Change NetEq::GetAudio to use AudioFrame

webrtc/modules/audio_coding/neteq/neteq_unittest.cc

 /*
  *  Copyright (c) 2011 The WebRTC project authors. All Rights Reserved.
  *
  *  Use of this source code is governed by a BSD-style license
  *  that can be found in the LICENSE file in the root of the source
  *  tree. An additional intellectual property rights grant can be found
  *  in the file PATENTS.  All contributing project authors may
  *  be found in the AUTHORS file in the root of the source tree.
  */
 
@@ skipping 11 matching lines @@
 #include <memory>
 #include <set>
 #include <string>
 #include <vector>
 
 #include "gflags/gflags.h"
 #include "testing/gtest/include/gtest/gtest.h"
 #include "webrtc/modules/audio_coding/neteq/tools/audio_loop.h"
 #include "webrtc/modules/audio_coding/neteq/tools/rtp_file_source.h"
 #include "webrtc/modules/audio_coding/codecs/pcm16b/pcm16b.h"
+#include "webrtc/modules/include/module_common_types.h"
 #include "webrtc/test/testsupport/fileutils.h"
 #include "webrtc/typedefs.h"
 
 #ifdef WEBRTC_NETEQ_UNITTEST_BITEXACT
 #ifdef WEBRTC_ANDROID_PLATFORM_BUILD
 #include "external/webrtc/webrtc/modules/audio_coding/neteq/neteq_unittest.pb.h"
 #else
 #include "webrtc/audio_coding/neteq/neteq_unittest.pb.h"
 #endif
 #endif
@@ skipping 230 matching lines @@
 
 class NetEqDecodingTest : public ::testing::Test {
  protected:
   // NetEQ must be polled for data once every 10 ms. Thus, neither of the
   // constants below can be changed.
   static const int kTimeStepMs = 10;
   static const size_t kBlockSize8kHz = kTimeStepMs * 8;
   static const size_t kBlockSize16kHz = kTimeStepMs * 16;
   static const size_t kBlockSize32kHz = kTimeStepMs * 32;
   static const size_t kBlockSize48kHz = kTimeStepMs * 48;
-  static const size_t kMaxBlockSize = kBlockSize48kHz;
   static const int kInitSampleRateHz = 8000;
 
   NetEqDecodingTest();
   virtual void SetUp();
   virtual void TearDown();
   void SelectDecoders(NetEqDecoder* used_codec);
   void LoadDecoders();
   void OpenInputFile(const std::string &rtp_file);
-  void Process(size_t* out_len);
+  void Process();
 
   void DecodeAndCompare(const std::string& rtp_file,
                         const std::string& ref_file,
                         const std::string& stat_ref_file,
                         const std::string& rtcp_ref_file);
 
   static void PopulateRtpInfo(int frame_index,
                               int timestamp,
                               WebRtcRTPHeader* rtp_info);
   static void PopulateCng(int frame_index,
@@ skipping 14 matching lines @@
 
   void DuplicateCng();
 
   uint32_t PlayoutTimestamp();
 
   NetEq* neteq_;
   NetEq::Config config_;
   std::unique_ptr<test::RtpFileSource> rtp_source_;
   std::unique_ptr<test::Packet> packet_;
   unsigned int sim_clock_;
-  int16_t out_data_[kMaxBlockSize];
+  AudioFrame out_frame_;
   int output_sample_rate_;
   int algorithmic_delay_ms_;
 };
 
 // Allocating the static const so that it can be passed by reference.
 const int NetEqDecodingTest::kTimeStepMs;
 const size_t NetEqDecodingTest::kBlockSize8kHz;
 const size_t NetEqDecodingTest::kBlockSize16kHz;
 const size_t NetEqDecodingTest::kBlockSize32kHz;
-const size_t NetEqDecodingTest::kMaxBlockSize;
 const int NetEqDecodingTest::kInitSampleRateHz;
 
 NetEqDecodingTest::NetEqDecodingTest()
     : neteq_(NULL),
       config_(),
       sim_clock_(0),
       output_sample_rate_(kInitSampleRateHz),
       algorithmic_delay_ms_(0) {
   config_.sample_rate_hz = kInitSampleRateHz;
-  memset(out_data_, 0, sizeof(out_data_));
 }
 
 void NetEqDecodingTest::SetUp() {
   neteq_ = NetEq::Create(config_);
   NetEqNetworkStatistics stat;
   ASSERT_EQ(0, neteq_->NetworkStatistics(&stat));
   algorithmic_delay_ms_ = stat.current_buffer_size_ms;
   ASSERT_TRUE(neteq_);
   LoadDecoders();
 }
@@ skipping 42 matching lines @@
                                            "cng-nb", 13));
   // Load CNG 16 kHz.
   ASSERT_EQ(0, neteq_->RegisterPayloadType(NetEqDecoder::kDecoderCNGwb,
                                            "cng-wb", 98));
 }
 
 void NetEqDecodingTest::OpenInputFile(const std::string &rtp_file) {
   rtp_source_.reset(test::RtpFileSource::Create(rtp_file));
 }
 
-void NetEqDecodingTest::Process(size_t* out_len) {
+void NetEqDecodingTest::Process() {
   // Check if time to receive.
   while (packet_ && sim_clock_ >= packet_->time_ms()) {
     if (packet_->payload_length_bytes() > 0) {
       WebRtcRTPHeader rtp_header;
       packet_->ConvertHeader(&rtp_header);
 #ifndef WEBRTC_CODEC_ISAC
       // Ignore payload type 104 (iSAC-swb) if ISAC is not supported.
       if (rtp_header.header.payloadType != 104)
 #endif
       ASSERT_EQ(0, neteq_->InsertPacket(
                        rtp_header,
                        rtc::ArrayView<const uint8_t>(
                            packet_->payload(), packet_->payload_length_bytes()),
                        static_cast<uint32_t>(packet_->time_ms() *
                                              (output_sample_rate_ / 1000))));
     }
     // Get next packet.
     packet_.reset(rtp_source_->NextPacket());
   }
 
