Chrome M40 WebRTC Release Notes.

1967 views
Skip to first unread message

Vikas Marwaha

unread,
Dec 12, 2014, 6:53:11 PM12/12/14
to discuss...@googlegroups.com

M40


Main things to pay attention to:

  • The G.722 codec is now supported in the WebRTC stack and Chrome. This is useful for wideband audio in low-compute environments.

  • BlackMagic video capture devices are now supported in Chrome for Mac OS X.

  • VP8 HW decode is now supported in Chrome for Android on Samsung Exynos devices running KitKat and later.

  • The Android AppRTCDemo client now uses WebSockets for signaling. Previous versions of AppRTCDemo will stop working in the near future.

  • Now possible to build 64-bit builds for webrtc on iOS.


Features:



  • Issue 3898: Added support for 64-bit builds on iOS for webrtc.


  • Issue 3937: (Sample Apps) Switched Android AppRTCDemo to use websockets signaling instead of the channel API.

  • Issue 3939: (Sample Apps) Improvements to user experience for Android AppRTCDemo.


  • Issue 3953: (Sample Apps) Added settings screen to AppRTCDemo with options to choose video resolution, display type, GAE server URL, camera fps.


  • Issue 369796: (Chrome-specific) Performance improvements in getStats, setRemoteDescription and setLocalDescription as they no longer run on Chrome’s UI thread.


  • Issue 408089: (Chrome-specific) Switched the video stack on Mac over to AVFoundation (for Lion and up).


  • Issue 421232: (Chrome-specific) Enabled hardware VP8 decoder on Samsung devices with Android 4.4 and above.


  • Improved handling of FEC packets, avoid queuing FEC packets which are not to be used into NetEq packet buffer.


  • Default Opus target bitrate adjusted to match maxplaybackrate; if maxplaybackrate set to a lower value (8000 Hz, e.g.), the bitrate will be set accordingly.


  • Improved pacing for screenshare (and regular video), resulting in lower send-side delay.


  • Added support for adaptive denoiser for VP8.



Bugs:


  • Issue 3951: Fixed problem in SDP for G722 codec, it is now properly advertised as 8Khz instead of 16Khz.


  • Issue 3279: (Sample Apps) Fixed problem, the volume buttons in AppRTCDemo were not affecting output audio volume for OpenSL.


  • Issue 3936: (Sample Apps) Fixed problem with busted aspect ratio in AppRTCDemo on Android when rotating to portrait mode.


  • Issue 430001: (Chrome-specific) Fixed flaky renderer crashes in WebRTC on Nexus 5.


  • Issue 428706: (Chrome-specific) Fixed one way audio issues on Mac.

  • Issue 347371: (Chrome-specific) Fixed support for Black Magic Devices (BMD) on Mac OS X. Fixed by integrating the BMD SDK in Chromium.


/Vikas

Iñaki Baz Castillo

unread,
Dec 12, 2014, 7:30:27 PM12/12/14
to discuss...@googlegroups.com
May I know why changes in the AppRTCDemo demo application are
announced as part of Chrome M40?
Just wondering.

2014-12-13 0:53 GMT+01:00 'Vikas Marwaha' via discuss-webrtc
<discuss...@googlegroups.com>:
> --
>
> ---
> You received this message because you are subscribed to the Google Groups
> "discuss-webrtc" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to discuss-webrt...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.



--
Iñaki Baz Castillo
<i...@aliax.net>

Philipp Hancke

unread,
Dec 12, 2014, 11:33:27 PM12/12/14
to discuss...@googlegroups.com
2014-12-12 16:29 GMT-08:00 Iñaki Baz Castillo <i...@aliax.net>:
May I know why changes in the AppRTCDemo demo application are
announced as part of Chrome M40?
Just wondering.

Because lots of people read these emails. Unless you're volunteering to help writing separate release notes (warning: you might end up with more work when complaining).

oh btw: adapter.js shims MediaStreamTrack.getSources for Firefox since last week. See https://github.com/GoogleChrome/webrtc/pull/272
If you're using adapter.js this will make your life easier as it hides another difference between Chrome and Firefox
Reply all
Reply to author
Forward
0 new messages