generic-library/webrtc
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Revert 7846 "Adding DTX to WebRTC Opus wrapper"

Browse Source 
> Adding DTX to WebRTC Opus wrapper
> 
> This is a step toward adding Opus DTX support in WebRTC.
> 
> Note that opus_encode() returns 1 byte in case of DTX, then the packet does not need to be transmitted. See
> 
> https://mf4.xiph.org/jenkins/view/opus/job/opus/ws/doc/html/group__opus__encoder.html
> 
> We transmit the first 1-byte packet to let decoder be in-sync
> 
> BUG=webrtc:1014
> R=henrik.lundin@webrtc.org
> 
> Review URL: https://webrtc-codereview.appspot.com/13219004

TBR=minyue@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/34449004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7848 4adac7df-926f-26a2-2b94-8c16560cd09d
This commit is contained in:
minyue@webrtc.org
2014-12-09 15:11:15 +00:00
parent f244760827
commit 19dd129c69
5 changed files with 108 additions and 390 deletions
Download Patch File Download Diff File Expand all files Collapse all files

352
webrtc/modules/audio_coding/codecs/opus/opus_unittest.cc
View File

@@ -12,50 +12,34 @@
#include "testing/gtest/include/gtest/gtest.h"
#include "webrtc/modules/audio_coding/codecs/opus/interface/opus_interface.h"
#include "webrtc/modules/audio_coding/codecs/opus/opus_inst.h"
#include "webrtc/modules/audio_coding/neteq/tools/audio_loop.h"
#include "webrtc/test/testsupport/fileutils.h"
namespace webrtc {
using test::AudioLoop;
// Number of samples in a 60 ms stereo frame, sampled at 48 kHz.
const int kOpusMaxFrameSamples = 48 * 60 * 2;
// Maximum number of bytes in output bitstream.
const size_t kMaxBytes = 1000;
// Sample rate of Opus.
const int kOpusRateKhz = 48;
// Number of samples-per-channel in a 20 ms frame, sampled at 48 kHz.
const int kOpus20msFrameSamples = kOpusRateKhz * 20;
const int kOpus20msFrameSamples = 48 * 20;
// Number of samples-per-channel in a 10 ms frame, sampled at 48 kHz.
const int kOpus10msFrameSamples = kOpusRateKhz * 10;
const int kOpus10msFrameSamples = 48 * 10;
class OpusTest : public ::testing::Test {
protected:
OpusTest();
virtual void SetUp();
void TestSetMaxPlaybackRate(opus_int32 expect, int32_t set);
void TestDtxEffect(bool dtx);
// Prepare |speech_data_| for encoding, read from a hard-coded file.
// After preparation, |speech_data_.GetNextBlock()| returns a pointer to a
// block of |block_length_ms| milliseconds. The data is looped every
// |loop_length_ms| milliseconds.
void PrepareSpeechData(int channel, int block_length_ms, int loop_length_ms);
int EncodeDecode(WebRtcOpusEncInst* encoder,
const int16_t* input_audio,
const int input_samples,
WebRtcOpusDecInst* decoder,
int16_t* output_audio,
int16_t* audio_type);
WebRtcOpusEncInst* opus_mono_encoder_;
WebRtcOpusEncInst* opus_stereo_encoder_;
WebRtcOpusDecInst* opus_mono_decoder_;
WebRtcOpusDecInst* opus_stereo_decoder_;
AudioLoop speech_data_;
int16_t speech_data_[kOpusMaxFrameSamples];
int16_t output_data_[kOpusMaxFrameSamples];
uint8_t bitstream_[kMaxBytes];
int encoded_bytes_;
};
OpusTest::OpusTest()
@@ -65,16 +49,17 @@ OpusTest::OpusTest()
opus_stereo_decoder_(NULL) {
}
void OpusTest::PrepareSpeechData(int channel, int block_length_ms,
int loop_length_ms) {
void OpusTest::SetUp() {