   // Get audio from NetEq.
   NetEqOutputType type;
-  size_t num_channels;
-  ASSERT_EQ(0, neteq_->GetAudio(kMaxBlockSize, out_data_, out_len,
-                                &num_channels, &type));
-  ASSERT_TRUE((*out_len == kBlockSize8kHz) ||
-              (*out_len == kBlockSize16kHz) ||
-              (*out_len == kBlockSize32kHz) ||
-              (*out_len == kBlockSize48kHz));
-  output_sample_rate_ = static_cast<int>(*out_len / 10 * 1000);
+  ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+  ASSERT_TRUE((out_frame_.samples_per_channel_ == kBlockSize8kHz) ||
+              (out_frame_.samples_per_channel_ == kBlockSize16kHz) ||
+              (out_frame_.samples_per_channel_ == kBlockSize32kHz) ||
+              (out_frame_.samples_per_channel_ == kBlockSize48kHz));
+  output_sample_rate_ = out_frame_.sample_rate_hz_;
   EXPECT_EQ(output_sample_rate_, neteq_->last_output_sample_rate_hz());
 
   // Increase time.
   sim_clock_ += kTimeStepMs;
 }
 
 void NetEqDecodingTest::DecodeAndCompare(const std::string& rtp_file,
                                          const std::string& ref_file,
                                          const std::string& stat_ref_file,
                                          const std::string& rtcp_ref_file) {
@@ skipping 16 matching lines @@
     rtcp_out_file = webrtc::test::OutputPath() + "neteq_rtcp_stats.dat";
   }
   RefFiles rtcp_stat_files(rtcp_ref_file, rtcp_out_file);
 
   packet_.reset(rtp_source_->NextPacket());
   int i = 0;
   while (packet_) {
     std::ostringstream ss;
     ss << "Lap number " << i++ << " in DecodeAndCompare while loop";
     SCOPED_TRACE(ss.str());  // Print out the parameter values on failure.
-    size_t out_len = 0;
-    ASSERT_NO_FATAL_FAILURE(Process(&out_len));
-    ASSERT_NO_FATAL_FAILURE(ref_files.ProcessReference(out_data_, out_len));
+    ASSERT_NO_FATAL_FAILURE(Process());
+    ASSERT_NO_FATAL_FAILURE(ref_files.ProcessReference(
+        out_frame_.data_, out_frame_.samples_per_channel_));
 
     // Query the network statistics API once per second
     if (sim_clock_ % 1000 == 0) {
       // Process NetworkStatistics.
       NetEqNetworkStatistics network_stats;
       ASSERT_EQ(0, neteq_->NetworkStatistics(&network_stats));
       ASSERT_NO_FATAL_FAILURE(
           network_stat_files.ProcessReference(network_stats));
       // Compare with CurrentDelay, which should be identical.
       EXPECT_EQ(network_stats.current_buffer_size_ms, neteq_->CurrentDelayMs());
@@ skipping 119 matching lines @@
     WebRtcRTPHeader rtp_info;
     rtp_info.header.sequenceNumber = i;
     rtp_info.header.timestamp = i * kSamples;
     rtp_info.header.ssrc = 0x1234;  // Just an arbitrary SSRC.
     rtp_info.header.payloadType = 94;  // PCM16b WB codec.
     rtp_info.header.markerBit = 0;
     ASSERT_EQ(0, neteq_->InsertPacket(rtp_info, payload, 0));
   }
   // Pull out all data.
   for (size_t i = 0; i < num_frames; ++i) {
-    size_t out_len;
-    size_t num_channels;
     NetEqOutputType type;
-    ASSERT_EQ(0, neteq_->GetAudio(kMaxBlockSize, out_data_, &out_len,
-                                  &num_channels, &type));
-    ASSERT_EQ(kBlockSize16kHz, out_len);
+    ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+    ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
   }
 
   NetEqNetworkStatistics stats;
   EXPECT_EQ(0, neteq_->NetworkStatistics(&stats));
   // Since all frames are dumped into NetEQ at once, but pulled out with 10 ms
   // spacing (per definition), we expect the delay to increase with 10 ms for
   // each packet. Thus, we are calculating the statistics for a series from 10
   // to 300, in steps of 10 ms.
   EXPECT_EQ(155, stats.mean_waiting_time_ms);
   EXPECT_EQ(155, stats.median_waiting_time_ms);
@@ skipping 19 matching lines @@
     int num_packets = (frame_index % 10 == 0 ? 2 : 1);
     for (int n = 0; n < num_packets; ++n) {
       uint8_t payload[kPayloadBytes] = {0};
       WebRtcRTPHeader rtp_info;
       PopulateRtpInfo(frame_index, frame_index * kSamples, &rtp_info);
       ASSERT_EQ(0, neteq_->InsertPacket(rtp_info, payload, 0));
       ++frame_index;
     }
 
     // Pull out data once.
-    size_t out_len;
-    size_t num_channels;
     NetEqOutputType type;
-    ASSERT_EQ(0, neteq_->GetAudio(kMaxBlockSize, out_data_, &out_len,
-                                  &num_channels, &type));
-    ASSERT_EQ(kBlockSize16kHz, out_len);
+    ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+    ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
   }
 
   NetEqNetworkStatistics network_stats;
   ASSERT_EQ(0, neteq_->NetworkStatistics(&network_stats));
   EXPECT_EQ(-103196, network_stats.clockdrift_ppm);
 }
 
