Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!apple!sun-barr!sun!pepper!cmcmanis
From: cmcmanis%pepper@Sun.COM (Chuck McManis)
Newsgroups: comp.sys.amiga
Subject: Re: "baud" == "bits"/"second"
Message-ID: <114038@sun.Eng.Sun.COM>
Date: 6 Jul 89 20:33:42 GMT
References: <42300@bbn.COM>
Sender: news@sun.Eng.Sun.COM
Reply-To: cmcmanis@sun.UUCP (Chuck McManis)
Organization: Sun Microsystems, Mountain View
Lines: 27

If I may be so bold to put in another "bit" of information :

As most of the followups pointed out so far, baud != Bits/Second.
Instead baud = "signalling transitions" and BPS = data bits crossing
a given interface over time. One of the more subtle points that was
missed was that on regular serial interfaces, something like 9600 baud
is *not* 9600 bits/second. This is due to the fact that the "start bit"
and "stop bit" in the serial stream are often defined as "bauds" but 
not as "bits". (You don't see them in the resulting byte of information,
although you can and do reconstruct them to send them back out over 
the line). This explains why the maximum throughput of a 9600 baud line
is 960 ASCII characters per second, and yet since characters have 8
bits in them (7 + parity) that yields only 7680 bits/second of actual
data. The entire mess is further cloudied by the fact, as was mentioned
in the followups, that things like 2400 baud modems, may communicate with
the computer at 2400 baud, but talk to each other at 600 baud. So how
could that be? The answer is that a baud is not constrained to a simple
binary value like a bit is. Rather, these modems use quadrature encoding
(or something like that) which gives each baud one of four different 
values, allowing each baud to contain two bits. 

Fun isn't it?

--Chuck McManis
uucp: {anywhere}!sun!cmcmanis   BIX: cmcmanis  ARPAnet: cmcmanis@sun.com
These opinions are my own and no one elses, but you knew that didn't you.
"A most excellent barbarian ... Genghis Kahn!"