[Telepathy] change speexenc settings for empathy

[Alexey Fisher]

Am Samstag, den 09.04.2011, 10:42 -0400 schrieb Olivier Crête:
> On Sat, 2011-04-09 at 10:33 -0400, Olivier Crête wrote:
> > So the way to keep on receiving data it to send out packets
> > continuously, even if there is no actual data to be sent. It does waste
> > bandwidth, but for now it's the best we can do.. Unless someone can
> > figure out a way to reliably disable it while in a call.
> 
> I guess the way to do that is to use /dev/network_latency

Whole WMM concept is about stream tagging. I assume: if we tag steam as
voice we will also get lo latency.

part from this page http://www.dd-wrt.com/wiki/index.php/WMM :
"AC_VO queue is laso used for WMM Power Save, also called Unscheduled
Automatic Power Save Delivery U-APSD
http://en.wikipedia.org/wiki/IEEE_802.11e-2005#APSD. Only IP packets
with ToS value 0xb8 DSCP Expedited Forwarding are buffered by AP, and
waiting for other RTP packets from STA to be released. "

I tried:
... ! udpsink qos=1 qos-dscp=8 port=...

And it looks like my dd-wrt tagged the stream as voice. At leas
wireshark recognize it. I still can't confirm if driver recognize it or
not.

-- 
Regards,
        Alexey