 TEST_F(NetEqDecodingTest, TestAverageInterArrivalTimePositive) {
   const int kNumFrames = 5000;  // Needed for convergence.
   int frame_index = 0;
   const size_t kSamples = 10 * 16;
   const size_t kPayloadBytes = kSamples * 2;
   for (int i = 0; i < kNumFrames; ++i) {
     // Insert one packet each time, except every 10th time where we don't insert
     // any packet. This will create a positive clock-drift of approx. 11%.
     int num_packets = (i % 10 == 9 ? 0 : 1);
     for (int n = 0; n < num_packets; ++n) {
       uint8_t payload[kPayloadBytes] = {0};
       WebRtcRTPHeader rtp_info;
       PopulateRtpInfo(frame_index, frame_index * kSamples, &rtp_info);
       ASSERT_EQ(0, neteq_->InsertPacket(rtp_info, payload, 0));
       ++frame_index;
     }
 
     // Pull out data once.
-    size_t out_len;
-    size_t num_channels;
     NetEqOutputType type;
-    ASSERT_EQ(0, neteq_->GetAudio(kMaxBlockSize, out_data_, &out_len,
-                                  &num_channels, &type));
-    ASSERT_EQ(kBlockSize16kHz, out_len);
+    ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+    ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
   }
 
   NetEqNetworkStatistics network_stats;
   ASSERT_EQ(0, neteq_->NetworkStatistics(&network_stats));
   EXPECT_EQ(110946, network_stats.clockdrift_ppm);
 }
 
 void NetEqDecodingTest::LongCngWithClockDrift(double drift_factor,
                                               double network_freeze_ms,
                                               bool pull_audio_during_freeze,
                                               int delay_tolerance_ms,
                                               int max_time_to_speech_ms) {
   uint16_t seq_no = 0;
   uint32_t timestamp = 0;
   const int kFrameSizeMs = 30;
   const size_t kSamples = kFrameSizeMs * 16;
   const size_t kPayloadBytes = kSamples * 2;
   double next_input_time_ms = 0.0;
   double t_ms;
-  size_t out_len;
-  size_t num_channels;
   NetEqOutputType type;
 
   // Insert speech for 5 seconds.
   const int kSpeechDurationMs = 5000;
   for (t_ms = 0; t_ms < kSpeechDurationMs; t_ms += 10) {
     // Each turn in this for loop is 10 ms.
     while (next_input_time_ms <= t_ms) {
       // Insert one 30 ms speech frame.
       uint8_t payload[kPayloadBytes] = {0};
       WebRtcRTPHeader rtp_info;
       PopulateRtpInfo(seq_no, timestamp, &rtp_info);
       ASSERT_EQ(0, neteq_->InsertPacket(rtp_info, payload, 0));
       ++seq_no;
       timestamp += kSamples;
       next_input_time_ms += static_cast<double>(kFrameSizeMs) * drift_factor;
     }
     // Pull out data once.
-    ASSERT_EQ(0, neteq_->GetAudio(kMaxBlockSize, out_data_, &out_len,
-                                  &num_channels, &type));
-    ASSERT_EQ(kBlockSize16kHz, out_len);
+    ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+    ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
   }
 
   EXPECT_EQ(kOutputNormal, type);
   int32_t delay_before = timestamp - PlayoutTimestamp();
 
   // Insert CNG for 1 minute (= 60000 ms).
   const int kCngPeriodMs = 100;
   const int kCngPeriodSamples = kCngPeriodMs * 16;  // Period in 16 kHz samples.
   const int kCngDurationMs = 60000;
   for (; t_ms < kSpeechDurationMs + kCngDurationMs; t_ms += 10) {
     // Each turn in this for loop is 10 ms.
     while (next_input_time_ms <= t_ms) {
       // Insert one CNG frame each 100 ms.
       uint8_t payload[kPayloadBytes];
       size_t payload_len;
       WebRtcRTPHeader rtp_info;
       PopulateCng(seq_no, timestamp, &rtp_info, payload, &payload_len);
       ASSERT_EQ(0, neteq_->InsertPacket(
                        rtp_info,
                        rtc::ArrayView<const uint8_t>(payload, payload_len), 0));
       ++seq_no;
       timestamp += kCngPeriodSamples;
       next_input_time_ms += static_cast<double>(kCngPeriodMs) * drift_factor;
     }
     // Pull out data once.
-    ASSERT_EQ(0, neteq_->GetAudio(kMaxBlockSize, out_data_, &out_len,
-                                  &num_channels, &type));
-    ASSERT_EQ(kBlockSize16kHz, out_len);
+    ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+    ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
   }
 
   EXPECT_EQ(kOutputCNG, type);
 
   if (network_freeze_ms > 0) {
     // First keep pulling audio for |network_freeze_ms| without inserting
     // any data, then insert CNG data corresponding to |network_freeze_ms|
     // without pulling any output audio.
     const double loop_end_time = t_ms + network_freeze_ms;
     for (; t_ms < loop_end_time; t_ms += 10) {
       // Pull out data once.
-      ASSERT_EQ(0,
-                neteq_->GetAudio(
-                    kMaxBlockSize, out_data_, &out_len, &num_channels, &type));
-      ASSERT_EQ(kBlockSize16kHz, out_len);
+      ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+      ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
       EXPECT_EQ(kOutputCNG, type);
     }
     bool pull_once = pull_audio_during_freeze;
     // If |pull_once| is true, GetAudio will be called once half-way through
     // the network recovery period.
     double pull_time_ms = (t_ms + next_input_time_ms) / 2;
     while (next_input_time_ms <= t_ms) {
       if (pull_once && next_input_time_ms >= pull_time_ms) {
         pull_once = false;
         // Pull out data once.
-        ASSERT_EQ(
-            0,
-            neteq_->GetAudio(
-                kMaxBlockSize, out_data_, &out_len, &num_channels, &type));
-        ASSERT_EQ(kBlockSize16kHz, out_len);
+        ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+        ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
         EXPECT_EQ(kOutputCNG, type);
         t_ms += 10;
       }
       // Insert one CNG frame each 100 ms.
       uint8_t payload[kPayloadBytes];
       size_t payload_len;
       WebRtcRTPHeader rtp_info;
       PopulateCng(seq_no, timestamp, &rtp_info, payload, &payload_len);
       ASSERT_EQ(0, neteq_->InsertPacket(
                        rtp_info,
@@ skipping 12 matching lines @@
       // Insert one 30 ms speech frame.
       uint8_t payload[kPayloadBytes] = {0};
       WebRtcRTPHeader rtp_info;
       PopulateRtpInfo(seq_no, timestamp, &rtp_info);
       ASSERT_EQ(0, neteq_->InsertPacket(rtp_info, payload, 0));
       ++seq_no;
       timestamp += kSamples;
       next_input_time_ms += kFrameSizeMs * drift_factor;
     }
     // Pull out data once.
-    ASSERT_EQ(0, neteq_->GetAudio(kMaxBlockSize, out_data_, &out_len,
-                                  &num_channels, &type));
-    ASSERT_EQ(kBlockSize16kHz, out_len);
+    ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+    ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
     // Increase clock.
     t_ms += 10;
   }
 
