Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!nbires!hao!hplabs!ucbvax!CORY.BERKELEY.EDU!dillon
From: dillon@CORY.BERKELEY.EDU (Matt Dillon)
Newsgroups: comp.sys.amiga
Subject: Re: processing of digitized audio on PCs
Message-ID: <8701222104.AA03060@cory.Berkeley.EDU>
Date: Thu, 22-Jan-87 16:04:47 EST
Article-I.D.: cory.8701222104.AA03060
Posted: Thu Jan 22 16:04:47 1987
Date-Received: Fri, 23-Jan-87 06:45:08 EST
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: University of California at Berkeley
Lines: 35


	30Khz @ 14 bits?  Maybe for music, but that kind of resolution is
wasted on simple voice.  8Khz @ 8 bits gives you almost undistinguishable
voice replay. (With the addum that the circuit must be completely shielded
and properly designed to avoid noise.  But since this is an absolute 
requirement for 14 bits anyway, it doesn't detract at all from my point).
Generally, ADCPM encoding (a sort-of delta encoding) is used to compact the
data:

			VOICE (delta encoding of bits, not nec. ADCPM)

	BITS	RESULT
	1	character only
	2	badly garbled
	3	barely understood
	4	understood quite well, but obviously synthesized
	5	somewhat better
	6	even better 
	7	almost indistinguishable from tape
	8+	indistinguishable from tape


-------------

	As far as music is concerned, your now talking 12 bits @ 60Khz
to get CD sound (though I would like to point at the even CD's have their
failings in the higher frequency ranges).  I have no idea what kind of data 
compaction is employed by sound-cd's.

(P.S. I haven't looked at the specs for CD's recently, so I may be incorrect
in my specifications here).

					-Matt