Wrote a new NetEq unit test to test a network freeze during comfort
noise playout. The network freezes and resumes during the silence
period, and then resumes speech. It was verified that the delay
increased due to the freeze, and this CL contains a fix for that
problem.
BUG=2995
R=turaj@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/9849004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@5701 4adac7df-926f-26a2-2b94-8c16560cd09d
In practice, this will have only marginal effect. The length_limit
was increased from 6.7 ms to 10 ms. This is compared with the
input_length, which is equal to the decoded frame size. Thus,
this change will only affect encoded frame sizes in this range
(including 10 ms).
BUG=2696
R=tina.legrand@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/9969004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@5700 4adac7df-926f-26a2-2b94-8c16560cd09d
This CL implements a unit test to cover an case where comfort noise
packets should be discarded. The situation arises when NetEq gets a
duplicate comfort noise packet. Without this check, the duplicate would
be decoded, and a the timing would shift.
As it turned out, the corner-case funcionality was not completely
accurate in NetEq4. This is because decision_logic_::cng_state_ is set
after the corner-case check. In the old NetEq3, the corresponding state
was changed before the check. This is now fixed.
R=turaj@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/9639005
git-svn-id: http://webrtc.googlecode.com/svn/trunk@5685 4adac7df-926f-26a2-2b94-8c16560cd09d
Adding option to use mock or real objects instead of mocks.
This will help future testing efforts, where each test case can
select whether a mock or a real object should be used.
Adding new test InsertPacketsUntilBufferIsFull.
Removing a few uniteresting mock call warning.
R=turaj@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/9839004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@5684 4adac7df-926f-26a2-2b94-8c16560cd09d
std::memcpy -> memcpy for instance. This change was motivated by a
compile report complaining that std::rand() was used instead of rand(),
probably with a stdlib.h include instead of cstdlib. Use of C functions
without the std:: prefix is a lot more common, so removing std:: to
address this.
BUG=
R=tommi@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/9549004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@5658 4adac7df-926f-26a2-2b94-8c16560cd09d
The variable playout_timestamp_ was not updated to the latest decoded
timestamp while comfort noise was played. Instead, it was upadted using
dead reckoning, which caused it to drift away from the timestamps of the
incoming CNG packets. Now it is updated also during comfort noise
playout.
Since the change is only in NetEq4, this change also makes the test
PlaysOutAudioAndVideoInSync use both ACM1/NetEq3 and ACM2/NetEq4.
Re-enabling one NetEq unit test that is no longer failing thanks to this CL.
BUG=2932
R=stefan@webrtc.org, turaj@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/8799004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@5649 4adac7df-926f-26a2-2b94-8c16560cd09d
Converting the test to a method within the test fixture, and setting
up two tests that call this method. One for positive and one for
negative clock drift. The one with positive clock drift is disabled
for now since it does not pass, but will be re-enabled shortly.
This change is only made for NetEq4.
R=tlegrand@google.com
Review URL: https://webrtc-codereview.appspot.com/8599004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@5541 4adac7df-926f-26a2-2b94-8c16560cd09d
Add another test to NetEqPerformanceTest with no packet losses or
clock drift. The purpose of this test would be to focus on the
"clean" code path, i.e., the path taken when there are no network
problems. The reason is that this code path is presumably much
lighter in complexity, and regressions could easily drown in the
heavier code involved when combating losses and drift.
BUG=2859
R=kjellander@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/7689005
git-svn-id: http://webrtc.googlecode.com/svn/trunk@5452 4adac7df-926f-26a2-2b94-8c16560cd09d
The performance test is based on the neteq4_speed_test application. The
bulk of the test code is extracted into a test class, and included into
the neteq_unittest_tools target. The actual gtest that runs the
performance test is implemented in neteq_performance_unittest.cc, and
built as a part of webrtc_perf_tests.
The old stand-alone test application is now made dependent on the new
test class, to avoid code duplication.
BUG=2397
R=andrew@webrtc.org, kjellander@webrtc.org, turaj@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/6749004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@5362 4adac7df-926f-26a2-2b94-8c16560cd09d
The new version of Opus doesn't generate the same number of bytes encoding the test vectors in audio_decoder_unittest. Therefore the test was updated not to check the length of the encoded packet, to prepare for the coming roll of Opus. Same change was applied to iSAC, which can also generate different number of bytes on different platforms.
BUG=1459
R=henrik.lundin@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/4609004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@5195 4adac7df-926f-26a2-2b94-8c16560cd09d
The current implementation using std::vector is too slow.
This CL introduces a new implementation, using a regular
array as data container.
