Hi, I've been having a good old play with mod_rtmp, which is a very exciting addition, thanks! The main feedback I get when testing with colleagues is the latency, which seems to be somewhere in the order of 1.5 seconds each way. My testbed is an RTMP client > FS server > PSTN so I could three points where jitter buffers and encoders would be maintained. Ping time between servers is in the order of 50-60ms. Is there a way of controlling the size of the jitter buffer in mod_rtmp to reduce it? jitter_buffer_msec doesn't seem to change anything in this scenario. On another note, it seems the option is rtmp.conf.xml to allow calls without rgistration doesn't seem to enable this option, I always seem to have to get the client to log in. Cheers, Neil. -- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/RTMP-buffer-size-tp6554200p655... Sent from the freeswitch-users mailing list archive at Nabble.com.
Doing some more analysis, the noticable latency only seems to be on the receive side of the Flash control i.e. through the speakers of the PC using the Flash control. From the Flash control out to the PSTN the delay seems reasonably low. I can't compile the flash control from the source provided as I don't have Flex. The only thing noticable I can see if the incomingNetStream.bufferTime which seems to be 0.2 seconds in freeswitch.mxml. Is there any chance of having this configuable? Thanks, Neil. -- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/RTMP-buffer-size-tp6554200p655... Sent from the freeswitch-users mailing list archive at Nabble.com.
we have to come up with a way to have the server send pings to the client and when the round trip is too large send another message to the client to make it restart the audio layer by closing and re-opening the port. there is not much as far as options with this because they do not expose much on the client side. On Thu, Jul 7, 2011 at 1:16 PM, peely <freeswitch@peely.com> wrote:
Doing some more analysis, the noticable latency only seems to be on the receive side of the Flash control i.e. through the speakers of the PC using the Flash control. From the Flash control out to the PSTN the delay seems reasonably low.
I can't compile the flash control from the source provided as I don't have Flex. The only thing noticable I can see if the incomingNetStream.bufferTime which seems to be 0.2 seconds in freeswitch.mxml.
Is there any chance of having this configuable?
Thanks,
Neil.
-- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/RTMP-buffer-size-tp6554200p655... Sent from the freeswitch-users mailing list archive at Nabble.com.
_______________________________________________ Join us at ClueCon 2011, Aug 9-11, Chicago http://www.cluecon.com 877-7-4ACLUE
FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-- Anthony Minessale II FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ Twitter: http://twitter.com/FreeSWITCH_wire AIM: anthm MSN:anthony_minessale@hotmail.com GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com IRC: irc.freenode.net #freeswitch FreeSWITCH Developer Conference sip:888@conference.freeswitch.org googletalk:conf+888@conference.freeswitch.org pstn:+19193869900
Thanks for your response. In the end I dug around and found that Flex has an open source compiler. I had a poke around the NetStreamInfo object which you can create from the NetStream, it does have some useful data and showed up to 300ms of buffered data. Seems that up to twice the specified NetStream.bufferTime can be built up according to Adobe documentation. Whilst I managed to get the latency down a decent amount by capping the buffer with a NetStream.bufferTimeMax, it appears that the lion's share of the latency was flash onwards in my browser! I solved that by executing "export PULSE_LATENCY_MSEC=20" prior to /usr/bin/chromium-browser, this has a considerable effect, certainly on Ubuntu which uses PulseAudio. N. -- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/RTMP-buffer-size-tp6554200p656... Sent from the freeswitch-users mailing list archive at Nabble.com.
do you have a patch for the change you made? On Fri, Jul 8, 2011 at 9:33 AM, peely <freeswitch@peely.com> wrote:
Thanks for your response.
In the end I dug around and found that Flex has an open source compiler.
I had a poke around the NetStreamInfo object which you can create from the NetStream, it does have some useful data and showed up to 300ms of buffered data. Seems that up to twice the specified NetStream.bufferTime can be built up according to Adobe documentation.
Whilst I managed to get the latency down a decent amount by capping the buffer with a NetStream.bufferTimeMax, it appears that the lion's share of the latency was flash onwards in my browser!
I solved that by executing "export PULSE_LATENCY_MSEC=20" prior to /usr/bin/chromium-browser, this has a considerable effect, certainly on Ubuntu which uses PulseAudio.
N.
-- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/RTMP-buffer-size-tp6554200p656... Sent from the freeswitch-users mailing list archive at Nabble.com.
_______________________________________________ Join us at ClueCon 2011, Aug 9-11, Chicago http://www.cluecon.com 877-7-4ACLUE
FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-- Anthony Minessale II FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ Twitter: http://twitter.com/FreeSWITCH_wire AIM: anthm MSN:anthony_minessale@hotmail.com GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com IRC: irc.freenode.net #freeswitch FreeSWITCH Developer Conference sip:888@conference.freeswitch.org googletalk:conf+888@conference.freeswitch.org pstn:+19193869900
Sure, I'll generate one when I get back home. You want it throwing on pastebin? -- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/RTMP-buffer-size-tp6554200p656... Sent from the freeswitch-users mailing list archive at Nabble.com.
you could attach it to here http://jira.freeswitch.org/browse/FS-3368 On Fri, Jul 8, 2011 at 9:51 AM, peely <freeswitch@peely.com> wrote:
Sure, I'll generate one when I get back home.
You want it throwing on pastebin?
-- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/RTMP-buffer-size-tp6554200p656... Sent from the freeswitch-users mailing list archive at Nabble.com.
_______________________________________________ Join us at ClueCon 2011, Aug 9-11, Chicago http://www.cluecon.com 877-7-4ACLUE
FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-- Anthony Minessale II FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ Twitter: http://twitter.com/FreeSWITCH_wire AIM: anthm MSN:anthony_minessale@hotmail.com GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com IRC: irc.freenode.net #freeswitch FreeSWITCH Developer Conference sip:888@conference.freeswitch.org googletalk:conf+888@conference.freeswitch.org pstn:+19193869900
participants (2)
-
Anthony Minessale
-
peely