   // Check that the speech starts again within reasonable time.
   double time_until_speech_returns_ms = t_ms - speech_restart_time_ms;
   EXPECT_LT(time_until_speech_returns_ms, max_time_to_speech_ms);
   int32_t delay_after = timestamp - PlayoutTimestamp();
   // Compare delay before and after, and make sure it differs less than 20 ms.
   EXPECT_LE(delay_after, delay_before + delay_tolerance_ms * 16);
@@ skipping 102 matching lines @@
 TEST_F(NetEqDecodingTest, MAYBE_DecoderError) {
   const size_t kPayloadBytes = 100;
   uint8_t payload[kPayloadBytes] = {0};
   WebRtcRTPHeader rtp_info;
   PopulateRtpInfo(0, 0, &rtp_info);
   rtp_info.header.payloadType = 103;  // iSAC, but the payload is invalid.
   EXPECT_EQ(0, neteq_->InsertPacket(rtp_info, payload, 0));
   NetEqOutputType type;
   // Set all of |out_data_| to 1, and verify that it was set to 0 by the call
   // to GetAudio.
-  for (size_t i = 0; i < kMaxBlockSize; ++i) {
-    out_data_[i] = 1;
+  for (size_t i = 0; i < AudioFrame::kMaxDataSizeSamples; ++i) {
+    out_frame_.data_[i] = 1;
   }
-  size_t num_channels;
-  size_t samples_per_channel;
-  EXPECT_EQ(NetEq::kFail,
-            neteq_->GetAudio(kMaxBlockSize, out_data_,
-                             &samples_per_channel, &num_channels, &type));
+  EXPECT_EQ(NetEq::kFail, neteq_->GetAudio(&out_frame_, &type));
   // Verify that there is a decoder error to check.
   EXPECT_EQ(NetEq::kDecoderErrorCode, neteq_->LastError());
 
   enum NetEqDecoderError {
     ISAC_LENGTH_MISMATCH = 6730,
     ISAC_RANGE_ERROR_DECODE_FRAME_LENGTH = 6640
   };
 #if defined(WEBRTC_CODEC_ISAC)
   EXPECT_EQ(ISAC_LENGTH_MISMATCH, neteq_->LastDecoderError());
 #elif defined(WEBRTC_CODEC_ISACFX)
   EXPECT_EQ(ISAC_RANGE_ERROR_DECODE_FRAME_LENGTH, neteq_->LastDecoderError());
 #endif
   // Verify that the first 160 samples are set to 0, and that the remaining
   // samples are left unmodified.
   static const int kExpectedOutputLength = 160;  // 10 ms at 16 kHz sample rate.
   for (int i = 0; i < kExpectedOutputLength; ++i) {
     std::ostringstream ss;
     ss << "i = " << i;
     SCOPED_TRACE(ss.str());  // Print out the parameter values on failure.
-    EXPECT_EQ(0, out_data_[i]);
+    EXPECT_EQ(0, out_frame_.data_[i]);
   }
-  for (size_t i = kExpectedOutputLength; i < kMaxBlockSize; ++i) {
+  for (size_t i = kExpectedOutputLength; i < AudioFrame::kMaxDataSizeSamples;
+       ++i) {
     std::ostringstream ss;
     ss << "i = " << i;
     SCOPED_TRACE(ss.str());  // Print out the parameter values on failure.
-    EXPECT_EQ(1, out_data_[i]);
+    EXPECT_EQ(1, out_frame_.data_[i]);
   }
 }
 
 TEST_F(NetEqDecodingTest, GetAudioBeforeInsertPacket) {
   NetEqOutputType type;
   // Set all of |out_data_| to 1, and verify that it was set to 0 by the call
   // to GetAudio.
-  for (size_t i = 0; i < kMaxBlockSize; ++i) {
-    out_data_[i] = 1;
+  for (size_t i = 0; i < AudioFrame::kMaxDataSizeSamples; ++i) {
+    out_frame_.data_[i] = 1;
   }
-  size_t num_channels;
-  size_t samples_per_channel;
-  EXPECT_EQ(0, neteq_->GetAudio(kMaxBlockSize, out_data_,
-                                &samples_per_channel,
-                                &num_channels, &type));
+  EXPECT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
   // Verify that the first block of samples is set to 0.
   static const int kExpectedOutputLength =
       kInitSampleRateHz / 100;  // 10 ms at initial sample rate.
   for (int i = 0; i < kExpectedOutputLength; ++i) {
     std::ostringstream ss;
     ss << "i = " << i;
     SCOPED_TRACE(ss.str());  // Print out the parameter values on failure.
-    EXPECT_EQ(0, out_data_[i]);
+    EXPECT_EQ(0, out_frame_.data_[i]);
   }
   // Verify that the sample rate did not change from the initial configuration.
   EXPECT_EQ(config_.sample_rate_hz, neteq_->last_output_sample_rate_hz());
 }
 
