BL> The dB level figure is the amplitude of what comes back when the
BL> Courier sends out the frequency code and the other modem responds.

PE> Well how do we measure line noise then?

BL> That's the problem... we don't. Line noise is the great unknown.

Its listed in the stats Bob. Poor old Bob.

PE> Why have you people been saying I have noisy lines based
PE> on the dB level when instead you should be saying that I
PE> have weak repeaters on my STD link, or something like that.

BL> Yes... or simply a long run of twisted pair,
BL> or poor connections, or lots of things wrong.

You do/nt even KNOW there is anything wrong with PAULS line, in fact
the USR/USR data indicates that Bill has a MUCH worse line than Paul.
Which would certainly go a long way towards explaining why only he saw
handshaking failures when calling the Spirit with a V32bis Sportster.

Corse there is a small possibility that the Courier/Sportster stats are
dud too, just like we KNOW the Courier/M34F stats are. Tho Bills evidence
on that does make it likely that Courier/Sportster stats are valid.

BL> It is very unlikely that there will be
BL> anything wrong with the STD link itself.

Yeah, the world has changed forever on that now with fibre optic links.

BL> I am amazed that the USR is so good that it
BL> can measure levels at various freqiencies...

ALL V34 modems MUST be able to do that, thats what V34 is all about
Bob. Currently only the USRs make it available to the user TOO.

BL> but it is *very* hard to make a meaninglul measurement of line noise.

Its not actually that vital to do so with V34. The variation
in loss with frequency is MUCH more important. You can measure
it well enough as it affects a modern trellis protocol.
@EOT:

---