Update sha1 files and unittests for Opus v1.1.2

Update sha1 files and unittests for Opus v1.1.2

This was committed as part of
https://codereview.webrtc.org/1657343002/

[flim-webrtc, 2016-01-25 15:00:53]

flim@webrtc.org changed reviewers:
+ minyue@webrtc.org

[flim-webrtc, 2016-01-25 15:00:53]

Here are some changes intended to fix breaking tests when Opus 1.1.2 is upgraded
later in Chromium. A dummy CL was created in [1] with opus v1.1.2 to check that
the tests are passing. Could you please help to review? 

[1] https://codereview.webrtc.org/1612443002/

[minyue-webrtc, 2016-01-25 21:33:12]

On 2016/01/25 15:00:53, flim-webrtc wrote:
> Here are some changes intended to fix breaking tests when Opus 1.1.2 is
upgraded
> later in Chromium. A dummy CL was created in [1] with opus v1.1.2 to check
that
> the tests are passing. Could you please help to review? 
> 
> [1] https://codereview.webrtc.org/1612443002/

LG but how would you keep current bots green?

[minyue-webrtc, 2016-01-25 21:36:41]

one more comment

https://codereview.webrtc.org/1629413002/diff/1/resources/audio_coding/neteq4...
File resources/audio_coding/neteq4_opus_ref.pcm.sha1 (right):

https://codereview.webrtc.org/1629413002/diff/1/resources/audio_coding/neteq4...
resources/audio_coding/neteq4_opus_ref.pcm.sha1:1:
c23004d91ffbe5e7a1f24620fc89b58c0426040f
get neteq_unittest.cc and comment that these files are generated by using Opus
1.2 to decode RTC file generated by Opus 1.1.

[flim-webrtc, 2016-01-26 09:51:28]

Thanks for the review. I'm planning to have this landed only when webrtc is
rolled into chromium with opus 1.1.2 (i.e. when it starts breaking the tests)

https://codereview.webrtc.org/1629413002/diff/1/resources/audio_coding/neteq4...
File resources/audio_coding/neteq4_opus_ref.pcm.sha1 (right):

https://codereview.webrtc.org/1629413002/diff/1/resources/audio_coding/neteq4...
resources/audio_coding/neteq4_opus_ref.pcm.sha1:1:
c23004d91ffbe5e7a1f24620fc89b58c0426040f
On 2016/01/25 21:36:40, minyue-webrtc wrote:
> get neteq_unittest.cc and comment that these files are generated by using Opus
> 1.2 to decode RTC file generated by Opus 1.1.

Done.

[minyue-webrtc, 2016-01-26 14:30:23]

https://codereview.webrtc.org/1629413002/diff/20001/webrtc/modules/audio_codi...
File webrtc/modules/audio_coding/neteq/neteq_unittest.cc (right):

https://codereview.webrtc.org/1629413002/diff/20001/webrtc/modules/audio_codi...
webrtc/modules/audio_coding/neteq/neteq_unittest.cc:571: // The pcm files were
generated by using Opus v1.1.2 to decode the RTC
not only the pcm files but also stats files, right?

[flim-webrtc, 2016-01-26 16:47:06]

On 2016/01/26 14:30:23, minyue-webrtc wrote:
>
https://codereview.webrtc.org/1629413002/diff/20001/webrtc/modules/audio_codi...
> File webrtc/modules/audio_coding/neteq/neteq_unittest.cc (right):
> 
>
https://codereview.webrtc.org/1629413002/diff/20001/webrtc/modules/audio_codi...
> webrtc/modules/audio_coding/neteq/neteq_unittest.cc:571: // The pcm files were
> generated by using Opus v1.1.2 to decode the RTC
> not only the pcm files but also stats files, right?

You're right, thanks for catching this!

[minyue-webrtc, 2016-01-28 09:42:05]

lgtm if you submit it after Opus 1.2 was pulled. and there is one more inline
question

https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
File webrtc/modules/audio_coding/neteq/audio_classifier_unittest.cc (right):

https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
webrtc/modules/audio_coding/neteq/audio_classifier_unittest.cc:64: #if
defined(WEBRTC_ARCH_ARM) || defined(WEBRTC_ARCH_ARM64)
why is only mono affected?

[tlegrand-webrtc, 2016-01-29 14:26:31]

tina.legrand@webrtc.org changed reviewers:
+ tina.legrand@webrtc.org

[tlegrand-webrtc, 2016-01-29 14:26:32]

Mybe Minyue can comment on what it means that you needed to change the bounds
for noise levels in the DTX test?

https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
File webrtc/modules/audio_coding/codecs/opus/opus_unittest.cc (right):

https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
webrtc/modules/audio_coding/codecs/opus/opus_unittest.cc:211: const uint16_t
kOutputValueBound = 8;
It looks like noise during DTX got worse with the new version?