 class NetEqBgnTest : public NetEqDecodingTest {
  protected:
   virtual void TestCondition(double sum_squared_noise,
                              bool should_be_faded) = 0;
 
   void CheckBgn(int sampling_rate_hz) {
     size_t expected_samples_per_channel = 0;
     uint8_t payload_type = 0xFF;  // Invalid.
     if (sampling_rate_hz == 8000) {
       expected_samples_per_channel = kBlockSize8kHz;
       payload_type = 93;  // PCM 16, 8 kHz.
     } else if (sampling_rate_hz == 16000) {
       expected_samples_per_channel = kBlockSize16kHz;
       payload_type = 94;  // PCM 16, 16 kHZ.
     } else if (sampling_rate_hz == 32000) {
       expected_samples_per_channel = kBlockSize32kHz;
       payload_type = 95;  // PCM 16, 32 kHz.
     } else {
       ASSERT_TRUE(false);  // Unsupported test case.
     }
 
     NetEqOutputType type;
-    int16_t output[kBlockSize32kHz];  // Maximum size is chosen.
+    AudioFrame output;
     test::AudioLoop input;
     // We are using the same 32 kHz input file for all tests, regardless of
     // |sampling_rate_hz|. The output may sound weird, but the test is still
     // valid.
     ASSERT_TRUE(input.Init(
         webrtc::test::ResourcePath("audio_coding/testfile32kHz", "pcm"),
         10 * sampling_rate_hz,  // Max 10 seconds loop length.
         expected_samples_per_channel));
 
     // Payload of 10 ms of PCM16 32 kHz.
     uint8_t payload[kBlockSize32kHz * sizeof(int16_t)];
     WebRtcRTPHeader rtp_info;
     PopulateRtpInfo(0, 0, &rtp_info);
     rtp_info.header.payloadType = payload_type;
 
-    size_t number_channels = 0;
-    size_t samples_per_channel = 0;
-
     uint32_t receive_timestamp = 0;
     for (int n = 0; n < 10; ++n) {  // Insert few packets and get audio.
       auto block = input.GetNextBlock();
       ASSERT_EQ(expected_samples_per_channel, block.size());
       size_t enc_len_bytes =
           WebRtcPcm16b_Encode(block.data(), block.size(), payload);
       ASSERT_EQ(enc_len_bytes, expected_samples_per_channel * 2);
 
-      number_channels = 0;
-      samples_per_channel = 0;
       ASSERT_EQ(0, neteq_->InsertPacket(rtp_info, rtc::ArrayView<const uint8_t>(
                                                       payload, enc_len_bytes),
                                         receive_timestamp));
-      ASSERT_EQ(0,
-                neteq_->GetAudio(kBlockSize32kHz,
-                                 output,
-                                 &samples_per_channel,
-                                 &number_channels,
-                                 &type));
-      ASSERT_EQ(1u, number_channels);
-      ASSERT_EQ(expected_samples_per_channel, samples_per_channel);
+      output.Reset();
+      ASSERT_EQ(0, neteq_->GetAudio(&output, &type));
+      ASSERT_EQ(1u, output.num_channels_);
+      ASSERT_EQ(expected_samples_per_channel, output.samples_per_channel_);
       ASSERT_EQ(kOutputNormal, type);
 
       // Next packet.
       rtp_info.header.timestamp += expected_samples_per_channel;
       rtp_info.header.sequenceNumber++;
       receive_timestamp += expected_samples_per_channel;
     }
 
-    number_channels = 0;
-    samples_per_channel = 0;
+    output.Reset();
 
     // Get audio without inserting packets, expecting PLC and PLC-to-CNG. Pull
     // one frame without checking speech-type. This is the first frame pulled
     // without inserting any packet, and might not be labeled as PLC.
-    ASSERT_EQ(0,
-              neteq_->GetAudio(kBlockSize32kHz,
-                               output,
-                               &samples_per_channel,
-                               &number_channels,
-                               &type));
-    ASSERT_EQ(1u, number_channels);
-    ASSERT_EQ(expected_samples_per_channel, samples_per_channel);
+    ASSERT_EQ(0, neteq_->GetAudio(&output, &type));
+    ASSERT_EQ(1u, output.num_channels_);
+    ASSERT_EQ(expected_samples_per_channel, output.samples_per_channel_);
 
     // To be able to test the fading of background noise we need at lease to
     // pull 611 frames.
     const int kFadingThreshold = 611;
 
     // Test several CNG-to-PLC packet for the expected behavior. The number 20
     // is arbitrary, but sufficiently large to test enough number of frames.
     const int kNumPlcToCngTestFrames = 20;
     bool plc_to_cng = false;
     for (int n = 0; n < kFadingThreshold + kNumPlcToCngTestFrames; ++n) {
-      number_channels = 0;
-      samples_per_channel = 0;
-      memset(output, 1, sizeof(output));  // Set to non-zero.
-      ASSERT_EQ(0,
-                neteq_->GetAudio(kBlockSize32kHz,
-                                 output,
-                                 &samples_per_channel,
-                                 &number_channels,
-                                 &type));
-      ASSERT_EQ(1u, number_channels);
-      ASSERT_EQ(expected_samples_per_channel, samples_per_channel);
+      output.Reset();
+      memset(output.data_, 1, sizeof(output.data_));  // Set to non-zero.
+      ASSERT_EQ(0, neteq_->GetAudio(&output, &type));
+      ASSERT_EQ(1u, output.num_channels_);
+      ASSERT_EQ(expected_samples_per_channel, output.samples_per_channel_);
       if (type == kOutputPLCtoCNG) {
         plc_to_cng = true;
         double sum_squared = 0;
-        for (size_t k = 0; k < number_channels * samples_per_channel; ++k)
-          sum_squared += output[k] * output[k];
+        for (size_t k = 0;
+             k < output.num_channels_ * output.samples_per_channel_; ++k)
+          sum_squared += output.data_[k] * output.data_[k];
         TestCondition(sum_squared, n > kFadingThreshold);
       } else {
         EXPECT_EQ(kOutputPLC, type);
       }
     }
     EXPECT_TRUE(plc_to_cng);  // Just to be sure that PLC-to-CNG has occurred.
   }
 };
 
