audio_conference_mixer/source/audio_conference_mixer_impl.cc

/*
 *  Copyright (c) 2012 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.
 */

#include "webrtc/modules/audio_conference_mixer/interface/audio_conference_mixer_defines.h"
#include "webrtc/modules/audio_conference_mixer/source/audio_conference_mixer_impl.h"
#include "webrtc/modules/audio_conference_mixer/source/audio_frame_manipulator.h"
#include "webrtc/modules/audio_processing/include/audio_processing.h"
#include "webrtc/modules/utility/interface/audio_frame_operations.h"
#include "webrtc/system_wrappers/interface/critical_section_wrapper.h"
#include "webrtc/system_wrappers/interface/trace.h"

namespace webrtc {
namespace {

struct ParticipantFramePair {
  MixerParticipant* participant;
  AudioFrame* audioFrame;
};

typedef std::list<ParticipantFramePair*> ParticipantFramePairList;

// Mix |frame| into |mixed_frame|, with saturation protection and upmixing.
// These effects are applied to |frame| itself prior to mixing. Assumes that
// |mixed_frame| always has at least as many channels as |frame|. Supports
// stereo at most.
//
// TODO(andrew): consider not modifying |frame| here.
void MixFrames(AudioFrame* mixed_frame, AudioFrame* frame, bool use_limiter) {
  assert(mixed_frame->num_channels_ >= frame->num_channels_);
  if (use_limiter) {
    // Divide by two to avoid saturation in the mixing.
    // This is only meaningful if the limiter will be used.
    *frame >>= 1;
  }
  if (mixed_frame->num_channels_ > frame->num_channels_) {
    // We only support mono-to-stereo.
    assert(mixed_frame->num_channels_ == 2 &&
           frame->num_channels_ == 1);
    AudioFrameOperations::MonoToStereo(frame);
  }

  *mixed_frame += *frame;
}

// Return the max number of channels from a |list| composed of AudioFrames.
int MaxNumChannels(const AudioFrameList* list) {
  int max_num_channels = 1;
  for (AudioFrameList::const_iterator iter = list->begin();
       iter != list->end();
       ++iter) {
    max_num_channels = std::max(max_num_channels, (*iter)->num_channels_);
  }
  return max_num_channels;
}

void SetParticipantStatistics(ParticipantStatistics* stats,
                              const AudioFrame& frame) {
    stats->participant = frame.id_;
    stats->level = 0;  // TODO(andrew): to what should this be set?
}

}  // namespace

MixerParticipant::MixerParticipant()
    : _mixHistory(new MixHistory()) {
}

MixerParticipant::~MixerParticipant() {
    delete _mixHistory;
}

int32_t MixerParticipant::IsMixed(bool& mixed) const {
    return _mixHistory->IsMixed(mixed);
}

MixHistory::MixHistory()
    : _isMixed(0) {
}

MixHistory::~MixHistory() {
}

int32_t MixHistory::IsMixed(bool& mixed) const {
    mixed = _isMixed;
    return 0;
}

int32_t MixHistory::WasMixed(bool& wasMixed) const {
    // Was mixed is the same as is mixed depending on perspective. This function
    // is for the perspective of AudioConferenceMixerImpl.
    return IsMixed(wasMixed);
}

int32_t MixHistory::SetIsMixed(const bool mixed) {
    _isMixed = mixed;
    return 0;
}

void MixHistory::ResetMixedStatus() {
    _isMixed = false;
}

AudioConferenceMixer* AudioConferenceMixer::Create(int id) {
    AudioConferenceMixerImpl* mixer = new AudioConferenceMixerImpl(id);
    if(!mixer->Init()) {
        delete mixer;
        return NULL;
    }
    return mixer;
}