FILE* input_file;
const std::string file_name =
webrtc::test::ResourcePath("audio_coding/speech_mono_32_48kHz", "pcm");
if (loop_length_ms < block_length_ms) {
loop_length_ms = block_length_ms;
}
EXPECT_TRUE(speech_data_.Init(file_name,
loop_length_ms * kOpusRateKhz * channel,
block_length_ms * kOpusRateKhz * channel));
input_file = fopen(file_name.c_str(), "rb");
ASSERT_TRUE(input_file != NULL);
ASSERT_EQ(kOpusMaxFrameSamples,
static_cast<int32_t>(fread(speech_data_, sizeof(int16_t),
kOpusMaxFrameSamples, input_file)));
fclose(input_file);
input_file = NULL;
}
void OpusTest::TestSetMaxPlaybackRate(opus_int32 expect, int32_t set) {
@@ -91,155 +76,6 @@ void OpusTest::TestSetMaxPlaybackRate(opus_int32 expect, int32_t set) {
EXPECT_EQ(expect, bandwidth);
}
int OpusTest::EncodeDecode(WebRtcOpusEncInst* encoder,
const int16_t* input_audio,
const int input_samples,
WebRtcOpusDecInst* decoder,
int16_t* output_audio,
int16_t* audio_type) {
encoded_bytes_ = WebRtcOpus_Encode(encoder,
input_audio,
input_samples, kMaxBytes,
bitstream_);
return WebRtcOpus_Decode(decoder, bitstream_,
encoded_bytes_, output_audio,
audio_type);
}
// Test if encoder/decoder can enter DTX mode properly and do not enter DTX when
// they should not. This test is signal dependent.
void OpusTest::TestDtxEffect(bool dtx) {
PrepareSpeechData(1, 20, 2000);
// Create encoder memory.
EXPECT_EQ(0, WebRtcOpus_EncoderCreate(&opus_mono_encoder_, 1));
EXPECT_EQ(0, WebRtcOpus_DecoderCreate(&opus_mono_decoder_, 1));
// Set bitrate.
EXPECT_EQ(0, WebRtcOpus_SetBitRate(opus_mono_encoder_, 32000));
// Set input audio as silence.
int16_t silence[kOpus20msFrameSamples] = {0};
// Setting DTX.
EXPECT_EQ(0, dtx ? WebRtcOpus_EnableDtx(opus_mono_encoder_) :
WebRtcOpus_DisableDtx(opus_mono_encoder_));
int16_t audio_type;
int16_t output_data_decode[kOpus20msFrameSamples];
for (int i = 0; i < 100; ++i) {
EXPECT_EQ(kOpus20msFrameSamples,
EncodeDecode(opus_mono_encoder_, speech_data_.GetNextBlock(),
kOpus20msFrameSamples, opus_mono_decoder_,
output_data_decode, &audio_type));
// If not DTX, it should never enter DTX mode. If DTX, we do not care since
// whether it enters DTX depends on the signal type.
if (!dtx) {
EXPECT_GT(encoded_bytes_, 1);
EXPECT_EQ(0, opus_mono_encoder_->in_dtx_mode);
EXPECT_EQ(0, opus_mono_decoder_->in_dtx_mode);
EXPECT_EQ(0, audio_type); // Speech.
}
}
// We input some silent segments. In DTX mode, the encoder will stop sending.
// However, DTX may happen after a while.
for (int i = 0; i < 21; ++i) {
EXPECT_EQ(kOpus20msFrameSamples,
EncodeDecode(opus_mono_encoder_, silence,
kOpus20msFrameSamples, opus_mono_decoder_,
output_data_decode, &audio_type));
if (!dtx) {
EXPECT_GT(encoded_bytes_, 1);
EXPECT_EQ(0, opus_mono_encoder_->in_dtx_mode);
EXPECT_EQ(0, opus_mono_decoder_->in_dtx_mode);
EXPECT_EQ(0, audio_type); // Speech.
}
}
// For this input signal, DTX happens now.
EXPECT_EQ(kOpus20msFrameSamples,
EncodeDecode(opus_mono_encoder_, silence,
kOpus20msFrameSamples, opus_mono_decoder_,
output_data_decode, &audio_type));
if (dtx) {
EXPECT_EQ(1, encoded_bytes_); // Send 1 byte.
EXPECT_EQ(1, opus_mono_encoder_->in_dtx_mode);
EXPECT_EQ(1, opus_mono_decoder_->in_dtx_mode);
EXPECT_EQ(2, audio_type); // Comfort noise.