 class NetEqBgnTestOn : public NetEqBgnTest {
@@ skipping 137 matching lines @@
 // packets should not produce error, statistics should not show any packet loss
 // and sync-packets should decode to zero.
 // TODO(turajs) we will have a better test if we have a referece NetEq, and
 // when Sync packets are inserted in "test" NetEq we insert all-zero payload
 // in reference NetEq and compare the output of those two.
 TEST_F(NetEqDecodingTest, SyncPacketDecode) {
   WebRtcRTPHeader rtp_info;
   PopulateRtpInfo(0, 0, &rtp_info);
   const size_t kPayloadBytes = kBlockSize16kHz * sizeof(int16_t);
   uint8_t payload[kPayloadBytes];
-  int16_t decoded[kBlockSize16kHz];
+  AudioFrame output;
   int algorithmic_frame_delay = algorithmic_delay_ms_ / 10 + 1;
   for (size_t n = 0; n < kPayloadBytes; ++n) {
     payload[n] = (rand() & 0xF0) + 1;  // Non-zero random sequence.
   }
   // Insert some packets which decode to noise. We are not interested in
   // actual decoded values.
   NetEqOutputType output_type;
-  size_t num_channels;
-  size_t samples_per_channel;
   uint32_t receive_timestamp = 0;
   for (int n = 0; n < 100; ++n) {
     ASSERT_EQ(0, neteq_->InsertPacket(rtp_info, payload, receive_timestamp));
-    ASSERT_EQ(0, neteq_->GetAudio(kBlockSize16kHz, decoded,
-                                  &samples_per_channel, &num_channels,
-                                  &output_type));
-    ASSERT_EQ(kBlockSize16kHz, samples_per_channel);
-    ASSERT_EQ(1u, num_channels);
+    ASSERT_EQ(0, neteq_->GetAudio(&output, &output_type));
+    ASSERT_EQ(kBlockSize16kHz, output.samples_per_channel_);
+    ASSERT_EQ(1u, output.num_channels_);
 
     rtp_info.header.sequenceNumber++;
     rtp_info.header.timestamp += kBlockSize16kHz;
     receive_timestamp += kBlockSize16kHz;
   }
   const int kNumSyncPackets = 10;
 
   // Make sure sufficient number of sync packets are inserted that we can
   // conduct a test.
   ASSERT_GT(kNumSyncPackets, algorithmic_frame_delay);
   // Insert sync-packets, the decoded sequence should be all-zero.
   for (int n = 0; n < kNumSyncPackets; ++n) {
     ASSERT_EQ(0, neteq_->InsertSyncPacket(rtp_info, receive_timestamp));
-    ASSERT_EQ(0, neteq_->GetAudio(kBlockSize16kHz, decoded,
-                                  &samples_per_channel, &num_channels,
-                                  &output_type));
-    ASSERT_EQ(kBlockSize16kHz, samples_per_channel);
-    ASSERT_EQ(1u, num_channels);
+    ASSERT_EQ(0, neteq_->GetAudio(&output, &output_type));
+    ASSERT_EQ(kBlockSize16kHz, output.samples_per_channel_);
+    ASSERT_EQ(1u, output.num_channels_);
     if (n > algorithmic_frame_delay) {
-      EXPECT_TRUE(IsAllZero(decoded, samples_per_channel * num_channels));
+      EXPECT_TRUE(IsAllZero(
+          output.data_, output.samples_per_channel_ * output.num_channels_));
     }
     rtp_info.header.sequenceNumber++;
     rtp_info.header.timestamp += kBlockSize16kHz;
     receive_timestamp += kBlockSize16kHz;
   }
 
   // We insert regular packets, if sync packet are not correctly buffered then
   // network statistics would show some packet loss.
   for (int n = 0; n <= algorithmic_frame_delay + 10; ++n) {
     ASSERT_EQ(0, neteq_->InsertPacket(rtp_info, payload, receive_timestamp));
-    ASSERT_EQ(0, neteq_->GetAudio(kBlockSize16kHz, decoded,
-                                  &samples_per_channel, &num_channels,
-                                  &output_type));
+    ASSERT_EQ(0, neteq_->GetAudio(&output, &output_type));
     if (n >= algorithmic_frame_delay + 1) {
       // Expect that this frame contain samples from regular RTP.
-      EXPECT_TRUE(IsAllNonZero(decoded, samples_per_channel * num_channels));
+      EXPECT_TRUE(IsAllNonZero(
+          output.data_, output.samples_per_channel_ * output.num_channels_));
     }
     rtp_info.header.sequenceNumber++;
     rtp_info.header.timestamp += kBlockSize16kHz;
     receive_timestamp += kBlockSize16kHz;
   }
   NetEqNetworkStatistics network_stats;
   ASSERT_EQ(0, neteq_->NetworkStatistics(&network_stats));
   // Expecting a "clean" network.
   EXPECT_EQ(0, network_stats.packet_loss_rate);
   EXPECT_EQ(0, network_stats.expand_rate);
   EXPECT_EQ(0, network_stats.accelerate_rate);
   EXPECT_LE(network_stats.preemptive_rate, 150);
 }
 
