Bug 4985 revealed two flaws
1. Opus duration estimate did not return correct length for DTX packets,
2. NetEq DoCodecInternalCng did not assign enough buffer.
P.S.
Generalizing problem 1, current NetEq decode function checks memory size by calling the duration estimate function. This is not ideal. A better way is to let codec's decode function to receive buffer size and return failure if it is not enough. This can be made in a separate CL.
BUG=webrtc:4985
R=henrik.lundin@webrtc.org
Review URL: https://codereview.webrtc.org/1334303005 .
Cr-Commit-Position: refs/heads/master@{#10031}
The Init() method was previously used to initialize and reset
decoders, and returned an error code. The new Reset() method is used
for reset only; the constructor is now responsible for fully
initializing the AudioDecoder.
Reset() doesn't return an error code; it turned out that none of the
functions it ended up calling could actually fail, so this CL removes
their error return codes as well.
R=henrik.lundin@webrtc.org
Review URL: https://codereview.webrtc.org/1319683002 .
Cr-Commit-Position: refs/heads/master@{#9798}
This reverts portions of commit cb180976dd0e9672cde4523d87b5f4857478b5e9, which
reverted commit 83ad33a8aed1fb00e422b6abd33c3e8942821c24. Specifically, the
files in webrtc/modules/audio_coding/codecs/ that are not in ilbc/ or isac/, as
well as webrtc/modules/audio_coding/main/test/opus_test.cc, are relanded.
The original commit message is below:
Upconvert various types to int.
Per comments from HL/kwiberg on https://webrtc-codereview.appspot.com/42569004 , when there is existing usage of mixed types (int16_t, int, etc.), we'd prefer to standardize on larger types like int and phase out use of int16_t.
Specifically, "Using int16 just because we're sure all reasonable values will fit in 16 bits isn't usually meaningful in C."
This converts some existing uses of int16_t (and, in a few cases, other types such as uint16_t) to int (or, in a few places, int32_t). Other locations will be converted to size_t in a separate change.
BUG=none
TBR=kwiberg
Review URL: https://codereview.webrtc.org/1179093003
Cr-Commit-Position: refs/heads/master@{#9424}
Per comments from HL/kwiberg on https://webrtc-codereview.appspot.com/42569004 , when there is existing usage of mixed types (int16_t, int, etc.), we'd prefer to standardize on larger types like int and phase out use of int16_t.
Specifically, "Using int16 just because we're sure all reasonable values will fit in 16 bits isn't usually meaningful in C."
This converts some existing uses of int16_t (and, in a few cases, other types such as uint16_t) to int (or, in a few places, int32_t). Other locations will be converted to size_t in a separate change.
BUG=none
R=andrew@webrtc.org, kwiberg@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/54629004
Cr-Commit-Position: refs/heads/master@{#9405}
A better solution than forcing OPUS_APPLICATION_VOIP when enabling DTX has been found, which is to set OPUS_SIGNAL_VOICE.
This reduces the uncertainty of entering DTX over silence period of audio.
This CL contains the setup of OPUS_SIGNAL_VOICE and decoupling opus application mode with DTX.
BUG=4559
R=henrik.lundin@webrtc.org, henrika@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/46959004
Cr-Commit-Position: refs/heads/master@{#9168}
As a by-product, a generic tool for testing and comparing the complexity of codecs is added, and new audio files are added to the resources.
Three complexity tests are included
1. Default Opus complexity
2. Opus complexity knob
3. Default iSAC complexity (to compare with Opus)
The complexity tests are only meant for development reasons
and not to be run at bots.
The .isolate file is only needed for the APK packaging and test execution on Android.
TEST=passes all trybots
BUG=
R=kjellander@webrtc.org, tina.legrand@webrtc.org, turaj@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/8969004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@5655 4adac7df-926f-26a2-2b94-8c16560cd09d
The function WebRtcOpus_DurationEst returned the number of samples
per packet in the native 48 kHz sample rate, while the decoder
function returns data in 32 kHz sample rate. This creates a discrepancy
that makes NetEQ's lip-sync functionality add too little delay.
BUG=1334
TEST=try bots
Review URL: https://webrtc-codereview.appspot.com/1069006
git-svn-id: http://webrtc.googlecode.com/svn/trunk@3403 4adac7df-926f-26a2-2b94-8c16560cd09d
The Opus audio codec targets applications for pure conversations as well as other types of audio (e.g. music), and there are two different settings to use for this (VoIP and AUDIO). In the current implementation of Opus in WebRTC we use VoIP only.
I this CL I have changed default setting to AUDIO in the case of stereo, and kept VoIP as default in case of mono.
Next step is to add an API to choose application mode.
BUG=issue1239
Review URL: https://webrtc-codereview.appspot.com/1007006
git-svn-id: http://webrtc.googlecode.com/svn/trunk@3319 4adac7df-926f-26a2-2b94-8c16560cd09d
During call setup Opus should always be signaled as a 48000 Hz stereo codec, not depending on what we plan to send, or how we plan to decode received packets.
The previous implementation had different payload types for mono and stereo, which breaks the proposed standard.
While working on this CL I ran in to the problem reported earlier, that we could get a crash related to deleting decoder memory. This should now be solved in Patch Set 3.
BUG=issue1013, issue1112
Review URL: https://webrtc-codereview.appspot.com/933022
git-svn-id: http://webrtc.googlecode.com/svn/trunk@3177 4adac7df-926f-26a2-2b94-8c16560cd09d