} else {
EXPECT_GT(encoded_bytes_, 1);
EXPECT_EQ(0, opus_mono_encoder_->in_dtx_mode);
EXPECT_EQ(0, opus_mono_decoder_->in_dtx_mode);
EXPECT_EQ(0, audio_type); // Speech.
}
// DTX mode is maintained 400 ms.
for (int i = 0; i < 20; ++i) {
EXPECT_EQ(kOpus20msFrameSamples,
EncodeDecode(opus_mono_encoder_, silence,
kOpus20msFrameSamples, opus_mono_decoder_,
output_data_decode, &audio_type));
if (dtx) {
EXPECT_EQ(0, encoded_bytes_); // Send 0 byte.
EXPECT_EQ(1, opus_mono_encoder_->in_dtx_mode);
EXPECT_EQ(1, opus_mono_decoder_->in_dtx_mode);
EXPECT_EQ(2, audio_type); // Comfort noise.
} else {
EXPECT_GT(encoded_bytes_, 1);
EXPECT_EQ(0, opus_mono_encoder_->in_dtx_mode);
EXPECT_EQ(0, opus_mono_decoder_->in_dtx_mode);
EXPECT_EQ(0, audio_type); // Speech.
}
}
// Quit DTX after 400 ms
EXPECT_EQ(kOpus20msFrameSamples,
EncodeDecode(opus_mono_encoder_, silence,
kOpus20msFrameSamples, opus_mono_decoder_,
output_data_decode, &audio_type));
EXPECT_GT(encoded_bytes_, 1);
EXPECT_EQ(0, opus_mono_encoder_->in_dtx_mode);
EXPECT_EQ(0, opus_mono_decoder_->in_dtx_mode);
EXPECT_EQ(0, audio_type); // Speech.
// Enters DTX again immediately.
EXPECT_EQ(kOpus20msFrameSamples,
EncodeDecode(opus_mono_encoder_, silence,
kOpus20msFrameSamples, opus_mono_decoder_,
output_data_decode, &audio_type));
if (dtx) {
EXPECT_EQ(1, encoded_bytes_); // Send 1 byte.
EXPECT_EQ(1, opus_mono_encoder_->in_dtx_mode);
EXPECT_EQ(1, opus_mono_decoder_->in_dtx_mode);
EXPECT_EQ(2, audio_type); // Comfort noise.
} else {
EXPECT_GT(encoded_bytes_, 1);
EXPECT_EQ(0, opus_mono_encoder_->in_dtx_mode);
EXPECT_EQ(0, opus_mono_decoder_->in_dtx_mode);
EXPECT_EQ(0, audio_type); // Speech.
}
silence[0] = 10000;
if (dtx) {
// Verify that encoder/decoder can jump out from DTX mode.
EXPECT_EQ(kOpus20msFrameSamples,
EncodeDecode(opus_mono_encoder_, silence,
kOpus20msFrameSamples, opus_mono_decoder_,
output_data_decode, &audio_type));
EXPECT_GT(encoded_bytes_, 1);
EXPECT_EQ(0, opus_mono_encoder_->in_dtx_mode);
EXPECT_EQ(0, opus_mono_decoder_->in_dtx_mode);
EXPECT_EQ(0, audio_type); // Speech.
}
// Free memory.
EXPECT_EQ(0, WebRtcOpus_EncoderFree(opus_mono_encoder_));
EXPECT_EQ(0, WebRtcOpus_DecoderFree(opus_mono_decoder_));
}
// Test failing Create.
TEST_F(OpusTest, OpusCreateFail) {
// Test to see that an invalid pointer is caught.
@@ -274,8 +110,6 @@ TEST_F(OpusTest, OpusCreateFree) {
}
TEST_F(OpusTest, OpusEncodeDecodeMono) {
PrepareSpeechData(1, 20, 20);
// Create encoder memory.
