[Asterisk-doc] How to use the Asterisk IRC Channel Politely

[Leif Madsen]

On Sun, 2003-12-21 at 22:07, Leif Madsen wrote:
> On Sat, 2003-12-20 at 23:12, Peter Grace wrote:
> > Guys,
> > 
> > ~    Check this out and let me know what you think:
> > 
> > *How to Politely Use the Asterisk IRC Channel*

Oops.. forgot to paste!

Olle,

I'm not sure who the original author was, so this reply might not be
reaching the appropriate person.

Comments:
1. The text is excellent and accurate. If I read this as a newbie, it
creates some questions in my mind that should probably be addressed in
the text up front. I'll try to address those below. Some of this might
be better added as an appendix (or whatever) to the original document.

2. The average * user probably does not have a clue why these settings
vary from one implementation to another. Therefore include something
like:

"Echo cancellation settings are directly related to the pstn line in use
at your location. For example, the distance between the * machine and 
the pstn central switching office determines the amplitude of the voice
signal in both directions. If asterisk is close to the central office, 
the settings for rxgain and txgain will be different then if asterisk 
were further away from the pstn office. Therefore, one person may get
better results with rxgain=-3 and txgain=-3, while another asterisk
implementation may require rxgain=0 and txgain=0. The ztmonitor 
application should be considered an aid to adjust rxgain and txgain,
with initial changes limited to increments of 3."

3. Changes to the rxgain and txgain settings "require" a complete
restart
of asterisk and not just a simple reload. Therefore, the changes will be
disruptive to asterisk users. (Note: lots of folks on the list have
been caught doing simple reloads; it doesn't work.)

4. Another element that impacts how effective any echo cancellor will
work, whether the canceller is part of asterisk or built into channel
banks, is the "quality" of the analog pstn cabling. For example, if the
inhouse pstn analog wiring connecting to the asterisk system is passing
over the top of a florescent lighting fixture, electrical noise from
the ballast in the fixture will be injected into the analog wiring in
such a way as to fool the echo canceller into using incorrect settings.
Inhouse pstn wiring that passes next to a PC video monitor has been
known to inject such noise. One asterisk user found a high quality 
analog telephone that was bridged onto the analog pstn line was actually
defective, injecting power line noise (hum) onto the pstn line causing 
the asterisk echo canceller to be ineffective. A simple test is to 
connect the asterisk box directly to the inhouse telephone entrance 
block and disconnect "all" other inhouse wiring. If the echo disappears 
after making the adjustments noted in this document, then suspect wiring
and/or bridged analog telephone devices as the route-cause for
generating 
noise. (The electical noise that fools an echo canceller may be such 
that a normal user cannot actually hear that noise under normal 
conditions. Test equipment does exist that can measure the noise,
however
the good testing equipment generally costs more then $500 US.)

5. One advanced asterisk user found that reversing the tip & ring wires
of the pstn analog line where it attaches to the asterisk system cleared
his echo problem.  It is highly likely this user had poor inhouse analog
wiring where noise was being injected from some unknown source, and 
reversing the tip & ring fooled the echo canceller into using settings 
that were more effective for his users. Locating the source of the noise
and correcting that problem would be a better choice then simply
assuming 
the line reversal corrected the problem.

6. Finally, it should be noted that echo problems "can" be caused by
the digital SIP phones although these cases are rare in todays SIP
world. 
Two such cases are:
a. the audio sounds heard in the SIP phone handset will sometimes travel
through the handset handle and be picked up by the microphone within the
handset. This might be the case should one find that a particular
vendor's SIP phone has echo when another vendor's SIP phone does not
have echo. One simple resolution is to disassemble the handset and stuff
foam in the handset handle passageway (where the earpiece wiring passes)
and around the earpiece and microphone assembles, reducing the audio
coupling between the earpiece and the microphone.
b. it is also possible for one vendor's SIP phone to have electronic
transmission levels within the phone at much higher sound levels then 
another vendor's SIP phone. This might be the case if "all" of vendor-A 
SIP phones have echo while all of vendor-B SIP phones have no echo.
(Some vendor's SIP phones have adjustable sound level settings that can
be used to correct this situation.)

Hope the above helps in some way....

Rich
radamson@routers.com