Tighten up webrtc audio tests.

Tighten up webrtc audio tests.

See http://goo.gl/LDae10. I'm trying to improve this algorithm. We used to
check that some value was high enough and that the average was in the right
ballpark. The new algorithm will count falling edges in the peaks instead,
as a means to detect the peaks in the audio signal. The algorithm reliably
yields one peak per second of sampled audio, which is correct because the
audio signal beeps at 1 Hz. To get some margin, I'm sampling four seconds
and checking that I get at least two peaks, which should be a rock solid 
assumption even on slow machines.

BUG=357287, 357626
R=perkj@chromium.org
NOTRY=True
(passed all trybots, just got stuck on a broken win trybot)

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=260878

[perkj_chrome, 2014-03-28 13:29:17]

4 seconds per test sounds like a lot. Why not increase the  the frequency of the
beeps?

https://codereview.chromium.org/216523003/diff/1/content/test/data/media/peer...
File content/test/data/media/peerconnection-call.html (right):

https://codereview.chromium.org/216523003/diff/1/content/test/data/media/peer...
content/test/data/media/peerconnection-call.html:334:
gatherAudioLevelSamples(gSecondConnection, 40, 10, function(samples) {
why is this 40 samples to check for playing while you can decide that its not
playing in just 20? 

Do these test really need to run for 4s? How about sending the beeps more
frequently?

https://codereview.chromium.org/216523003/diff/1/content/test/data/media/webr...
File content/test/data/media/webrtc_test_audio.js (right):

https://codereview.chromium.org/216523003/diff/1/content/test/data/media/webr...
content/test/data/media/webrtc_test_audio.js:31: var maxLevel = 32768;
where to this numbers come frome? What do they mean?  Can you document?

[phoglund_chromium, 2014-03-28 13:55:26]

PTAL

https://codereview.chromium.org/216523003/diff/1/content/test/data/media/peer...
File content/test/data/media/peerconnection-call.html (right):

https://codereview.chromium.org/216523003/diff/1/content/test/data/media/peer...
content/test/data/media/peerconnection-call.html:295:
gatherAudioLevelSamples(gSecondConnection, 40, 10, function(samples) {
Ok, I doubled the beep frequency and halved the gathering times (4 secs -> 2
secs).

https://codereview.chromium.org/216523003/diff/1/content/test/data/media/peer...
content/test/data/media/peerconnection-call.html:334:
gatherAudioLevelSamples(gSecondConnection, 40, 10, function(samples) {
On 2014/03/28 13:29:17, perkj wrote:
> why is this 40 samples to check for playing while you can decide that its not
> playing in just 20? 
> 
> Do these test really need to run for 4s? How about sending the beeps more
> frequently? 
We need more of a safety margin when determining if we are playing. If we sample
over 4 seconds we can expect 4 peaks, but I can imagine we'll only see 3 in some
cases because the sampling lines up badly or we are running on a slow machine.
If we see only zeroes for two seconds we can be pretty sure nothing is coming.

https://codereview.chromium.org/216523003/diff/1/content/test/data/media/webr...
File content/test/data/media/webrtc_test_audio.js (right):

https://codereview.chromium.org/216523003/diff/1/content/test/data/media/webr...
content/test/data/media/webrtc_test_audio.js:31: var maxLevel = 32768;
On 2014/03/28 13:29:17, perkj wrote:
> where to this numbers come frome? What do they mean?  Can you document? 

Done.

[phoglund_chromium, 2014-03-28 14:38:34]

The tests will now analyze audio in two-second intervals rather than 4-second;
check the code if you like.

[perkj_chrome, 2014-03-31 07:52:39]

On 2014/03/28 14:38:34, phoglund wrote:
> The tests will now analyze audio in two-second intervals rather than 4-second;
> check the code if you like.

lgtm

[phoglund_chromium, 2014-03-31 07:59:21]

The CQ bit was checked by phoglund@chromium.org

[commit-bot: I haz the power, 2014-03-31 07:59:27]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/phoglund@chromium.org/216523003/40001

[commit-bot: I haz the power, 2014-03-31 15:50:46]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-03-31 15:50:47]

The commit queue went berserk retrying too often for a
seemingly flaky test on builder android_dbg:
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...

[phoglund_chromium, 2014-03-31 16:22:56]

The CQ bit was checked by phoglund@chromium.org

[commit-bot: I haz the power, 2014-03-31 16:27:33]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/phoglund@chromium.org/216523003/40001

[commit-bot: I haz the power, 2014-04-01 00:15:18]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-04-01 00:15:18]

The commit queue went berserk retrying too often for a
seemingly flaky test on builder android_dbg:
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...

[phoglund_chromium, 2014-04-01 07:02:26]

The CQ bit was checked by phoglund@chromium.org

[commit-bot: I haz the power, 2014-04-01 07:03:01]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/phoglund@chromium.org/216523003/40001

[phoglund_chromium, 2014-04-01 08:20:11]

The CQ bit was checked by phoglund@chromium.org

[commit-bot: I haz the power, 2014-04-01 08:20:37]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/phoglund@chromium.org/216523003/60001

[phoglund_chromium, 2014-04-01 13:31:54]

The CQ bit was checked by phoglund@chromium.org

[commit-bot: I haz the power, 2014-04-01 13:32:08]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/phoglund@chromium.org/216523003/80001

[phoglund_chromium, 2014-04-01 15:44:43]

The CQ bit was unchecked by phoglund@chromium.org

[phoglund_chromium, 2014-04-01 15:45:28]

The CQ bit was checked by phoglund@chromium.org

[commit-bot: I haz the power, 2014-04-01 15:46:23]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/phoglund@chromium.org/216523003/80001

[commit-bot: I haz the power, 2014-04-01 15:47:57]

Change committed as 260878

[Alexander Potapenko, 2014-04-01 16:00:05]

On 2014/04/01 15:47:57, I haz the power (commit-bot) wrote:
> Change committed as 260878

BTW the link in the change description is invalid (already?). Please don't leave
invalid links in commit messages :(

[phoglund_chromium, 2014-04-01 16:10:27]

On 2014/04/01 16:00:05, Alexander Potapenko wrote:
> On 2014/04/01 15:47:57, I haz the power (commit-bot) wrote:
> > Change committed as 260878
> 
> BTW the link in the change description is invalid (already?). Please don't
leave
> invalid links in commit messages :(

Huh that is weird. Sorry about that. Here is the correct link:
https://chromium.googlecode.com/issues/attachment?aid=3572870003000&name=plot....

[rlarocque, 2014-04-01 19:05:43]

A revert of this CL has been created in
https://codereview.chromium.org/220403006/ by rlarocque@chromium.org.

The reason for reverting is: This might have resulted in some flaky tests.  This
failure happened two builds after the patch landed.

http://build.chromium.org/p/chromium.chromiumos/builders/Linux%20ChromiumOS%2...

See bug for more details.

BUG=358759.