EXPECT_EQ(0, WebRtcOpus_EncoderCreate(&opus_mono_encoder_, 1));
EXPECT_EQ(0, WebRtcOpus_DecoderCreate(&opus_mono_decoder_, 1));
@@ -287,12 +121,16 @@ TEST_F(OpusTest, OpusEncodeDecodeMono) {
EXPECT_EQ(1, WebRtcOpus_DecoderChannels(opus_mono_decoder_));
// Encode & decode.
int16_t encoded_bytes;
int16_t audio_type;
int16_t output_data_decode[kOpus20msFrameSamples];
int16_t output_data_decode[kOpusMaxFrameSamples];
encoded_bytes = WebRtcOpus_Encode(opus_mono_encoder_, speech_data_,
kOpus20msFrameSamples, kMaxBytes,
bitstream_);
EXPECT_EQ(kOpus20msFrameSamples,
EncodeDecode(opus_mono_encoder_, speech_data_.GetNextBlock(),
kOpus20msFrameSamples, opus_mono_decoder_,
output_data_decode, &audio_type));
WebRtcOpus_Decode(opus_mono_decoder_, bitstream_,
encoded_bytes, output_data_decode,
&audio_type));
// Free memory.
EXPECT_EQ(0, WebRtcOpus_EncoderFree(opus_mono_encoder_));
@@ -300,8 +138,6 @@ TEST_F(OpusTest, OpusEncodeDecodeMono) {
}
TEST_F(OpusTest, OpusEncodeDecodeStereo) {
PrepareSpeechData(2, 20, 20);
// Create encoder memory.
EXPECT_EQ(0, WebRtcOpus_EncoderCreate(&opus_stereo_encoder_, 2));
EXPECT_EQ(0, WebRtcOpus_DecoderCreate(&opus_stereo_decoder_, 2));
@@ -313,12 +149,16 @@ TEST_F(OpusTest, OpusEncodeDecodeStereo) {
EXPECT_EQ(2, WebRtcOpus_DecoderChannels(opus_stereo_decoder_));
// Encode & decode.
int16_t encoded_bytes;
int16_t audio_type;
int16_t output_data_decode[kOpus20msFrameSamples * 2];
int16_t output_data_decode[kOpusMaxFrameSamples];
encoded_bytes = WebRtcOpus_Encode(opus_stereo_encoder_, speech_data_,
kOpus20msFrameSamples, kMaxBytes,
bitstream_);
EXPECT_EQ(kOpus20msFrameSamples,
EncodeDecode(opus_stereo_encoder_, speech_data_.GetNextBlock(),
kOpus20msFrameSamples, opus_stereo_decoder_,
output_data_decode, &audio_type));
WebRtcOpus_Decode(opus_stereo_decoder_, bitstream_,
encoded_bytes, output_data_decode,
&audio_type));
// Free memory.
EXPECT_EQ(0, WebRtcOpus_EncoderFree(opus_stereo_encoder_));
@@ -367,25 +207,28 @@ TEST_F(OpusTest, OpusSetComplexity) {
// Encode and decode one frame (stereo), initialize the decoder and
// decode once more.
TEST_F(OpusTest, OpusDecodeInit) {
PrepareSpeechData(2, 20, 20);
// Create encoder memory.
EXPECT_EQ(0, WebRtcOpus_EncoderCreate(&opus_stereo_encoder_, 2));
EXPECT_EQ(0, WebRtcOpus_DecoderCreate(&opus_stereo_decoder_, 2));
// Encode & decode.
int16_t encoded_bytes;
int16_t audio_type;
int16_t output_data_decode[kOpus20msFrameSamples * 2];
int16_t output_data_decode[kOpusMaxFrameSamples];
encoded_bytes = WebRtcOpus_Encode(opus_stereo_encoder_, speech_data_,
kOpus20msFrameSamples, kMaxBytes,
bitstream_);
EXPECT_EQ(kOpus20msFrameSamples,
EncodeDecode(opus_stereo_encoder_, speech_data_.GetNextBlock(),
kOpus20msFrameSamples, opus_stereo_decoder_,
output_data_decode, &audio_type));
WebRtcOpus_Decode(opus_stereo_decoder_, bitstream_,
encoded_bytes, output_data_decode,
&audio_type));
EXPECT_EQ(0, WebRtcOpus_DecoderInit(opus_stereo_decoder_));
EXPECT_EQ(kOpus20msFrameSamples,
WebRtcOpus_Decode(opus_stereo_decoder_, bitstream_,
encoded_bytes_, output_data_decode,
encoded_bytes, output_data_decode,
&audio_type));
// Free memory.