[minyue-webrtc, 2016-01-29 14:46:36]

https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
File webrtc/modules/audio_coding/codecs/opus/opus_unittest.cc (right):

https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
webrtc/modules/audio_coding/codecs/opus/opus_unittest.cc:211: const uint16_t
kOutputValueBound = 8;
On 2016/01/29 14:26:32, tlegrand-webrtc wrote:
> It looks like noise during DTX got worse with the new version? 

I have discussed this with Felicia. Yes, the value increased, which is because
the decaying of CNG becomes slower in Opus 1.2. We have checked the long time
behavior and no noise pumping showed up.

[tlegrand-webrtc, 2016-01-29 14:48:03]

On 2016/01/29 14:46:36, minyue-webrtc wrote:
>
https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
> File webrtc/modules/audio_coding/codecs/opus/opus_unittest.cc (right):
> 
>
https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
> webrtc/modules/audio_coding/codecs/opus/opus_unittest.cc:211: const uint16_t
> kOutputValueBound = 8;
> On 2016/01/29 14:26:32, tlegrand-webrtc wrote:
> > It looks like noise during DTX got worse with the new version? 
> 
> I have discussed this with Felicia. Yes, the value increased, which is because
> the decaying of CNG becomes slower in Opus 1.2. We have checked the long time
> behavior and no noise pumping showed up.

Great! Thanks for the clarification.

[flim-webrtc, 2016-02-02 11:58:40]

Thanks for the reviews.

https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
File webrtc/modules/audio_coding/neteq/audio_classifier_unittest.cc (right):

https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
webrtc/modules/audio_coding/neteq/audio_classifier_unittest.cc:64: #if
defined(WEBRTC_ARCH_ARM) || defined(WEBRTC_ARCH_ARM64)
On 2016/01/28 09:42:05, minyue-webrtc wrote:
> why is only mono affected?

Not sure why it's just mono. However, the differences in the output were just a
couple of packets at the transition edges that have changed, so I don't think it
will significantly affect existing behaviour/performance.

[kjellander_webrtc, 2016-02-02 17:46:36]

On 2016/02/02 11:58:40, flim-webrtc wrote:
> Thanks for the reviews.
> 
>
https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
> File webrtc/modules/audio_coding/neteq/audio_classifier_unittest.cc (right):
> 
>
https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
> webrtc/modules/audio_coding/neteq/audio_classifier_unittest.cc:64: #if
> defined(WEBRTC_ARCH_ARM) || defined(WEBRTC_ARCH_ARM64)
> On 2016/01/28 09:42:05, minyue-webrtc wrote:
> > why is only mono affected?
> 
> Not sure why it's just mono. However, the differences in the output were just
a
> couple of packets at the transition edges that have changed, so I don't think
it
> will significantly affect existing behaviour/performance.

I'm submitting this CL as a part of the roll that brings in Opus v.1.1.2 into
WebRTC in https://codereview.webrtc.org/1657343002/

I'm going to close this CL once that's commited.

[kjellander_webrtc, 2016-02-02 22:46:48]

On 2016/02/02 17:46:36, kjellander (webrtc) wrote:
> On 2016/02/02 11:58:40, flim-webrtc wrote:
> > Thanks for the reviews.
> > 
> >
>
https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
> > File webrtc/modules/audio_coding/neteq/audio_classifier_unittest.cc (right):
> > 
> >
>
https://codereview.webrtc.org/1629413002/diff/40001/webrtc/modules/audio_codi...
> > webrtc/modules/audio_coding/neteq/audio_classifier_unittest.cc:64: #if
> > defined(WEBRTC_ARCH_ARM) || defined(WEBRTC_ARCH_ARM64)
> > On 2016/01/28 09:42:05, minyue-webrtc wrote:
> > > why is only mono affected?
> > 
> > Not sure why it's just mono. However, the differences in the output were
just
> a
> > couple of packets at the transition edges that have changed, so I don't
think
> it
> > will significantly affect existing behaviour/performance.
> 
> I'm submitting this CL as a part of the roll that brings in Opus v.1.1.2 into
> WebRTC in https://codereview.webrtc.org/1657343002/
> 
> I'm going to close this CL once that's commited.

This is now committed in https://codereview.webrtc.org/1657343002/ after passing
all trybots. Nice job with preparing this CL!

[kjellander_webrtc, 2016-02-02 22:47:16]

Description was changed from

==========
Update sha1 files and unittests for Opus v1.1.2

BUG=
==========

to

==========
Update sha1 files and unittests for Opus v1.1.2

This was committed as part of
https://codereview.webrtc.org/1657343002/
==========