Add aec-refined-adaptive-filter command line switch

Add aec-refined-adaptive-filter command line switch

The acoustic echo canceler (AEC) in WebRTC sometimes end up in bad
states. A new tuning was added to WebRTC, and should be tested on
selected platforms. This CL adds a command line switch to turn this
new tuning on or off, facilitating the testing.

Tuning added to WebRTC: https://chromium.googlesource.com/external/webrtc.git/+/0332c2db39d6f5c780ce9e92b850bcb57e24e7f8

Tuning rolled into Chrome: https://crrev.com/8cf9068d949adc9e5e5f13346f776520cbb29c4d

BUG=603411, webrtc:5777, webrtc:5778
Committed: https://crrev.com/8ed6849361610a82033e269f1fb8fe17a179c6e6
Cr-Commit-Position: refs/heads/master@{#388078}

[hlundin-chromium, 2016-04-18 08:26:28]

Description was changed from

==========
Add aec-refined-adaptive-filter command line switch

The acoustic echo canceler (AEC) in WebRTC sometimes end up in bad
states. A new tuning was added to WebRTC, and should be tested on
selected platforms. This CL adds a command line switch to turn this
new tuning on or off, facilitating the testing.

Tuning added to WebRTC:
https://chromium.googlesource.com/external/webrtc.git/+/0332c2db39d6f5c780ce9...

Tuning rolled into Chrome:
https://crrev.com/8cf9068d949adc9e5e5f13346f776520cbb29c4d

BUG=603411,webrtc:5777,webrtc:5778
==========

to

==========
Add aec-refined-adaptive-filter command line switch

The acoustic echo canceler (AEC) in WebRTC sometimes end up in bad
states. A new tuning was added to WebRTC, and should be tested on
selected platforms. This CL adds a command line switch to turn this
new tuning on or off, facilitating the testing.

Tuning added to WebRTC:
https://chromium.googlesource.com/external/webrtc.git/+/0332c2db39d6f5c780ce9...

Tuning rolled into Chrome:
https://crrev.com/8cf9068d949adc9e5e5f13346f776520cbb29c4d

BUG=603411,webrtc:5777,webrtc:5778
==========

[hlundin-chromium, 2016-04-18 08:26:28]

hlundin@chromium.org changed reviewers:
+ avi@chromium.org

[hlundin-chromium, 2016-04-18 08:29:17]

hlundin@chromium.org changed reviewers:
+ tommi@chromium.org

[hlundin-chromium, 2016-04-18 08:29:18]

Avi, Tommi,

Please, review this change.

Thanks!

[Avi (use Gerrit), 2016-04-18 13:47:58]

lgtm

[hlundin-chromium, 2016-04-18 19:10:28]

hlundin@chromium.org changed reviewers:
+ aluebs@chromium.org
- tommi@chromium.org

[hlundin-chromium, 2016-04-18 19:10:29]

aluebs, please review content/renderer/media/media_stream_audio_processor.cc.

Thanks!

[tommi (sloooow) - chröme, 2016-04-18 19:45:42]

tommi@chromium.org changed reviewers:
+ tommi@chromium.org

[tommi (sloooow) - chröme, 2016-04-18 19:45:42]

lgtm

[hlundin-chromium, 2016-04-18 20:49:39]

The CQ bit was checked by hlundin@chromium.org

[commit-bot: I haz the power, 2016-04-18 20:50:09]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1894963002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1894963002/1

[aluebs-chromium, 2016-04-18 20:52:49]

lgtm % with some small questions

https://codereview.chromium.org/1894963002/diff/1/content/renderer/media/medi...
File content/renderer/media/media_stream_audio_processor.cc (right):

https://codereview.chromium.org/1894963002/diff/1/content/renderer/media/medi...
content/renderer/media/media_stream_audio_processor.cc:117: return
base::CommandLine::ForCurrentProcess()->HasSwitch(
Just out of curiosity, does this kind of switches allows you to run a Finch
experiment? Or isn't that the plan?

https://codereview.chromium.org/1894963002/diff/1/content/renderer/media/medi...
content/renderer/media/media_stream_audio_processor.cc:529: }
Do you want to rely on the default WebRTC behavior when the switch is not
specified? Because you could force it to false in the else-clause depending on
the behavior you find better.

[commit-bot: I haz the power, 2016-04-18 23:39:56]

Description was changed from

==========
Add aec-refined-adaptive-filter command line switch

The acoustic echo canceler (AEC) in WebRTC sometimes end up in bad
states. A new tuning was added to WebRTC, and should be tested on
selected platforms. This CL adds a command line switch to turn this
new tuning on or off, facilitating the testing.

Tuning added to WebRTC:
https://chromium.googlesource.com/external/webrtc.git/+/0332c2db39d6f5c780ce9...

