<HTML>
<FONT SIZE="2" POINTSIZE="11" DEFAULT="SIZE">I'm told that Asterisks wideband capability is exclusively based upon 16 khz sampling. Higher sampling rates, like you might find with CELT, are downsampled for mixing at 16 khz.<BR>
<BR>
My guess is that not everyone will be happy with this, especially vendors trumpeting codecs with higher sampling rates. To handle their output will be more CPU intensive.<BR>
<BR>
I'm not an Asterisk insider. I just listen closely when the real insiders speak up ;-)<BR>
<BR>
Michael<BR>
<BR>
--Original Message Text---<BR>
<B>From:</B> Doken, Serhad<BR>
<B>Date:</B> Thu, 18 Jun 2009 14:15:42 -0700<BR>
<BR>
<BR>
<BR>
Thanks Michael. I guess prior to 1.6.2, Asterisk was downgrading streams to SLIN before mixing and then mixed stream got upgraded to WB. <BR>
<BR>
<BR>
<BR>
My question is, with this release, is Asterisk converting WB codecs to SLIN16 and mix them that way ? That seems to be the logical way to me just wanted an insider expert to confirm/deny that. <BR>
<BR>
<BR>
<BR>
Is this the right list to ask that question/find the right contact before I delve into the code knee deep ? <BR>
<BR>
<BR>
<BR>
Serhad Doken <BR>
<BR>
<BR>
<BR>
<B>From:</B> asterisk-users-bounces@lists.digium.com [mailto:asterisk-users-bounces@lists.digium.com] <B>On Behalf Of </B>Michael Graves<BR>
<B>Sent:</B> Thursday, June 18, 2009 6:14 AM<BR>
<B>To:</B> Asterisk Users Mailing List - Non-Commercial Discussion<BR>
<B>Subject:</B> Re: [asterisk-users] Wideband (G722) MeetMe <BR>
<BR>
<BR>
<BR>
<BR>
<BR>
--Original Message Text---<BR>
<B>From:</B> Doken, Serhad<BR>
<B>Date:</B> Wed, 17 Jun 2009 16:07:12 -0700<BR>
<BR>
<BR>
<BR>
Hi, <BR>
<BR>
I wanted to follow up on this thread about WB support on the MeetMe bridge that is in 1.6.2. Does it only work for G722 or any WB codec ? <BR>
<BR>
I am working with another 16k WB codec that I can transcode to 722 and vice versa. I was curious if the 1.6.2 MeetMe bridge can also mix 722 with any other WB codec natively(without downscaling). <BR>
<BR>
Thanks, <BR>
<BR>
Serhad Doken <BR>
<BR>
While not an expert in Asterisk internal, it seems unlikely that Asterisk is mixing signals in encoded space. It's most likely converting the stream to slin for mixing then encoding back into whatever is most appropriate for each end-point.<BR>
<BR>
Michael <BR>
<BR>
--<BR>
Michael Graves<BR>
mgraves<at>mstvp.com<BR>
http://blog.mgraves.org<BR>
o713-861-4005<BR>
c713-201-1262<BR>
sip:mgraves@mstvp.onsip.com<BR>
skype mjgraves<BR>
fwd 54245 <BR>
<BR>
<BR>
<BR>
</HTML>
<HTML>
<LEFT>
<FONT FACE="Arial" COLOR="#000001" SIZE="2" POINTSIZE="11" DEFAULT="ALL">
--<br>
Michael Graves<br>
mgraves<at>mstvp.com<br>
http://blog.mgraves.org<br>
o713-861-4005<br>
c713-201-1262<br>
sip:mgraves@mstvp.onsip.com<br>
skype mjgraves<br>
fwd 54245<br>
<br>
</HTML>