AudioConferenceMixerImpl::AudioConferenceMixerImpl(int id)
    : _scratchParticipantsToMixAmount(0),
      _scratchMixedParticipants(),
      _scratchVadPositiveParticipantsAmount(0),
      _scratchVadPositiveParticipants(),
      _id(id),
      _minimumMixingFreq(kLowestPossible),
      _mixReceiver(NULL),
      _mixerStatusCallback(NULL),
      _amountOf10MsBetweenCallbacks(1),
      _amountOf10MsUntilNextCallback(0),
      _mixerStatusCb(false),
      _outputFrequency(kDefaultFrequency),
      _sampleSize(0),
      _audioFramePool(NULL),
      _participantList(),
      _additionalParticipantList(),
      _numMixedParticipants(0),
      use_limiter_(true),
      _timeStamp(0),
      _timeScheduler(kProcessPeriodicityInMs),
      _mixedAudioLevel(),
      _processCalls(0) {}

bool AudioConferenceMixerImpl::Init() {
    _crit.reset(CriticalSectionWrapper::CreateCriticalSection());
    if (_crit.get() == NULL)
        return false;

    _cbCrit.reset(CriticalSectionWrapper::CreateCriticalSection());
    if(_cbCrit.get() == NULL)
        return false;

    Config config;
    config.Set<ExperimentalAgc>(new ExperimentalAgc(false));
    _limiter.reset(AudioProcessing::Create(config));
    if(!_limiter.get())
        return false;

    MemoryPool<AudioFrame>::CreateMemoryPool(_audioFramePool,
                                             DEFAULT_AUDIO_FRAME_POOLSIZE);
    if(_audioFramePool == NULL)
        return false;

    if(SetOutputFrequency(kDefaultFrequency) == -1)
        return false;

    if(_limiter->gain_control()->set_mode(GainControl::kFixedDigital) !=
        _limiter->kNoError)
        return false;

    // We smoothly limit the mixed frame to -7 dbFS. -6 would correspond to the
    // divide-by-2 but -7 is used instead to give a bit of headroom since the
    // AGC is not a hard limiter.
    if(_limiter->gain_control()->set_target_level_dbfs(7) != _limiter->kNoError)
        return false;

    if(_limiter->gain_control()->set_compression_gain_db(0)
        != _limiter->kNoError)
        return false;

    if(_limiter->gain_control()->enable_limiter(true) != _limiter->kNoError)
        return false;

    if(_limiter->gain_control()->Enable(true) != _limiter->kNoError)
        return false;

    return true;
}

AudioConferenceMixerImpl::~AudioConferenceMixerImpl() {
    MemoryPool<AudioFrame>::DeleteMemoryPool(_audioFramePool);
    assert(_audioFramePool == NULL);
}

int32_t AudioConferenceMixerImpl::ChangeUniqueId(const int32_t id) {
    _id = id;
    return 0;
}

// Process should be called every kProcessPeriodicityInMs ms
int32_t AudioConferenceMixerImpl::TimeUntilNextProcess() {
    int32_t timeUntilNextProcess = 0;
    CriticalSectionScoped cs(_crit.get());
    if(_timeScheduler.TimeToNextUpdate(timeUntilNextProcess) != 0) {
        WEBRTC_TRACE(kTraceError, kTraceAudioMixerServer, _id,
                     "failed in TimeToNextUpdate() call");
        // Sanity check
        assert(false);
        return -1;
    }
    return timeUntilNextProcess;
}

int32_t AudioConferenceMixerImpl::Process() {
    size_t remainingParticipantsAllowedToMix =
        kMaximumAmountOfMixedParticipants;
    {
        CriticalSectionScoped cs(_crit.get());
        assert(_processCalls == 0);
        _processCalls++;

        // Let the scheduler know that we are running one iteration.
        _timeScheduler.UpdateScheduler();
    }

    AudioFrameList mixList;
    AudioFrameList rampOutList;
    AudioFrameList additionalFramesList;
    std::map<int, MixerParticipant*> mixedParticipantsMap;
    {
        CriticalSectionScoped cs(_cbCrit.get());

