Once apon a time Ian Smith said, 'Specifics on modem retraining' to Simon Byrnand...



 SB> Can anyone tell me whether it is neccessary to enable

 SB> retraining at both ends of a link to actually enable it, or

 SB> just one end?



 IS> It depends on the modems and chipsets/firmware used, some use quite

 IS> different algorithms for line quality versus channel bitrate

 IS> adjustments.  Some are more or less adjustable by the user, and some

 IS> are not.



Another thing I didnt mention is the protocol. I understand that V34

ALWAYS has retraining enabled (?) and that its only V32/bis that can

have it enabled/disabled.



 IS> In Rockwell-chip modems %E0 both disables local and rejects remote

 IS> retrains, as I understand it.  Noone seems to use that, except perhaps

 IS> on leased lines.

 

I have seen a number of modems, including my old 14.4K Cirrus-Logic based

Dynalink that actually have %E0 in the factory defaults. (AT&F) DOH!



 IS> %E1 accepts and responds to both fallback and

 IS> fallforward requests from the remote modem, but only initiates

 IS> fallback requests when detecting unacceptable signal quality (at a

 IS> desired bit error rate), while %E2 allows the modem to both accept and

 IS> initiate fallforward and fallback requests as necessary.



My old modem didnt have %E2 as a valid setting, but it did have %G1 to

enable fallforward/fallback.



 IS> Other modems allow user setting of acceptable bit error rate

 IS> threshold, action on poor signal quality, the period of detecting

 IS> improved signal quality before renegotiation attemps, etc.  Differing

 IS> revisions of some modems/chipsets may also behave quite differently

 IS> with different types of line impairments.



Yup, there are definately no clear cut answer with anything to do with modems,

people saying things like "oh its a Rockwell modem, it will do such
and such"

is generalising somewhat...



 SB> OR, is it either the calling end, or the receiving end whos

 SB> retrain on/off setting decides whether retraining is

 SB> enabled for that link?



 IS> I can see why test results might surprise.  Many people

 IS> (understandably, given the usual poor standard of technical

 IS> documentation in modem manuals) seem to think that %E1 will stop their

 IS> modem falling forward when requested to do so by a remote (%E2-using,

 IS> in Rockwell parlance) modem, and perhaps avoid using it due to that

 IS> misunderstanding.  It's demonstrably not so.



Interesting. Yes, the manual for my Dyanlink was pretty sparse, and was

basically just a glorified command listing. The modem I had before that,

which was an old Rockwell-based GVC 14.4 Fax/Data had an absolutely excellent

book. It literally had chapters dedicated to explaining things like retraining,

error correction, compression, etc...but I dont have that book anymore :(



 SB> All the modem books I've seen dont say much in the way of

 SB> specifics about this.



 IS> Indeed, had me baffled for ages.  Remote Rockwell-chip modem (older

 IS> Maestro 144FM, as it happens) was then on very poor, quite variable

 IS> lines.  Left to its own devices with %E2 set, it would merrily spend

 IS> most of its time requesting and getting rate changes (much faster than

 IS> full retrains), moving next to no data meanwhile.  Many failed

 IS> connects, much cash being wasted.

 IS> Using %E1 on the 'problem' modem allowed my modem to control rate

 IS> adjustment attempts, keeping line speed down and successful throughput

 IS> rates up, while allowing fallforward when, as often happened, the line

 IS> was seen to improve later on in the call, allowing shifting up to 12k

 IS> or better from as low as 4800 sometimes.  The line concerned is now

 IS> very good, but not all are ..



Hmm, so what setting do you recommend in general? Or is there no "Best"

setting?



 IS> Keeping a (local, at least) connection healthy matters much more than

 IS> top speed, at least when you're interested in minimising cost.  Other

 IS> people out on poorer lines have found %E1 at their end greatly

 IS> improving connectivity, even with some assorted more recent Rockwell

 IS> gadgets, to here anyway ..



Thanks for your informative reply, Ian.



Regards,

Simon



... Some people will believe anything if you whisper it to them.

--- FMail/386 1.02