I work for an Avaya and Cisco business partner certifying networks for
IP-Tel.

These companies recommend G.711 for LAN and G.729 for WAN. 

The testing that I do shows that in a “perfect network” i.e. virtually no
losses, G.711 renders a MOS (Mean Opinion Score) of 4.4 and G.729 renders
4.1. I don’t think the human ear can really tell the difference.

When Voicemail is involved, Avaya best practices says to use G.711 because
the repeated application of the algebraic compression algorithm causes loss
of voice quality. Avaya says to use G.711 for voicemail, even over a WAN. I
don’t have Cisco’s position on that.

 

 

  _____  

From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Andreas Fink
Sent: Saturday, February 10, 2007 12:17 PM
To: [EMAIL PROTECTED]; Community support list for Wireshark
Subject: [Norton AntiSpam] Re: [Wireshark-users] Help. I do not know much
about anything.... Iamtrying to see if a wireless connection between 2

 

 

On 10.02.2007, at 17:44, Frank Bulk wrote:





Andreas:

 

On what basis do you say that most modern IP phones use G.729?  Is there a
certain class of IP phones (PacketCable, Vonage, 8x8, enterprise (Cisco,
Avaya, etc), VoFi) that you had in mind?

 

Frank

 

 

G.729 is one of the best codecs when it comes to efficiency and providing
excellent quality. For software implementations its a bit problematic
because of patent issues. The hardware versions don't have this because you
simply buy chips which do G.729 and youre done because the chips vendor has
a license for G.729 and they usually do much bigger volumes than a small
software company developing a soft phone.

 

If you compare the different codecs out there:

G.726 ADPCM is also good but doesnt compress so much so uses quite a bit of
bandwith. Perfect choice for "a little bit of compression but not too much".

G.711 dont compress and waste a lot of bandwith (80kbps)

GSM codec is popular on soft phones and open source because its free to use
(even there are patents about it).

G.728 is not so good in quality and not so efficient than G.729 but less CPU
intensive.

G.723 is the one which compresses most (as low as 5.3kbps without IP
headers) but its clearly audible and quality is not considered very good.
But useful for applications where minimizing the  bandwidth is more
important than quality.

 

>From the experience with IP phones, especially WiFi IP phones, we learned
that those people who care about quality and provide phones which actually
work (there's unfortunately a lot of crap out there too), they usually
implement G.726, G.729, GSM and of course G.711 a-law and G.711 µ-law for
the guys with tons of bandwidth available.

 

The Cisco IP phones (Desk phones) as far as I know do G.711 and G.726. UTS
Starcom does G.711, G726 and G.729. The Hitachi WiFi Phones WIP3000 and
WIP5000A do G.711, G.729 and GSM.  Asterisk, the open source PABX does
G.711, G.729 (not for free), GSM and I think also G.726 and a few other
codecs. The Grandstream phone adapters also do G.729

 

I wont mention any "fake" wireless IP phones like the wireless Skype phones.
Those are nothing else than a microphone and a loudspeaker connected to the
computer and all the processing is done on the computer. So you get whatever
codec the soft phone has and the phone is just dumb. I prefer real WiFi
phones like the Hitachi Cable WIP 5000A which you can use on any accesspoint
and just work. And I prefer G.729 because its excellent quality.

 

Just my personal opinion...

 





 

Andreas Fink

Global Networks Schweiz AG

---------------------------------------------------------------

Tel: +41-61-6666333 Fax: +41-61-6666331  Mobile: +41-79-2457333

Address: Clarastrasse 3, 4058 Basel, Switzerland

E-Mail:  [EMAIL PROTECTED]

Homepage: http://www.global-networks.ch

---------------------------------------------------------------

ICQ: 8239353 MSN: [EMAIL PROTECTED] AIM: smsrelay Skype: andreasfink

Yahoo: finkconsulting SMS: +41792457333

 

Meet us in Barcelona 3GSM World Congress 2007: Hall 2-1 Stand 2.1F83

 

 





 

_______________________________________________
Wireshark-users mailing list
Wireshark-users@wireshark.org
http://www.wireshark.org/mailman/listinfo/wireshark-users

Reply via email to