Lays the base work and initial unit tests for renderer side audio mixing! Right

now nothing except unit tests are plumbed since mixing has poor performance with
the large buffer sizes we're currently using;  I.e., for a 140ms clip we attempt
to render the full 170ms buffer, adding 30ms of overhead to the pipeline.

Once another CL introduces resampling we can switch over to the low latency path
for audio and turn on mixing at the same time.  Chris also has some other ideas
on how we might fix this prior to introducing resampling.

Design revolves around each renderer creating an AudioRendererMixerInput which
funnels into a common AudioRendererMixer (based on the AudioParameters each
AudioRendererMixerInput is configured with).  Each AudioRendererMixer owns a
single AudioDevice which output from each AudioRendererMixerInput is mixed into.

As mentioned above, for the initial landing of this CL the code which creates an
AudioRendererMixerInput in RenderAudioSourceProvider has been disabled.  Please
look at https://chromiumcodereview.appspot.com/10636036/ if you're interested in
what this looks like.

BUG=133637
TEST=unit tests. audio_latency_perf. drum machine. manual testing. asan.

Review URL: https://chromiumcodereview.appspot.com/10544130

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@144819 0039d316-1c4b-4281-b951-d872f2087c98

0 files changed, 0 insertions, 0 deletions