Thanks for the hint on dtls_srtp. I once found this the hard way by debugging, but just couldn’t remember it this time. Maybe there could be a better error message than “no common audio codecs found”? After 2 days of fiddling, i was able to make a local WebRTC connection from Firefox to baresip. :D

I’m very keen on getting this to work, but i have 2 questions regarding the SIP protocol:

1) What is the minimum viable SIP exchange that baresip needs to happily make a connection?

I currently send an INVITE from the browser to baresip, get back 180 Ringing (ignoring that for now), then get receive the 200 Answering response.

That seems to be enough for baresip to start the session, but of course it expects an ACK, keeps re-sending the Answer, and times out after a while. But is this the only remaining piece in the SIP signalling communication?

2) How to send the ICE-Trickle candidates from the browser to baresip?

I worked around this by just waiting for the ICE gathering to finish in the browser, before sending the initial SDP Offer to baresip. What would be the correct way to deliver extra ICE candidate through SIP? is it feasible to do that without touching the SIP/SDP too much (i’m still trying to get away without a SIP client library for parsing, etc…)

The SDP Offer from the Chromium browser now contains this line: a=rtcp:9 IN IP4

baresip trys to connect through RTCP correct right IP, on Port 9, which obviously fails. Any idea why Chromium chooses this weird port number?

Firefox just refuses to make any connection to the websocket.

I’m trying to get away without using a SIP library, just

On 07.04.2017 18:33, Alfred E. Heggestad wrote:

On 07/04/17 04:26, Jonathan Sieber wrote:

I found the baresip_ws branch, respectively re/sip_websock, and was able to make a WebSocket connection from Chromium into baresip code work.

There seems to be a race condition where libre sometimes doesn’t “find” the incoming websocket, and starts a HTTP WS request to the remote address with ‘/path’

But eventually, the INVITE will arrive, but baresip replies with 488 Not Acceptable, because have_common_audio_codecs() returned false.

I learned something about the sdp parsing code, trying to find out why there wouldn’t be any codec candidates left in the audio stream object, but i’m entangled in unknown data structures and can not find out where exactly the sdp code rejects the candidates

What was the current state of the websocket branch(es) as a server, anyway? It’s seems like it’s really close to work and do something really cool with it… ;D

Hi Jonathan,

for webrtc interop the and modules must be loaded and used. You need to use the same RTP profile as the webrtc endpoint you are talking to, most likely UDP/TLS/RTP/SAVPF.

regarding SIP over Websockets, I did some work on this earlier but it is far from complete. I am not planning to work on this any time soon, so if we want to add support for this we need to ask the community for help.


SIP2WebRTC Bridge

Call me on

Mirror to or

EWindow Browser Access

Though this is beyond the functionality of a static window, we want interaction between the window and users visiting the website.

WebRTC Client


Set up HackMD! -> FreeBSD PhantomJS npm issue ->


Before i go home, i need to take the Web Client with me. Call me on


Found some trivial example code “WebRTC-Example” on github Problems with baresip connectivity

WebRTC // commercial product // doesnt support p2p ??


“Destination Address Required”

-> System can not send a packet to this address Possible Reasons:

Check for

net_interface          tap0

in .baresip/config

Connection stops after short time:

Video is not Fullscreen

Try software scaler module… module


Starting of Sylk, (commercial service provice Get an account on

Figure out audio codecs: {0/TLS/v4} 200 OK (SIP Thor on OpenSIPS XS 1.11) [1 binding] call: alloc with params laddr=, af=AF_INET ice: new session with STUN-server at (username=estar) call: no common audio codecs - rejected server logs are a blessing

G722/8000/1 a=rtpmap:109 opus/48000/2 a=rtpmap:9 G722/8000/1 a=rtpmap:0 PCMU/8000 a=rtpmap:8 PCMA/8000 a=setup:actpass

gmake CC=cc\ -g baresip
lldb37 baresip
-> the variable name reveals, it's actually about video codecs!
--> lldb37 stepped in a trap and led me in account_debug??
gdb baresip
b have_common_audio_codecs

sc = sdp_media_rformat(stream_sdpmedia(audio_strm(call->audio)), NULL);

10 minutes of tedious ack grepping –> re/src/sip/msg.c:media_decode

later: –> ((sdp_media*)m)->rfmtl is empty list

AAAH: —> The real problem is call->audio ->strm->sdp->rfmtl is empty How should it be filled? audio_alloc stream_alloc sdp_media_add media_alloc it’s not there??

—> rfmtl It should be filled later by sdp_decode

ack rfmtl -> b set

GDB hates me: (gdb) print call->audio->strm->sdp->rfmtl Segmentation fault

RESOLUTION: Use mediaenc=srtp instead of dtls_srtp

Actually baresip is saying there are no common encryption schemes, not audio codecs


Network Architecture

Good: Automatic naming by



icegath: {audio.1} STUN Request failed: Operation timed out icegath: {audio.2} STUN Request failed: Operation timed out ice: gather error: Operation timed out (0 ) call: medianat ‘ice’ failed: Operation timed out

—> Check for public v4/v6 / net_interface configu

; found no common audio codecs found -> module

ice: audio.1: no remote candidates found (address =