 // Test if the size of the packet buffer reported correctly when containing
 // sync packets. Also, test if network packets override sync packets. That is to
 // prefer decoding a network packet to a sync packet, if both have same sequence
 // number and timestamp.
 TEST_F(NetEqDecodingTest, SyncPacketBufferSizeAndOverridenByNetworkPackets) {
   WebRtcRTPHeader rtp_info;
   PopulateRtpInfo(0, 0, &rtp_info);
   const size_t kPayloadBytes = kBlockSize16kHz * sizeof(int16_t);
   uint8_t payload[kPayloadBytes];
-  int16_t decoded[kBlockSize16kHz];
+  AudioFrame output;
   for (size_t n = 0; n < kPayloadBytes; ++n) {
     payload[n] = (rand() & 0xF0) + 1;  // Non-zero random sequence.
   }
   // Insert some packets which decode to noise. We are not interested in
   // actual decoded values.
   NetEqOutputType output_type;
-  size_t num_channels;
-  size_t samples_per_channel;
   uint32_t receive_timestamp = 0;
   int algorithmic_frame_delay = algorithmic_delay_ms_ / 10 + 1;
   for (int n = 0; n < algorithmic_frame_delay; ++n) {
     ASSERT_EQ(0, neteq_->InsertPacket(rtp_info, payload, receive_timestamp));
-    ASSERT_EQ(0, neteq_->GetAudio(kBlockSize16kHz, decoded,
-                                  &samples_per_channel, &num_channels,
-                                  &output_type));
-    ASSERT_EQ(kBlockSize16kHz, samples_per_channel);
-    ASSERT_EQ(1u, num_channels);
+    ASSERT_EQ(0, neteq_->GetAudio(&output, &output_type));
+    ASSERT_EQ(kBlockSize16kHz, output.samples_per_channel_);
+    ASSERT_EQ(1u, output.num_channels_);
     rtp_info.header.sequenceNumber++;
     rtp_info.header.timestamp += kBlockSize16kHz;
     receive_timestamp += kBlockSize16kHz;
   }
   const int kNumSyncPackets = 10;
 
   WebRtcRTPHeader first_sync_packet_rtp_info;
   memcpy(&first_sync_packet_rtp_info, &rtp_info, sizeof(rtp_info));
 
   // Insert sync-packets, but no decoding.
@@ skipping 14 matching lines @@
   // Insert.
   for (int n = 0; n < kNumSyncPackets; ++n) {
     ASSERT_EQ(0, neteq_->InsertPacket(rtp_info, payload, receive_timestamp));
     rtp_info.header.sequenceNumber++;
     rtp_info.header.timestamp += kBlockSize16kHz;
     receive_timestamp += kBlockSize16kHz;
   }
 
   // Decode.
   for (int n = 0; n < kNumSyncPackets; ++n) {
-    ASSERT_EQ(0, neteq_->GetAudio(kBlockSize16kHz, decoded,
-                                  &samples_per_channel, &num_channels,
-                                  &output_type));
-    ASSERT_EQ(kBlockSize16kHz, samples_per_channel);
-    ASSERT_EQ(1u, num_channels);
-    EXPECT_TRUE(IsAllNonZero(decoded, samples_per_channel * num_channels));
+    ASSERT_EQ(0, neteq_->GetAudio(&output, &output_type));
+    ASSERT_EQ(kBlockSize16kHz, output.samples_per_channel_);
+    ASSERT_EQ(1u, output.num_channels_);
+    EXPECT_TRUE(IsAllNonZero(
+        output.data_, output.samples_per_channel_ * output.num_channels_));
   }
 }
 
 void NetEqDecodingTest::WrapTest(uint16_t start_seq_no,
                                  uint32_t start_timestamp,
                                  const std::set<uint16_t>& drop_seq_numbers,
                                  bool expect_seq_no_wrap,
                                  bool expect_timestamp_wrap) {
   uint16_t seq_no = start_seq_no;
   uint32_t timestamp = start_timestamp;
   const int kBlocksPerFrame = 3;  // Number of 10 ms blocks per frame.
   const int kFrameSizeMs = kBlocksPerFrame * kTimeStepMs;
   const int kSamples = kBlockSize16kHz * kBlocksPerFrame;
   const size_t kPayloadBytes = kSamples * sizeof(int16_t);
   double next_input_time_ms = 0.0;
-  int16_t decoded[kBlockSize16kHz];
-  size_t num_channels;
-  size_t samples_per_channel;
-  NetEqOutputType output_type;
   uint32_t receive_timestamp = 0;
 
   // Insert speech for 2 seconds.
   const int kSpeechDurationMs = 2000;
   int packets_inserted = 0;
   uint16_t last_seq_no;
   uint32_t last_timestamp;
   bool timestamp_wrapped = false;
   bool seq_no_wrapped = false;
   for (double t_ms = 0; t_ms < kSpeechDurationMs; t_ms += 10) {
@@ skipping 26 matching lines @@
 
       ++seq_no;
       timestamp += kSamples;
       receive_timestamp += kSamples;
       next_input_time_ms += static_cast<double>(kFrameSizeMs);
 
       seq_no_wrapped |= seq_no < last_seq_no;
       timestamp_wrapped |= timestamp < last_timestamp;
     }
     // Pull out data once.
-    ASSERT_EQ(0, neteq_->GetAudio(kBlockSize16kHz, decoded,
-                                  &samples_per_channel, &num_channels,
-                                  &output_type));
-    ASSERT_EQ(kBlockSize16kHz, samples_per_channel);
-    ASSERT_EQ(1u, num_channels);
+    AudioFrame output;
+    NetEqOutputType output_type;
+    ASSERT_EQ(0, neteq_->GetAudio(&output, &output_type));
+    ASSERT_EQ(kBlockSize16kHz, output.samples_per_channel_);
+    ASSERT_EQ(1u, output.num_channels_);
 
     // Expect delay (in samples) to be less than 2 packets.
     EXPECT_LE(timestamp - PlayoutTimestamp(),
               static_cast<uint32_t>(kSamples * 2));
   }
   // Make sure we have actually tested wrap-around.
   ASSERT_EQ(expect_seq_no_wrap, seq_no_wrapped);
   ASSERT_EQ(expect_timestamp_wrap, timestamp_wrapped);
 }
 