Tuning rolled into Chrome:
https://crrev.com/8cf9068d949adc9e5e5f13346f776520cbb29c4d

BUG=603411,webrtc:5777,webrtc:5778
==========

to

==========
Add aec-refined-adaptive-filter command line switch

The acoustic echo canceler (AEC) in WebRTC sometimes end up in bad
states. A new tuning was added to WebRTC, and should be tested on
selected platforms. This CL adds a command line switch to turn this
new tuning on or off, facilitating the testing.

Tuning added to WebRTC:
https://chromium.googlesource.com/external/webrtc.git/+/0332c2db39d6f5c780ce9...

Tuning rolled into Chrome:
https://crrev.com/8cf9068d949adc9e5e5f13346f776520cbb29c4d

BUG=603411,webrtc:5777,webrtc:5778
==========

[commit-bot: I haz the power, 2016-04-18 23:39:58]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-04-18 23:41:12]

Description was changed from

==========
Add aec-refined-adaptive-filter command line switch

The acoustic echo canceler (AEC) in WebRTC sometimes end up in bad
states. A new tuning was added to WebRTC, and should be tested on
selected platforms. This CL adds a command line switch to turn this
new tuning on or off, facilitating the testing.

Tuning added to WebRTC:
https://chromium.googlesource.com/external/webrtc.git/+/0332c2db39d6f5c780ce9...

Tuning rolled into Chrome:
https://crrev.com/8cf9068d949adc9e5e5f13346f776520cbb29c4d

BUG=603411,webrtc:5777,webrtc:5778
==========

to

==========
Add aec-refined-adaptive-filter command line switch

The acoustic echo canceler (AEC) in WebRTC sometimes end up in bad
states. A new tuning was added to WebRTC, and should be tested on
selected platforms. This CL adds a command line switch to turn this
new tuning on or off, facilitating the testing.

Tuning added to WebRTC:
https://chromium.googlesource.com/external/webrtc.git/+/0332c2db39d6f5c780ce9...

Tuning rolled into Chrome:
https://crrev.com/8cf9068d949adc9e5e5f13346f776520cbb29c4d

BUG=603411,webrtc:5777,webrtc:5778

Committed: https://crrev.com/8ed6849361610a82033e269f1fb8fe17a179c6e6
Cr-Commit-Position: refs/heads/master@{#388078}
==========

[commit-bot: I haz the power, 2016-04-18 23:41:13]

Patchset 1 (id:??) landed as
https://crrev.com/8ed6849361610a82033e269f1fb8fe17a179c6e6
Cr-Commit-Position: refs/heads/master@{#388078}

[hlundin-chromium, 2016-04-19 06:13:36]

Post-commit answers to Alex's questions.

https://codereview.chromium.org/1894963002/diff/1/content/renderer/media/medi...
File content/renderer/media/media_stream_audio_processor.cc (right):

https://codereview.chromium.org/1894963002/diff/1/content/renderer/media/medi...
content/renderer/media/media_stream_audio_processor.cc:117: return
base::CommandLine::ForCurrentProcess()->HasSwitch(
On 2016/04/18 20:52:49, aluebs-chromium wrote:
> Just out of curiosity, does this kind of switches allows you to run a Finch
> experiment? Or isn't that the plan?

The plan is to maybe run a Finch experiment, but not at this stage. For a Finch
experiment, you have to query base::FieldTrialList, I believe.

https://codereview.chromium.org/1894963002/diff/1/content/renderer/media/medi...
content/renderer/media/media_stream_audio_processor.cc:529: }
On 2016/04/18 20:52:49, aluebs-chromium wrote:
> Do you want to rely on the default WebRTC behavior when the switch is not
> specified? Because you could force it to false in the else-clause depending on
> the behavior you find better.

Ack.

I'm relying on the default behavior for now.

[aluebs-chromium, 2016-04-19 16:20:51]

https://codereview.chromium.org/1894963002/diff/1/content/renderer/media/medi...
File content/renderer/media/media_stream_audio_processor.cc (right):

https://codereview.chromium.org/1894963002/diff/1/content/renderer/media/medi...
content/renderer/media/media_stream_audio_processor.cc:117: return
base::CommandLine::ForCurrentProcess()->HasSwitch(
On 2016/04/19 06:13:35, hlundin1 wrote:
> On 2016/04/18 20:52:49, aluebs-chromium wrote:
> > Just out of curiosity, does this kind of switches allows you to run a Finch
> > experiment? Or isn't that the plan?
> 
> The plan is to maybe run a Finch experiment, but not at this stage. For a
Finch
> experiment, you have to query base::FieldTrialList, I believe.

Yes, that is what I thought. I was just curious :)

https://codereview.chromium.org/1894963002/diff/1/content/renderer/media/medi...
content/renderer/media/media_stream_audio_processor.cc:529: }
On 2016/04/19 06:13:35, hlundin1 wrote:
> On 2016/04/18 20:52:49, aluebs-chromium wrote:
> > Do you want to rely on the default WebRTC behavior when the switch is not
> > specified? Because you could force it to false in the else-clause depending
on
> > the behavior you find better.
> 
> Ack.
> 
> I'm relying on the default behavior for now.

Ack