Home | History | Annotate | Download | only in media
      1 /*
      2  * libjingle
      3  * Copyright 2004 Google Inc.
      4  *
      5  * Redistribution and use in source and binary forms, with or without
      6  * modification, are permitted provided that the following conditions are met:
      7  *
      8  *  1. Redistributions of source code must retain the above copyright notice,
      9  *     this list of conditions and the following disclaimer.
     10  *  2. Redistributions in binary form must reproduce the above copyright notice,
     11  *     this list of conditions and the following disclaimer in the documentation
     12  *     and/or other materials provided with the distribution.
     13  *  3. The name of the author may not be used to endorse or promote products
     14  *     derived from this software without specific prior written permission.
     15  *
     16  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED
     17  * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
     18  * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO
     19  * EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
     20  * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
     21  * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
     22  * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
     23  * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
     24  * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
     25  * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
     26  */
     27 
     28 #include "talk/base/gunit.h"
     29 #include "talk/base/thread.h"
     30 #include "talk/session/media/call.h"
     31 #include "talk/session/media/currentspeakermonitor.h"
     32 
     33 namespace cricket {
     34 
     35 static const uint32 kSsrc1 = 1001;
     36 static const uint32 kSsrc2 = 1002;
     37 static const uint32 kMinTimeBetweenSwitches = 10;
     38 // Due to limited system clock resolution, the CurrentSpeakerMonitor may
     39 // actually require more or less time between switches than that specified
     40 // in the call to set_min_time_between_switches.  To be safe, we sleep for
     41 // 90 ms more than the min time between switches before checking for a switch.
     42 // I am assuming system clocks do not have a coarser resolution than 90 ms.
     43 static const uint32 kSleepTimeBetweenSwitches = 100;
     44 
     45 class MockCall : public Call {
     46  public:
     47   MockCall() : Call(NULL) {}
     48 
     49   void EmitAudioMonitor(const AudioInfo& info) {
     50     GetAudioSourceProxy()->SignalAudioMonitor(GetAudioSourceProxy(), info);
     51   }
     52 };
     53 
     54 class CurrentSpeakerMonitorTest : public testing::Test,
     55     public sigslot::has_slots<> {
     56  public:
     57   CurrentSpeakerMonitorTest() {
     58     call_ = new MockCall();
     59     monitor_ = new CurrentSpeakerMonitor(call_->GetAudioSourceProxy(), NULL);
     60     // Shrink the minimum time betweeen switches to 10 ms so we don't have to
     61     // slow down our tests.
     62     monitor_->set_min_time_between_switches(kMinTimeBetweenSwitches);
     63     monitor_->SignalUpdate.connect(this, &CurrentSpeakerMonitorTest::OnUpdate);
     64     current_speaker_ = 0;
     65     num_changes_ = 0;
     66     monitor_->Start();
     67   }
     68 
     69   ~CurrentSpeakerMonitorTest() {
     70     delete monitor_;
     71     delete call_;
     72   }
     73 
     74  protected:
     75   MockCall* call_;
     76   CurrentSpeakerMonitor* monitor_;
     77   int num_changes_;
     78   uint32 current_speaker_;
     79 
     80   void OnUpdate(CurrentSpeakerMonitor* monitor, uint32 current_speaker) {
     81     current_speaker_ = current_speaker;
     82     num_changes_++;
     83   }
     84 };
     85 
     86 static void InitAudioInfo(AudioInfo* info, int input_level, int output_level) {
     87   info->input_level = input_level;
     88   info->output_level = output_level;
     89 }
     90 
     91 TEST_F(CurrentSpeakerMonitorTest, NoActiveStreams) {
     92   AudioInfo info;
     93   InitAudioInfo(&info, 0, 0);
     94   call_->EmitAudioMonitor(info);
     95 
     96   EXPECT_EQ(current_speaker_, 0U);
     97   EXPECT_EQ(num_changes_, 0);
     98 }
     99 
    100 TEST_F(CurrentSpeakerMonitorTest, MultipleActiveStreams) {
    101   AudioInfo info;
    102   InitAudioInfo(&info, 0, 0);
    103 
    104   info.active_streams.push_back(std::make_pair(kSsrc1, 3));
    105   info.active_streams.push_back(std::make_pair(kSsrc2, 7));
    106   call_->EmitAudioMonitor(info);
    107 
    108   // No speaker recognized because the initial sample is treated as possibly
    109   // just noise and disregarded.