@@ skipping 29 matching lines @@
   uint32_t timestamp = 0;
   const int kFrameSizeMs = 10;
   const int kSampleRateKhz = 16;
   const int kSamples = kFrameSizeMs * kSampleRateKhz;
   const size_t kPayloadBytes = kSamples * 2;
 
   const int algorithmic_delay_samples = std::max(
       algorithmic_delay_ms_ * kSampleRateKhz, 5 * kSampleRateKhz / 8);
   // Insert three speech packets. Three are needed to get the frame length
   // correct.
-  size_t out_len;
-  size_t num_channels;
   NetEqOutputType type;
   uint8_t payload[kPayloadBytes] = {0};
   WebRtcRTPHeader rtp_info;
   for (int i = 0; i < 3; ++i) {
     PopulateRtpInfo(seq_no, timestamp, &rtp_info);
     ASSERT_EQ(0, neteq_->InsertPacket(rtp_info, payload, 0));
     ++seq_no;
     timestamp += kSamples;
 
     // Pull audio once.
-    ASSERT_EQ(0,
-              neteq_->GetAudio(
-                  kMaxBlockSize, out_data_, &out_len, &num_channels, &type));
-    ASSERT_EQ(kBlockSize16kHz, out_len);
+    ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+    ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
   }
   // Verify speech output.
   EXPECT_EQ(kOutputNormal, type);
 
   // Insert same CNG packet twice.
   const int kCngPeriodMs = 100;
   const int kCngPeriodSamples = kCngPeriodMs * kSampleRateKhz;
   size_t payload_len;
   PopulateCng(seq_no, timestamp, &rtp_info, payload, &payload_len);
   // This is the first time this CNG packet is inserted.
   ASSERT_EQ(
       0, neteq_->InsertPacket(
              rtp_info, rtc::ArrayView<const uint8_t>(payload, payload_len), 0));
 
   // Pull audio once and make sure CNG is played.
-  ASSERT_EQ(0,
-            neteq_->GetAudio(
-                kMaxBlockSize, out_data_, &out_len, &num_channels, &type));
-  ASSERT_EQ(kBlockSize16kHz, out_len);
+  ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+  ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
   EXPECT_EQ(kOutputCNG, type);
   EXPECT_EQ(timestamp - algorithmic_delay_samples, PlayoutTimestamp());
 
   // Insert the same CNG packet again. Note that at this point it is old, since
   // we have already decoded the first copy of it.
   ASSERT_EQ(
       0, neteq_->InsertPacket(
              rtp_info, rtc::ArrayView<const uint8_t>(payload, payload_len), 0));
 
   // Pull audio until we have played |kCngPeriodMs| of CNG. Start at 10 ms since
   // we have already pulled out CNG once.
   for (int cng_time_ms = 10; cng_time_ms < kCngPeriodMs; cng_time_ms += 10) {
-    ASSERT_EQ(0,
-              neteq_->GetAudio(
-                  kMaxBlockSize, out_data_, &out_len, &num_channels, &type));
-    ASSERT_EQ(kBlockSize16kHz, out_len);
+    ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+    ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
     EXPECT_EQ(kOutputCNG, type);
     EXPECT_EQ(timestamp - algorithmic_delay_samples,
               PlayoutTimestamp());
   }
 
   // Insert speech again.
   ++seq_no;
   timestamp += kCngPeriodSamples;
   PopulateRtpInfo(seq_no, timestamp, &rtp_info);
   ASSERT_EQ(0, neteq_->InsertPacket(rtp_info, payload, 0));
 
   // Pull audio once and verify that the output is speech again.
-  ASSERT_EQ(0,
-            neteq_->GetAudio(
-                kMaxBlockSize, out_data_, &out_len, &num_channels, &type));
-  ASSERT_EQ(kBlockSize16kHz, out_len);
+  ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+  ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
   EXPECT_EQ(kOutputNormal, type);
   EXPECT_EQ(timestamp + kSamples - algorithmic_delay_samples,
             PlayoutTimestamp());
 }
 
 uint32_t NetEqDecodingTest::PlayoutTimestamp() {
   uint32_t playout_timestamp = 0;
   EXPECT_TRUE(neteq_->GetPlayoutTimestamp(&playout_timestamp));
   return playout_timestamp;
 }
@@ skipping 16 matching lines @@
 
   PopulateCng(seq_no, timestamp, &rtp_info, payload, &payload_len);
   ASSERT_EQ(
       NetEq::kOK,
       neteq_->InsertPacket(
           rtp_info, rtc::ArrayView<const uint8_t>(payload, payload_len), 0));
   ++seq_no;
   timestamp += kCngPeriodSamples;
 
   // Pull audio once and make sure CNG is played.
-  size_t out_len;
-  size_t num_channels;
   NetEqOutputType type;
-  ASSERT_EQ(0, neteq_->GetAudio(kMaxBlockSize, out_data_, &out_len,
-                                &num_channels, &type));
-  ASSERT_EQ(kBlockSize16kHz, out_len);
+  ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+  ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
   EXPECT_EQ(kOutputCNG, type);
 
   // Insert some speech packets.
   for (int i = 0; i < 3; ++i) {
     PopulateRtpInfo(seq_no, timestamp, &rtp_info);
     ASSERT_EQ(0, neteq_->InsertPacket(rtp_info, payload, 0));
     ++seq_no;
     timestamp += kSamples;
 
     // Pull audio once.
-    ASSERT_EQ(0, neteq_->GetAudio(kMaxBlockSize, out_data_, &out_len,
-                                  &num_channels, &type));
-    ASSERT_EQ(kBlockSize16kHz, out_len);
+    ASSERT_EQ(0, neteq_->GetAudio(&out_frame_, &type));
+    ASSERT_EQ(kBlockSize16kHz, out_frame_.samples_per_channel_);
   }
   // Verify speech output.
   EXPECT_EQ(kOutputNormal, type);
 }
 
 }  // namespace webrtc