HylaFAX The world's most advanced open source fax server
[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [hylafax-users] USR Courier V. Everything Performance

On 2004.12.21 07:54 "Donald, Adam" wrote:

I performed a garbage upgrade and attached a USR Courier V. Everything
to
hylafax, which has actually managed to receive all pages sent to it so
far
(still using the 13 page test) from the HP T45xi/deskjet FAX.

The Courier V. Everything modems usually have a different chipset in them than the average Sportster model does. They're still can't compare to new MultiTechs or MainPines, or even old Rockwell-chipsetted modems, but they're better than Sportsters.

My problem now is performance.

With the 56K voice and 33.6 fax modems, I could expect to receive all
13
pages (on occasions when errors were absent) in around 3 minutes.

You'll only get 33.6 fax if the modem supports it. Just because a modem supports 33.6 V.34 doesn't mean that it supports that speed in fax mode. From experience I can tell you that Courier V. Everything modems don't. The only analog modems that I know of that do support V.34 are new MultiTechs and MainPines.

With the Courier V. configured as a class 1 device, I can expect 13+
minutes
to receive all 13 pages.

It's probably ECM slowing things down.

With the Courier V. configured as a class 2.0 device, I can expect a
receive
time of just over 5 minutes (still significantly longer than the
sportsters).

The modems don't support ECM in Class 2.0. So, the image is going to come through more quickly, although with errors. With ECM you get perfect image quality, but it takes longer if there are problems. You can disable ECM on your fax machine or in Class 1 HylaFAX with "Class1ECMSupport" in your modem config file.

I have noticed that as a class 1 device, there appears to be some
difficulties encountered with HDLC frames/termination, and perhaps
this is
"fixable".

Here is a typical log from Courier V. configured as a class 1: 

Dec 21 09:01:40.29: [ 6380]: <-- [11:AT+FRM=146\r]
Dec 21 09:01:41.26: [ 6380]: --> [7:CONNECT]
Dec 21 09:01:42.40: [ 6380]: HDLC frame not byte-oriented.  Trailing

(132 more lines of HDLC errors)

Try faxing from a different fax machine. If you get these errors still then something's not right with the Courier V. Everything. If you don't get these errors then I would suspect that this first fax machine is actually dropping out of ECM mode after the first ECM frame. Note that it recovers later when it switches to V.17 12000...

Dec 21 09:02:56.46: [ 6380]: RECV recv CTC (continue to correct)

Dec 21 09:02:56.46: [ 6380]: <-- [9:AT+FRS=7\r]

Dec 21 09:02:56.57: [ 6380]: --> [2:OK]

Dec 21 09:02:56.57: [ 6380]: <-- [9:AT+FTH=3\r]

Dec 21 09:02:56.69: [ 6380]: --> [7:CONNECT]

Dec 21 09:02:56.69: [ 6380]: <-- data [3]

Dec 21 09:02:56.69: [ 6380]: <-- data [2]

Dec 21 09:02:57.91: [ 6380]: --> [2:OK]

Dec 21 09:02:57.91: [ 6380]: RECV send CTR (confirm continue to
correct)

Dec 21 09:02:57.91: [ 6380]: <-- [11:AT+FRM=121\r]

Dec 21 09:03:00.21: [ 6380]: --> [7:CONNECT]

Dec 21 09:03:00.55: [ 6380]: RECV received frame number 0

So either the modulators are incompatible or the fax sender messed up big-time. I would need detailed logs to discern any further. Maybe you can send this fax to me with that fax machine?

Dec 21 09:14:04.20: [ 6380]: RECV received RCP frame

Dec 21 09:14:07.65: [ 6380]: MODEM <Timeout>

Dec 21 09:14:09.65: [ 6380]: MODEM <Empty line>

Dec 21 09:14:11.65: [ 6380]: MODEM <Empty line>

These timeouts are not encouraging. Something is still wrong, it didn't detect the carrier drop.

A typical class 1 log contains 4300 lines, as opposed to 200 lines in
a
class 2 log.

That's normal. You can turn logging down when you're comfortable with things. Class 1 interacts with the faxing on a much lower level, which means that there is more activity going on.

Questions:

1.	Is there something that I can do to address the HDLC errors
(apart
from purchasing another modem) so I can run in class 1?

Disable ECM. ECM is a good thing, though, so we'd need to get to the bottom of the problem I mentioned above to make sense of the stuff any better.

2.	When in class 1, the remote wants 2-D MMR, but in class 2.0
the same
remote wants 1-D MH.

ECM is required to transmit with MMR. So the fact that in Class 2.0 ECM is not supported means that MMR is not supported either.

Does this account for the 2 minute speed
difference
between sportster receives and courier class 2.0 receives?

Kind of. ECM is accounting for it. MMR is actually a tighter compression than MH.

Is there
anything that would remedy this behavior?

Disabling ECM.

Lee.

____________________ HylaFAX(tm) Users Mailing List _______________________
 To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi
On UNIX: mail -s unsubscribe hylafax-users-request@xxxxxxxxxxx < /dev/null
 *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*

Project hosted by iFAX Solutions