@@ -412,61 +255,6 @@ TEST_F(OpusTest, OpusEnableDisableFec) {
EXPECT_EQ(0, WebRtcOpus_EncoderFree(opus_stereo_encoder_));
}
TEST_F(OpusTest, OpusEnableDisableDtx) {
// Test without creating encoder memory.
EXPECT_EQ(-1, WebRtcOpus_EnableDtx(opus_mono_encoder_));
EXPECT_EQ(-1, WebRtcOpus_DisableDtx(opus_stereo_encoder_));
// Create encoder memory, try with different bitrates.
EXPECT_EQ(0, WebRtcOpus_EncoderCreate(&opus_mono_encoder_, 1));
EXPECT_EQ(0, WebRtcOpus_EncoderCreate(&opus_stereo_encoder_, 2));
opus_int32 dtx;
// DTX is off by default.
opus_encoder_ctl(opus_mono_encoder_->encoder,
OPUS_GET_DTX(&dtx));
EXPECT_EQ(0, dtx);
opus_encoder_ctl(opus_stereo_encoder_->encoder,
OPUS_GET_DTX(&dtx));
EXPECT_EQ(0, dtx);
// Test to enable DTX.
EXPECT_EQ(0, WebRtcOpus_EnableDtx(opus_mono_encoder_));
opus_encoder_ctl(opus_mono_encoder_->encoder,
OPUS_GET_DTX(&dtx));
EXPECT_EQ(1, dtx);
EXPECT_EQ(0, WebRtcOpus_EnableDtx(opus_stereo_encoder_));
opus_encoder_ctl(opus_stereo_encoder_->encoder,
OPUS_GET_DTX(&dtx));
EXPECT_EQ(1, dtx);
// Test to disable DTX.
EXPECT_EQ(0, WebRtcOpus_DisableDtx(opus_mono_encoder_));
opus_encoder_ctl(opus_mono_encoder_->encoder,
OPUS_GET_DTX(&dtx));
EXPECT_EQ(0, dtx);
EXPECT_EQ(0, WebRtcOpus_DisableDtx(opus_stereo_encoder_));
opus_encoder_ctl(opus_stereo_encoder_->encoder,
OPUS_GET_DTX(&dtx));
EXPECT_EQ(0, dtx);
// Free memory.
EXPECT_EQ(0, WebRtcOpus_EncoderFree(opus_mono_encoder_));
EXPECT_EQ(0, WebRtcOpus_EncoderFree(opus_stereo_encoder_));
}
TEST_F(OpusTest, OpusDtxOff) {
TestDtxEffect(false);
}
TEST_F(OpusTest, OpusDtxOn) {
TestDtxEffect(true);
}
TEST_F(OpusTest, OpusSetPacketLossRate) {
// Test without creating encoder memory.
EXPECT_EQ(-1, WebRtcOpus_SetPacketLossRate(opus_mono_encoder_, 50));
@@ -515,8 +303,6 @@ TEST_F(OpusTest, OpusSetMaxPlaybackRate) {
// PLC in mono mode.
TEST_F(OpusTest, OpusDecodePlcMono) {
PrepareSpeechData(1, 20, 20);
// Create encoder memory.
EXPECT_EQ(0, WebRtcOpus_EncoderCreate(&opus_mono_encoder_, 1));
EXPECT_EQ(0, WebRtcOpus_DecoderCreate(&opus_mono_decoder_, 1));
@@ -528,15 +314,19 @@ TEST_F(OpusTest, OpusDecodePlcMono) {
EXPECT_EQ(1, WebRtcOpus_DecoderChannels(opus_mono_decoder_));
// Encode & decode.
int16_t encoded_bytes;
int16_t audio_type;
int16_t output_data_decode[kOpus20msFrameSamples];
int16_t output_data_decode[kOpusMaxFrameSamples];
encoded_bytes = WebRtcOpus_Encode(opus_mono_encoder_, speech_data_,
kOpus20msFrameSamples, kMaxBytes,
bitstream_);
EXPECT_EQ(kOpus20msFrameSamples,
EncodeDecode(opus_mono_encoder_, speech_data_.GetNextBlock(),
kOpus20msFrameSamples, opus_mono_decoder_,
output_data_decode, &audio_type));
WebRtcOpus_Decode(opus_mono_decoder_, bitstream_,
encoded_bytes, output_data_decode,
&audio_type));
// Call decoder PLC.