        int32_t lowFreq = GetLowestMixingFrequency();
        // SILK can run in 12 kHz and 24 kHz. These frequencies are not
        // supported so use the closest higher frequency to not lose any
        // information.
        // TODO(henrike): this is probably more appropriate to do in
        //                GetLowestMixingFrequency().
        if (lowFreq == 12000) {
            lowFreq = 16000;
        } else if (lowFreq == 24000) {
            lowFreq = 32000;
        }
        if(lowFreq <= 0) {
            CriticalSectionScoped cs(_crit.get());
            _processCalls--;
            return 0;
        } else {
            switch(lowFreq) {
            case 8000:
                if(OutputFrequency() != kNbInHz) {
                    SetOutputFrequency(kNbInHz);
                }
                break;
            case 16000:
                if(OutputFrequency() != kWbInHz) {
                    SetOutputFrequency(kWbInHz);
                }
                break;
            case 32000:
                if(OutputFrequency() != kSwbInHz) {
                    SetOutputFrequency(kSwbInHz);
                }
                break;
            case 48000:
                if(OutputFrequency() != kFbInHz) {
                    SetOutputFrequency(kFbInHz);
                }
                break;
            default:
                assert(false);

                CriticalSectionScoped cs(_crit.get());
                _processCalls--;
                return -1;
            }
        }

        UpdateToMix(&mixList, &rampOutList, &mixedParticipantsMap,
                    remainingParticipantsAllowedToMix);

        GetAdditionalAudio(&additionalFramesList);
        UpdateMixedStatus(mixedParticipantsMap);
        _scratchParticipantsToMixAmount = mixedParticipantsMap.size();
    }

    // Get an AudioFrame for mixing from the memory pool.
    AudioFrame* mixedAudio = NULL;
    if(_audioFramePool->PopMemory(mixedAudio) == -1) {
        WEBRTC_TRACE(kTraceMemory, kTraceAudioMixerServer, _id,
                     "failed PopMemory() call");
        assert(false);
        return -1;
    }

    bool timeForMixerCallback = false;
    int retval = 0;
    int32_t audioLevel = 0;
    {
        CriticalSectionScoped cs(_crit.get());

        // TODO(henrike): it might be better to decide the number of channels
        //                with an API instead of dynamically.

        // Find the max channels over all mixing lists.
        const int num_mixed_channels = std::max(MaxNumChannels(&mixList),
            std::max(MaxNumChannels(&additionalFramesList),
                     MaxNumChannels(&rampOutList)));

        mixedAudio->UpdateFrame(-1, _timeStamp, NULL, 0, _outputFrequency,
                                AudioFrame::kNormalSpeech,
                                AudioFrame::kVadPassive, num_mixed_channels);

        _timeStamp += _sampleSize;

        // We only use the limiter if it supports the output sample rate and
        // we're actually mixing multiple streams.
        use_limiter_ = _numMixedParticipants > 1 &&
                       _outputFrequency <= kAudioProcMaxNativeSampleRateHz;

        MixFromList(*mixedAudio, &mixList);
        MixAnonomouslyFromList(*mixedAudio, &additionalFramesList);
        MixAnonomouslyFromList(*mixedAudio, &rampOutList);

        if(mixedAudio->samples_per_channel_ == 0) {
            // Nothing was mixed, set the audio samples to silence.
            mixedAudio->samples_per_channel_ = _sampleSize;
            mixedAudio->Mute();
        } else {
            // Only call the limiter if we have something to mix.
            if(!LimitMixedAudio(*mixedAudio))
                retval = -1;
        }

        _mixedAudioLevel.ComputeLevel(mixedAudio->data_,_sampleSize);
        audioLevel = _mixedAudioLevel.GetLevel();

        if(_mixerStatusCb) {
            _scratchVadPositiveParticipantsAmount = 0;
            UpdateVADPositiveParticipants(&mixList);
            if(_amountOf10MsUntilNextCallback-- == 0) {
                _amountOf10MsUntilNextCallback = _amountOf10MsBetweenCallbacks;
                timeForMixerCallback = true;
            }
        }
    }

    {
        CriticalSectionScoped cs(_cbCrit.get());
        if(_mixReceiver != NULL) {
            const AudioFrame** dummy = NULL;
            _mixReceiver->NewMixedAudio(
                _id,
                *mixedAudio,
                dummy,
                0);
        }

        if((_mixerStatusCallback != NULL) &&
            timeForMixerCallback) {
            _mixerStatusCallback->MixedParticipants(
                _id,
                _scratchMixedParticipants,
                static_cast<uint32_t>(_scratchParticipantsToMixAmount));

