USB DACs with 24/192 via USB

I'll take 24/48 over 20/192 any day. 24-bit has 16 times the amplitude resolution as 20-bit

I would respectfully disagree, Johnny, and I certainly wouldn't apply that as a general rule. At best I think it would depend on the dynamic range of the music, and perhaps the high frequency content of the music as well. Yes 24-bit has 16 times the amplitude resolution of 20-bit, but 20-bit has a resolution of approximately 2^20 = 1 part in 1,048,576 = less than 0.0001%, assuming the bits are used effectively (i.e., assuming that some of them are not thrown away to provide overly conservative headroom in the recording process).

While on the other hand a 48kHz sample rate barely exceeds the theoretical minimum Nyquist rate (40kHz for a 20kHz signal bandwidth), and invites pretty much the same side effects of anti-aliasing and reconstruction filters which are generally recognized to have limited cd sound quality right from the start.

Regards,
-- Al

I think differences between 24/96 & 192 would have to be very slight.

That would be my instinct as well, and I would expect that differences in implementation and quality would generally overshadow the difference in sample rate.

Compared to redbook, though, although most of us are aware that 24 bits represents a 256-fold increase in information compared to 16 bits (at least potentially), I would view the benefit of the higher sample rate (even at 96kHz) to be at least as significant, if not more so. It is probably tempting to think of it as a little more than a doubling, but I think it is best viewed in relation to the Nyquist frequency (the 40kHz minimum sampling frequency which is theoretically required to capture a 20kHz bandwidth). Redbook's 44.1kHz exceeds the Nyquist rate by about 10% (which to me has always made it seem wondrous that it works as well as it does); 96kHz exceeds it by 140%, which should make possible vastly reduced side effects from anti-aliasing and reconstruction filters.

Regards,
-- Al