179,854Messages
9,130Senders
30Years
342mboxes

← back to listing · view thread

From:
Kent Williams
To:
Date:
Tue, 16 May 1995 10:29:53 -0500
Subject:
Re: Frequency Ranges
Msg-Id:
<9505161029.ZM20293@elvis.cadsi.com>
In-Reply-To:
<gapv96@udcf.gla.ac.uk (derek) "Frequency Ranges" (May 16, 3:44pm)>
Mbox:
idm.9505.gz
If you ask an audiologist, they'll tell you most adult human beings have hearing that rolls off somewhere between 16khz and 20khz, and that as they get older, their hearing rolls off even more. If you ask an audiophile he'll tell you his CD's sound different if you coat the edges with red magic marker, and that he can hear the difference between different brands of speaker wire. While you can't hear a signal at > 20khz, if one of sufficient strength is mixed with an audible signal, subharmonics and phase interference may produce audible artifacts. This is not the same as hearing a 30khz tone. There are two proposals to increase sampling rate & # bits per sample in digital audio that have some merit, but neither is made necessary by the need to transmit signals people can't hear. Increasing the sample width from 16 to 18 or 20 bits has the effect of increasing the resolution of very quiet passages. Increasing the sample rate beyond 44khz allows you to construct simpler filters to roll off the aliasing and other crap that creeps in over the Nyquist limit. Life is too short to waste on this sort of bullshit. Please stop this discussion or move it somewhere else. -- kent.williams@cadsi.com -- Can a filter sweep the noise floor? How to program computers: Sit still. Stare at Screen. Type. Repeat.