            _mixerStatusCallback->VADPositiveParticipants(
                _id,
                _scratchVadPositiveParticipants,
                _scratchVadPositiveParticipantsAmount);
            _mixerStatusCallback->MixedAudioLevel(_id,audioLevel);
        }
    }

    // Reclaim all outstanding memory.
    _audioFramePool->PushMemory(mixedAudio);
    ClearAudioFrameList(&mixList);
    ClearAudioFrameList(&rampOutList);
    ClearAudioFrameList(&additionalFramesList);
    {
        CriticalSectionScoped cs(_crit.get());
        _processCalls--;
    }
    return retval;
}

int32_t AudioConferenceMixerImpl::RegisterMixedStreamCallback(
    AudioMixerOutputReceiver& mixReceiver) {
    CriticalSectionScoped cs(_cbCrit.get());
    if(_mixReceiver != NULL) {
        return -1;
    }
    _mixReceiver = &mixReceiver;
    return 0;
}

int32_t AudioConferenceMixerImpl::UnRegisterMixedStreamCallback() {
    CriticalSectionScoped cs(_cbCrit.get());
    if(_mixReceiver == NULL) {
        return -1;
    }
    _mixReceiver = NULL;
    return 0;
}

int32_t AudioConferenceMixerImpl::SetOutputFrequency(
    const Frequency frequency) {
    CriticalSectionScoped cs(_crit.get());

    _outputFrequency = frequency;
    _sampleSize = (_outputFrequency*kProcessPeriodicityInMs) / 1000;

    return 0;
}

AudioConferenceMixer::Frequency
AudioConferenceMixerImpl::OutputFrequency() const {
    CriticalSectionScoped cs(_crit.get());
    return _outputFrequency;
}

int32_t AudioConferenceMixerImpl::RegisterMixerStatusCallback(
    AudioMixerStatusReceiver& mixerStatusCallback,
    const uint32_t amountOf10MsBetweenCallbacks) {
    if(amountOf10MsBetweenCallbacks == 0) {
        WEBRTC_TRACE(
            kTraceWarning,
            kTraceAudioMixerServer,
            _id,
            "amountOf10MsBetweenCallbacks(%d) needs to be larger than 0");
        return -1;
    }
    {
        CriticalSectionScoped cs(_cbCrit.get());
        if(_mixerStatusCallback != NULL) {
            WEBRTC_TRACE(kTraceWarning, kTraceAudioMixerServer, _id,
                         "Mixer status callback already registered");
            return -1;
        }
        _mixerStatusCallback = &mixerStatusCallback;
    }
    {
        CriticalSectionScoped cs(_crit.get());
        _amountOf10MsBetweenCallbacks  = amountOf10MsBetweenCallbacks;
        _amountOf10MsUntilNextCallback = 0;
        _mixerStatusCb                 = true;
    }
    return 0;
}

int32_t AudioConferenceMixerImpl::UnRegisterMixerStatusCallback() {
    {
        CriticalSectionScoped cs(_crit.get());
        if(!_mixerStatusCb)
        {
            WEBRTC_TRACE(kTraceWarning, kTraceAudioMixerServer, _id,
                         "Mixer status callback not registered");
            return -1;
        }
        _mixerStatusCb = false;
    }
    {
        CriticalSectionScoped cs(_cbCrit.get());
        _mixerStatusCallback = NULL;
    }
    return 0;
}

int32_t AudioConferenceMixerImpl::SetMixabilityStatus(
    MixerParticipant& participant,
    bool mixable) {
    if (!mixable) {
        // Anonymous participants are in a separate list. Make sure that the
        // participant is in the _participantList if it is being mixed.
        SetAnonymousMixabilityStatus(participant, false);
    }
    size_t numMixedParticipants;
    {
        CriticalSectionScoped cs(_cbCrit.get());
        const bool isMixed =
            IsParticipantInList(participant, &_participantList);
        // API must be called with a new state.
        if(!(mixable ^ isMixed)) {
            WEBRTC_TRACE(kTraceWarning, kTraceAudioMixerServer, _id,
                         "Mixable is aready %s",
                         isMixed ? "ON" : "off");
            return -1;
        }
        bool success = false;
        if(mixable) {
            success = AddParticipantToList(participant, &_participantList);
        } else {
            success = RemoveParticipantFromList(participant, &_participantList);
        }
        if(!success) {
            WEBRTC_TRACE(kTraceError, kTraceAudioMixerServer, _id,
                         "failed to %s participant",
                         mixable ? "add" : "remove");
            assert(false);
            return -1;
        }