In AudioMultiVector::ReadInterleavedFromIndex, a special case for
1 channel was implemented, to further reduce runtime. Finally,
AudioMultiVector::Channels was reimplemented.
The changes in this CL reduces the runtime of neteq4_speed_test
by 33%.
BUG=1363
R=turaj@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/3509004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@5115 4adac7df-926f-26a2-2b94-8c16560cd09d
Analogous to the recent libjingle change: http://cl/54929753-p10.
This supports scoped_ptr<T[]> and scoped_ptr<C, FreeDeleter> rather
than scoped_array and scoped_ptr_malloc respectively.
- Add Chromium's template-based COMPILE_ASSERT. We didn't have this
previously in order to support the macro in C. Instead, move the
existing macro to compile_assert_c.h.
- Additionally copy the move.h and template_util.h depedencies and add
the WARN_UNUSED_RESULT macro.
- Leave scoped_array and scoped_ptr_malloc for now, but mark as
deprecated.
- Remove scoped_ptr foo(NULL) use. The default constructor handles it.
- Remove the now redundant COMPILE_ASSERT from peerconnection_jni.cc.
- Add a CHECK_ARRAY_SIZE macro to rtp_format_vp8_unittest.cc to remove
some repeated code.
TESTED=trybots
R=pbos@webrtc.org, tommi@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/2449005
git-svn-id: http://webrtc.googlecode.com/svn/trunk@5015 4adac7df-926f-26a2-2b94-8c16560cd09d
Recent changes in GYP seem to have broken our previous
"hack" for getting the GYP rule for .isolate files
imported from the Chromium build/isolate.gypi.
The best solution for now is to remove the hack
and check in a copy of Chromium's src/build/isolate.gypi
in WebRTC's build/ dir instead. A similar approach is
used for our build/protoc.gypi file.
TEST=On Linux, I successfully ran:
gclient runhooks
ninja -C out/Release
and verified a bunch of .isolated files were created in
out/Release (which didn't happen before this patch).
I also renamed the build/isolate.gypi from Chromium to
ensure that our own is used and not that one (in case any
paths would be incorrect).
I also ran build/gyp_chromium in a Chromium checkout
with WebRTC in third_party/webrtc having this patch applied
to ensure GYP processing was still working.
Finally, I verified that the same project generation and
compilation from a Chromium checkout worked the way we build
our Android native tests, using:
. build/android/envsetup.sh
GYP_DEFINES="$GYP_DEFINES include_tests=1 enable_tracing=1" gclient runhooks
ninja -C out/Release android_builder_webrtc
BUG=1916
R=andrew@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/2338004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@4907 4adac7df-926f-26a2-2b94-8c16560cd09d
We encounter a sample-underrun if NetEq is initialized with a sampling rate fs =16000 and receive Opus packets with frame-size less than 5 ms. The reason is as follows.
Let say NetEq buffer has 4 packets of Opus each of size 2.5ms this means that internally timestamp of packets incremented by 80 (internally Opus treated as 32 kHz codec). Given the initial sampling rate of NetEq, at the first time that it wants to fetch packets, it targets to fetch 160 samples. Therefore, it will only extracts 2 packets. Decoding these packets give us exactly 160 samples (5 ms at 32 kHz), however, upon decoding the first packet the internal sampling rate will be updated to 32 kHz. So it is expected that sync buffer to deliver 320 samples while it does only have 160 samples (or maybe few more as it starts with some zeros). And we encounter and under-run.
Even if we ignore the under-run "assert(sync_buffer_->FutureLength() >= expand_->overlap_length())" (neteq_impl.cc::811) is trigered. I'm not sure what happens if we remove this assert perhaps NetEq will work fine in subsequent calls. However the first under-run is blocking ACM2 test to pass.
Here I have a solution to update sample rate as soon as a packet is inserted, if required. It not a very efficient approach as we do the same reset in NetEqImpl::Decode().
It is a bit tricky to reproduce this because the TOT ACM tests do not run ACM2. In https://webrtc-codereview.appspot.com/2192005/ I have a patch to run both ACMs. To reproduce the problem, one can patch that CL and run
$ out/Debug/modules_tests --gtest_filter=AudioCodingModuleTest.TestOpus
Note that we would not encounter any problem if NetEq4 is initiated with 32000 Hz sampling rate. You can test this by setting |kNeteqInitSampleRateHz| to 32000 in webrtc/modules/audio_coding/main/acm2/acm_receiver.cc
BUG=
R=andrew@webrtc.org, henrik.lundin@webrtc.org, kjellander@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/2306004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@4896 4adac7df-926f-26a2-2b94-8c16560cd09d