    110   EXPECT_EQ(current_speaker_, 0U);
    111   EXPECT_EQ(num_changes_, 0);
    112 
    113   info.active_streams.push_back(std::make_pair(kSsrc1, 3));
    114   info.active_streams.push_back(std::make_pair(kSsrc2, 7));
    115   call_->EmitAudioMonitor(info);
    116 
    117   EXPECT_EQ(current_speaker_, kSsrc2);
    118   EXPECT_EQ(num_changes_, 1);
    119 }
    120 
    121 // See: https://code.google.com/p/webrtc/issues/detail?id=2409
    122 TEST_F(CurrentSpeakerMonitorTest, DISABLED_RapidSpeakerChange) {
    123   AudioInfo info;
    124   InitAudioInfo(&info, 0, 0);
    125 
    126   info.active_streams.push_back(std::make_pair(kSsrc1, 3));
    127   info.active_streams.push_back(std::make_pair(kSsrc2, 7));
    128   call_->EmitAudioMonitor(info);
    129 
    130   EXPECT_EQ(current_speaker_, 0U);
    131   EXPECT_EQ(num_changes_, 0);
    132 
    133   info.active_streams.push_back(std::make_pair(kSsrc1, 3));
    134   info.active_streams.push_back(std::make_pair(kSsrc2, 7));
    135   call_->EmitAudioMonitor(info);
    136 
    137   EXPECT_EQ(current_speaker_, kSsrc2);
    138   EXPECT_EQ(num_changes_, 1);
    139 
    140   info.active_streams.push_back(std::make_pair(kSsrc1, 9));
    141   info.active_streams.push_back(std::make_pair(kSsrc2, 1));
    142   call_->EmitAudioMonitor(info);
    143 
    144   // We expect no speaker change because of the rapid change.
    145   EXPECT_EQ(current_speaker_, kSsrc2);
    146   EXPECT_EQ(num_changes_, 1);
    147 }
    148 
    149 TEST_F(CurrentSpeakerMonitorTest, SpeakerChange) {
    150   AudioInfo info;
    151   InitAudioInfo(&info, 0, 0);
    152 
    153   info.active_streams.push_back(std::make_pair(kSsrc1, 3));
    154   info.active_streams.push_back(std::make_pair(kSsrc2, 7));
    155   call_->EmitAudioMonitor(info);
    156 
    157   EXPECT_EQ(current_speaker_, 0U);
    158   EXPECT_EQ(num_changes_, 0);
    159 
    160   info.active_streams.push_back(std::make_pair(kSsrc1, 3));
    161   info.active_streams.push_back(std::make_pair(kSsrc2, 7));
    162   call_->EmitAudioMonitor(info);
    163 
    164   EXPECT_EQ(current_speaker_, kSsrc2);
    165   EXPECT_EQ(num_changes_, 1);
    166 
    167   // Wait so the changes don't come so rapidly.
    168   talk_base::Thread::SleepMs(kSleepTimeBetweenSwitches);
    169 
    170   info.active_streams.push_back(std::make_pair(kSsrc1, 9));
    171   info.active_streams.push_back(std::make_pair(kSsrc2, 1));
    172   call_->EmitAudioMonitor(info);
    173 
    174   EXPECT_EQ(current_speaker_, kSsrc1);
    175   EXPECT_EQ(num_changes_, 2);
    176 }
    177 
    178 TEST_F(CurrentSpeakerMonitorTest, InterwordSilence) {
    179   AudioInfo info;
    180   InitAudioInfo(&info, 0, 0);
    181 
    182   info.active_streams.push_back(std::make_pair(kSsrc1, 3));
    183   info.active_streams.push_back(std::make_pair(kSsrc2, 7));
    184   call_->EmitAudioMonitor(info);
    185 
    186   EXPECT_EQ(current_speaker_, 0U);
    187   EXPECT_EQ(num_changes_, 0);
    188 
    189   info.active_streams.push_back(std::make_pair(kSsrc1, 3));
    190   info.active_streams.push_back(std::make_pair(kSsrc2, 7));
    191   call_->EmitAudioMonitor(info);
    192 
    193   EXPECT_EQ(current_speaker_, kSsrc2);
    194   EXPECT_EQ(num_changes_, 1);
    195 
    196   info.active_streams.push_back(std::make_pair(kSsrc1, 3));
    197   info.active_streams.push_back(std::make_pair(kSsrc2, 7));
    198   call_->EmitAudioMonitor(info);
    199 
    200   EXPECT_EQ(current_speaker_, kSsrc2);
    201   EXPECT_EQ(num_changes_, 1);
    202 
    203   // Wait so the changes don't come so rapidly.
    204   talk_base::Thread::SleepMs(kSleepTimeBetweenSwitches);
    205 
    206   info.active_streams.push_back(std::make_pair(kSsrc1, 3));
    207   info.active_streams.push_back(std::make_pair(kSsrc2, 0));
    208   call_->EmitAudioMonitor(info);
    209 
    210   // Current speaker shouldn't have changed because we treat this as an inter-
    211   // word silence.
    212   EXPECT_EQ(current_speaker_, kSsrc2);
    213   EXPECT_EQ(num_changes_, 1);
    214 
    215   info.active_streams.push_back(std::make_pair(kSsrc1, 3));
    216   info.active_streams.push_back(std::make_pair(kSsrc2, 0));
    217   call_->EmitAudioMonitor(info);
    218 
    219   // Current speaker shouldn't have changed because we treat this as an inter-
    220   // word silence.
    221   EXPECT_EQ(current_speaker_, kSsrc2);
    222   EXPECT_EQ(num_changes_, 1);
    223 
    224   info.active_streams.push_back(std::make_pair(kSsrc1, 3));
    225   info.active_streams.push_back(std::make_pair(kSsrc2, 0));
    226   call_->EmitAudioMonitor(info);
    227 
    228   // At this point, we should have concluded that SSRC2 stopped speaking.
    229   EXPECT_EQ(current_speaker_, kSsrc1);
    230   EXPECT_EQ(num_changes_, 2);
    231 }
    232 
    233 }  // namespace cricket
    234