From Peter Shultz Sent Tue, Feb 16th 1999, 10:29
>Aw jeez, this is totally bogus. 16-bit 44.1 kHz is the most bare-ass, >hardly passing standard for digital audio. Okay, maybe not totally because >the 16-bit part is pretty good, but the 44.1 kHz part is atrocious. Interesting you should say this -- I find that the bit depth is more of a limitation than the sampling rate. Perhaps not in IDM, where most of the music hovers around roughly the same dynamic level throughout the duration of a track, but classical music is often recorded at extremely low levels (so that the loud bits don't get clipped), with the result that it uses only 12 or 13 of the 16 available bits, and quantization noise becomes a distinct problem. >Once you get to 22.05 kHz, >this is what any waveform is going to be from a 44.1 kHz audio recording: >\/\/\/\/\/\/\/\/\ >That's it. Absolutely no detail at all. It gets better the lower you go >from there, but that's why it sounds so bad. Okay, this isn't exactly right... Any waveform that looks like /\/\/\/\/\/\/\ is actually a superposition of a very large number of waves, with the lowest-frequency (in your example) at 22.05 khz. Therefore, all of the other frequency components to that sound are well above the range of human hearing and do not enter into the picture. (and there are numerous completely valid workarounds for Nyquist aliasing as well, so it's really no limitation). The *real* limitation of CD's, IMHO, is their restriction to two channels...