Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!usc!sdd.hp.com!spool2.mu.edu!uunet!gammafax!mikes
From: mikes@gammafax.gammalink.com (mike spann)
Newsgroups: comp.dcom.fax
Subject: Re: Fax transmission times
Keywords: fax, timings, benchmarks
Message-ID: <1279@gammafax.gammalink.com>
Date: 25 Jan 91 23:13:13 GMT
References: <1991Jan23.223517.531@uunet!bandw>
Reply-To: mikes@gammalink.com
Distribution: usa
Organization: gammafax
Lines: 52

In article <1991Jan23.223517.531@uunet!bandw> craig@bandw.uunet (Craig Goss) writes:
>Is anyone familiar with a standard method for measuring fax
>transmission times. I have seen advertisements from several 
>manufacturers of fax machines that claim a "x second"
>transmission time.

Their ain't no such thing.

>My guess is that one of the CCITT test documents is used
>to determine the number of seconds for a single page, but
>I have seen nothing which details the reqirements for such
>testing. Is there some science to these claims or are these
>figures just a pipe dream.

The numbers they quote are much like the numbers quoted in benchmarks.  They
are not quite lies, but they often indicate values that are rarely acheived
in real life.  (We have all heard about MIPS and SMIPS..)

The time necessary to send a page of facsimile data is dependant on several
factors including (but not limited to): the content of the image, the data
rate, the compression algorighm and the minimum scan line timing imposed
by the receiver.

A manufacturer will often play with all four factors to achieve very fast
times.  They will pick an image that compresses very well, they will use
a proprietary modulation technique, they will use a proprietary compression
technique and they will eliminate minimum scan line timings.  Needless to
say, these 'NON-CCITT' parameters will not work with any machine OTHER than
another BRAND-X machine, although the BRAND-X machines will default to CCITT
conformance when dealing with other machines.  This is why you will often
see the disclaimer (in tiny print) that the timings are only to another
BRAND-X machine.  Going to another machine will achieve slower transmissions.

This is not to say that these overly optimistic numbers are bad things to
look at.  A branch office that only sends faxes to the main office, or the
reciprocol operation, SHOULD look for a single vendor of fax machines that
have nice proprietary features.  This will reduce operating costs for the
majority of your traffic.  The performance to out-of-house sites will be no
worse than with any other fax machine.

BTW: the RECEIVING fax machine has the final say on how fast, what compression
and what scan line timing to use.  Like they say, a chain is only as strong
as the weakest link.  Having the fastest, most expensive fax machine in the
world won't speed things up when you try to send to an 'oldy-moldy' machine
that someone else has.

mike

-- 
Michael Spann                           mikes@gammalink.com
Voice:  +1-408-744-1430			Fax:    +1-408-744-1549
UUCP:   ...!uunet!gammafax!mikes   	CIS:    73747,441