Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!thunder.mcrcim.mcgill.edu!snorkelwacker.mit.edu!world!decwrl!netcomsv!gandrews
From: gandrews@netcom.COM (Greg Andrews)
Newsgroups: comp.dcom.modems
Subject: Re: Tuning slip on a T2500/NextStation
Message-ID: <1991Jun19.152509.2214@netcom.COM>
Date: 19 Jun 91 15:25:09 GMT
References: <231.285ED74C@zswamp.uucp>
Organization: Netcom - Online Communication Services  UNIX System {408 241-9760 guest}
Lines: 49

In article <231.285ED74C@zswamp.uucp> root@zswamp.uucp (Geoffrey Welsh) writes:
>In a letter to All, Booker Bense (benseb@grumpy.sdsc.edu ) wrote:
>
> >- A back of the envelope calculation reveals:
>
> > (19200 / 8 ) = 2400 bytes/sec * (2/3 data/tcp packet length 
> >) = 1.6 Kbytes /sec ftp transfer rate.
>
>   V.32 "back of the envelope calculation": 9600 bps / 8 = 1200 bytes/sec 
>(neglect MNP or LAP-M overhead) * 2/3 TCP data/packet = 800 CPS.
>

Sorry guys, but you can't divide by eight.  Async data has TEN bits per
byte.  A start bit, eight data bits, and a stop bit.  It makes no difference
whether the modems strip start bits and stop bits across the phone line.
When the data is exchanged between the modem and computer, it has all the
start and stop bits reattached.

PEP's data throughput on a clean line is 14400 bps, rendering 1440 cps
throughput without data compression.  V.32 without error correction will
give you 960 cps throughput max.  V.32 with error correction will give you
around 1100 cps throughput without data compression.  V.32bis without error
correction will give you 1440 cps max.  V.32bis with error correction should
give you 1650 cps without data compression.  (I've seen figures of 1700-1750
cps given, but I don't know if that was without compression)

>
>Lesson #2: when manufacturers say that the modem is capable of "up to" 19200 
>bps, they mean it has done this on cooked files which don't even faintly 
>resemble real world data.
>

Not true for PEP and V.32bis, I'm afraid.  It only takes a 1.3:1 compression
to boost PEP's 1440 cps up to 1920, and V.32bis with error correction would
need a bit less than 1.2:1.  Those ratios may not be achievable with pre-
compressed data, but most other types can easily be compressed by that much.

V.32 has a harder time of it, requiring that the data be compressible by
about 1.75:1.


>Geoffrey Welsh - Operator, Izot's Swamp BBS (FidoNet 1:221/171)
>root@zswamp.uucp or ..uunet!watmath!xenitec!zswamp!root

-- 
 .------------------------------------------------------------------------.
 |  Greg Andrews   |       UUCP: {apple,amdahl,claris}!netcom!gandrews    |
 |                 |   Internet: gandrews@netcom.COM                      |
 `------------------------------------------------------------------------'