RS> Yeah, its definitely the rather pronounced slope its got that is
RS> the problem when you call someone like Paul with an average line.

BG> Yeah, the line just runs out of bandwidth at the top end.

Its a bit more complicated than that, thats why these are all on.

Nonlinear Encoding       ON/ON
Precoding                ON/ON
Shaping                  OFF/ON
Preemphasis     (-dB )   4/4

BG> With Dave's lines, it's enough to allow consistent connects of
BG> 33600 or 31200, but to Paul is a different matter, with many calls
BG> barely making it at 26400.  I also note that Dave has polled Paul
BG> and seen the dreaded CONNECT 28800/NONE which I saw here quite
BG> frequently until upgrading the EPROM.  No idea why either.

Smells rather like a USR code wart, evidenced by the massive
change you got with a USR code change on that symptom.

RS> Your 7db on your local loop isnt the cause of your problems to Paul.

BG> I'm actually starting to think that Paul's NetComm
BG> isn't quite up to scratch either, especially now
BG> that Dave is seeing the 28800/NONE string himself.

Thats mad, particularly when your rom change
eliminated it completely, cant be Pauls Netcomm.

BG> Remember that his Courier would connect perfectly to mine every time.

All that proves is that the USR code does better with another USR on
the other end of the line. Thats almost ALWAYS true, largely because
that config is always much more thoroughly tested by the manufacturer.

RS> You also now have a problem that this call with a Courier at your
RS> place is completely different to what you get with the Sportster
RS> V34+ WITH that classic shift of the entire curve too. Urk.

BG> Dunno, I'll do some stats later and post them with the others below,
BG> just in case something has changed here over the past month or so. 

Yeah, although its an odd sort of change, there arent
to many things that Telstra is likely to have done which
would produce a change that looks like that in the curve.

BG> Damned if I can see why the Courier should
BG> measure any differently from the Sportster though.

Sure, it certainly shouldnt, but it clearly does with a Rockwell
on the other end and almost as certainly with a Motorola too, so its
quite possible there is some significant wart in the ATY11 stats yet.

RS> Corse you really should repeat the test today with both a Sportster
RS> V34+ and a Courier on your line, calling Dave, to eliminate the
RS> possibility that Telstra has changed something about your line.

BG> Yep, my thoughts exactly.  I'll do it first thing tomorrow morning.

Dunno, I meant using BOTH a Courier and your Sportster on the same line
at your place to call Dave. Maybe you mean your new Courier has arrived.
@EOT:

---