The event was hosted at Google by Chad Hart, Tsahi Levi, and Chris Koehncke. The schedule was packed with interesting talks from various contributors to the WebRTC ecosystem. The first session was all about user experience, Arin Sime showed a summary of the different user interactions across WebRTC apps. For example: to show the audio volumes when someone speaks (an indicator that someone is speaking and you may not be hearing), be able to move the picture in picture position: drag and drop, or minimize, and to use small icons on desktop and large and spaced apart on mobile.

Emil Ivov, the chief video architect at Atlassian delivered a fast tempo presentation on Selective Forwarding Units (SFU) and the differences between scalable video and simulcast video with a lot of references to various currently ongoing standardization activities.

Philipp Hancke from &yet observed the traffic from various messaging apps that offer video communicaiton (e.g., Whatsapp, Messenger, etc) both on the desktop and native mobile apps. His observations were that the mobile applications were using simple tricks to improve the quality of audio and video, for example varying the sampling rate, packetisation interval, or setting lower bitrates for audio, to trying to improve setup experience by connecting first via TURN relay server and then migrating the session to peer-to-peer. All-in-all a very good talk, peppered with anecdotal references. He’s published most of his findings on webrtchacks. Philipp’s talk was nicely complemented by Rob Brazier from Twilio, who talked about common points of failure and how Twilio manages these issues at scale.

One of the talks by Tim Panton (lego robot demo) and ORTC update (by Trent Johnsen, Bernard Aboba).

Browser Updates

The browsers: Firefox and Chrome made upcoming feature announcements, notable were related to the IP address privacy. They also revealed that about 720 companies currently use webrtc, up from 550 in February. And this year about 30 acquisitions were made (notable: Tropo, Acision).

Serge Lachapelle and Justin Uberti from Google announced forthcoming updates to Chrome, a summary are listed below:

• no more HTTP support for getUserMedia from chrome 48 i.e., in December 2015. Additionally, RSA will be deprecated, ECDSA to be default and support for DTLS1.2 also coming in Chrome. Firefox already supported DTLS.
• A new delay agnostic AEC is coming, which can be enabled by a new constraint: googDAEchoCandellation. It is available everywhere except on OS X. It will come to OS X in mid September. The new echo canceller attempts to recover when it fails, i.e., it will come back on, but it still has some niggles according to Google.
• Faster screen sharing and video capture
• Improved Android and iOS support

Nils Ohlmeier from Firefox announced new features as well, they are summarised below:

• Canvas as input for video capture comes in FF41/43
• Support for ICE over TCP arrives in FF41, Chrome already supports this.
• IPv6 support in FF 42, chrome already supports this.
• Several fixes to improve audio are coming in FF41-44.
• Lastly, moz- prefix to PeerConnection goes away in FF43.

Bernard Aboba from Microsoft and Trent Johnsen from Hookflash presented the vision of WebRTC NV and updates to ORTC. Especially highlighting the expected changes and additional objects that are not available in the WebRTC v1.0.

Addendum: (18/09/2015): Microsoft announced support for ORTC in a preview version of the Edge browser in Windows 10. We are excited to have another browser support WebRTC natively without a plugin. To try a demo on Edge, head over to their Test Drive Demo. See also Philipp Hancke’s post on ORTC support on webrtchacks.

Codec support

Real-time encoding and decoding support for VP9 is coming in Chrome 47. Until now Chrome just decoded VP9 streams (which went live on Youtube earlier this summer). Chrome partnered with Vidyo for SVC, which is to come later. This is interesting as Microsoft announced support for rendering VP9 a week ago.

At the event, Justin Uberti also announced that H.264 support is coming soon to Chrome perhaps in December (interoperability as a Christmas present). Firefox already supports H.264 via openH264, and Bernard Aboba from Microsoft announced that ORTC in Edge will initially support H.264 UC (based on RFC6190) and Opus. While this will create some incompatibility in video interop until all browsers support at least either of H.264 and VP8, however, audio does work smoothly. Finally, the end result is that we have more browsers in the ecosystem support WebRTC.

Summary

Overall there was a sense of excitement within the WebRTC community, especially within the developer and engineering community, in part due to the upcoming enhancements and the fact that the event provided them an opportunity to exchange notes on their deployment issues, etc. The video recordings of the event are available on Youtube.

If you would like a a personal demo of callstats.io or a free trial coupon email us at support-at-callstats.io.