume, and you generally get bad artifacting if you change the pitch and/or speed of a sound and don't have enough audio bandwidth to hide the sample aliasing and the like. However, in most situations, 20-bit 48KHz is pretty massive overkill; in the situations where it isn't, the sound engineer needs to be shot in the groin repeatedly and convinced to find a new line of work.
"Is not a quine" is not a quine.
I have a master's degree in science!
[ Hug Your Trikuare ]
[ Parent ]