int16_t plc_buffer[kOpus20msFrameSamples];
int16_t plc_buffer[kOpusMaxFrameSamples];
EXPECT_EQ(kOpus20msFrameSamples,
WebRtcOpus_DecodePlc(opus_mono_decoder_, plc_buffer, 1));
@@ -547,8 +337,6 @@ TEST_F(OpusTest, OpusDecodePlcMono) {
// PLC in stereo mode.
TEST_F(OpusTest, OpusDecodePlcStereo) {
PrepareSpeechData(2, 20, 20);
// Create encoder memory.
EXPECT_EQ(0, WebRtcOpus_EncoderCreate(&opus_stereo_encoder_, 2));
EXPECT_EQ(0, WebRtcOpus_DecoderCreate(&opus_stereo_decoder_, 2));
@@ -560,15 +348,19 @@ TEST_F(OpusTest, OpusDecodePlcStereo) {
EXPECT_EQ(2, WebRtcOpus_DecoderChannels(opus_stereo_decoder_));
// Encode & decode.
int16_t encoded_bytes;
int16_t audio_type;
int16_t output_data_decode[kOpus20msFrameSamples * 2];
int16_t output_data_decode[kOpusMaxFrameSamples];
encoded_bytes = WebRtcOpus_Encode(opus_stereo_encoder_, speech_data_,
kOpus20msFrameSamples, kMaxBytes,
bitstream_);
EXPECT_EQ(kOpus20msFrameSamples,
EncodeDecode(opus_stereo_encoder_, speech_data_.GetNextBlock(),
kOpus20msFrameSamples, opus_stereo_decoder_,
output_data_decode, &audio_type));
WebRtcOpus_Decode(opus_stereo_decoder_, bitstream_,
encoded_bytes, output_data_decode,
&audio_type));
// Call decoder PLC.
int16_t plc_buffer[kOpus20msFrameSamples * 2];
int16_t plc_buffer[kOpusMaxFrameSamples];
EXPECT_EQ(kOpus20msFrameSamples,
WebRtcOpus_DecodePlc(opus_stereo_decoder_, plc_buffer, 1));
@@ -579,29 +371,27 @@ TEST_F(OpusTest, OpusDecodePlcStereo) {
// Duration estimation.
TEST_F(OpusTest, OpusDurationEstimation) {
PrepareSpeechData(2, 20, 20);
// Create.
EXPECT_EQ(0, WebRtcOpus_EncoderCreate(&opus_stereo_encoder_, 2));
EXPECT_EQ(0, WebRtcOpus_DecoderCreate(&opus_stereo_decoder_, 2));
// 10 ms. We use only first 10 ms of a 20 ms block.
encoded_bytes_ = WebRtcOpus_Encode(opus_stereo_encoder_,
speech_data_.GetNextBlock(),
kOpus10msFrameSamples, kMaxBytes,
bitstream_);
int16_t encoded_bytes;
// 10 ms.
encoded_bytes = WebRtcOpus_Encode(opus_stereo_encoder_, speech_data_,
kOpus10msFrameSamples, kMaxBytes,
bitstream_);
EXPECT_EQ(kOpus10msFrameSamples,
WebRtcOpus_DurationEst(opus_stereo_decoder_, bitstream_,
encoded_bytes_));
encoded_bytes));
// 20 ms
encoded_bytes_ = WebRtcOpus_Encode(opus_stereo_encoder_,
speech_data_.GetNextBlock(),
kOpus20msFrameSamples, kMaxBytes,
bitstream_);
encoded_bytes = WebRtcOpus_Encode(opus_stereo_encoder_, speech_data_,
kOpus20msFrameSamples, kMaxBytes,
bitstream_);
EXPECT_EQ(kOpus20msFrameSamples,
WebRtcOpus_DurationEst(opus_stereo_decoder_, bitstream_,
encoded_bytes_));
encoded_bytes));
// Free memory.
EXPECT_EQ(0, WebRtcOpus_EncoderFree(opus_stereo_encoder_));