Xref: utzoo comp.dcom.fax:324 comp.dcom.modems:9885
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!think.com!mintaka!ai-lab!ai.mit.edu!hqm
From: hqm@ai.mit.edu (Henry Minsky)
Newsgroups: comp.dcom.fax,comp.dcom.modems
Subject: Re: Fax modems: internal vs external
Message-ID: <16055@life.ai.mit.edu>
Date: 21 May 91 21:11:43 GMT
References: <1991May21.182406.25973@beach.csulb.edu>
Sender: news@ai.mit.edu
Organization: MIT Artificial Intelligence Laboratory
Lines: 23

In article <1991May21.182406.25973@beach.csulb.edu>, sichermn@beach.csulb.edu (Jeff Sicherman) writes:
|> 
|>   I tend to prefer to get regular modems as external models for safety
|> (to the computer system), light status visibility, easier access and
|> portability, and upgradability of the serial port chip.
|> 
|>   However, I was wondering, given the bandwidth requirements of Fax
|> modems, especially when receiving a fax, is there a performance or
|> reliability penalty in choosing an external fax modem ? (Reliability
|> in the sense of lost image.) Can the modem to computer link be bumped
|> up higher than the fax rate (9600) to handle this ?
|> 
|> Jeff Sicherman

The actual speed of the fax is 9600 baud synchronous, which is faster
than 9600 baud async. This means that you really do need to run the
computer/fax serial line at 19200 in order not to get choked with
data on reception, or have long idle periods at the ends of lines. 

We run 19200 to an EVEREX everfax 24/96 D (model 968-51) from a sun3. 
It is true that at those data rates, you might expect some noise; we
haven't had a noticable problem, but apparently it was enough of a concern that the Class 2 standards committee is holding up the acceptance of the standard until they add some sort of reliable host-to-modem data-link layer to the protocol.