        size_t numMixedNonAnonymous = _participantList.size();
        if (numMixedNonAnonymous > kMaximumAmountOfMixedParticipants) {
            numMixedNonAnonymous = kMaximumAmountOfMixedParticipants;
        }
        numMixedParticipants =
            numMixedNonAnonymous + _additionalParticipantList.size();
    }
    // A MixerParticipant was added or removed. Make sure the scratch
    // buffer is updated if necessary.
    // Note: The scratch buffer may only be updated in Process().
    CriticalSectionScoped cs(_crit.get());
    _numMixedParticipants = numMixedParticipants;
    return 0;
}

int32_t AudioConferenceMixerImpl::MixabilityStatus(
    MixerParticipant& participant,
    bool& mixable) {
    CriticalSectionScoped cs(_cbCrit.get());
    mixable = IsParticipantInList(participant, &_participantList);
    return 0;
}

int32_t AudioConferenceMixerImpl::SetAnonymousMixabilityStatus(
    MixerParticipant& participant, const bool anonymous) {
    CriticalSectionScoped cs(_cbCrit.get());
    if(IsParticipantInList(participant, &_additionalParticipantList)) {
        if(anonymous) {
            return 0;
        }
        if(!RemoveParticipantFromList(participant,
                                      &_additionalParticipantList)) {
            WEBRTC_TRACE(kTraceError, kTraceAudioMixerServer, _id,
                         "unable to remove participant from anonymous list");
            assert(false);
            return -1;
        }
        return AddParticipantToList(participant, &_participantList) ? 0 : -1;
    }
    if(!anonymous) {
        return 0;
    }
    const bool mixable = RemoveParticipantFromList(participant,
                                                   &_participantList);
    if(!mixable) {
        WEBRTC_TRACE(
            kTraceWarning,
            kTraceAudioMixerServer,
            _id,
            "participant must be registered before turning it into anonymous");
        // Setting anonymous status is only possible if MixerParticipant is
        // already registered.
        return -1;
    }
    return AddParticipantToList(participant, &_additionalParticipantList) ?
        0 : -1;
}

int32_t AudioConferenceMixerImpl::AnonymousMixabilityStatus(
    MixerParticipant& participant, bool& mixable) {
    CriticalSectionScoped cs(_cbCrit.get());
    mixable = IsParticipantInList(participant,
                                  &_additionalParticipantList);
    return 0;
}

int32_t AudioConferenceMixerImpl::SetMinimumMixingFrequency(
    Frequency freq) {
    // Make sure that only allowed sampling frequencies are used. Use closest
    // higher sampling frequency to avoid losing information.
    if (static_cast<int>(freq) == 12000) {
         freq = kWbInHz;
    } else if (static_cast<int>(freq) == 24000) {
        freq = kSwbInHz;
    }

    if((freq == kNbInHz) || (freq == kWbInHz) || (freq == kSwbInHz) ||
       (freq == kLowestPossible)) {
        _minimumMixingFreq=freq;
        return 0;
    } else {
        WEBRTC_TRACE(kTraceError, kTraceAudioMixerServer, _id,
                     "SetMinimumMixingFrequency incorrect frequency: %i",freq);
        assert(false);
        return -1;
    }
}

// Check all AudioFrames that are to be mixed. The highest sampling frequency
// found is the lowest that can be used without losing information.
int32_t AudioConferenceMixerImpl::GetLowestMixingFrequency() {
    const int participantListFrequency =
        GetLowestMixingFrequencyFromList(&_participantList);
    const int anonymousListFrequency =
        GetLowestMixingFrequencyFromList(&_additionalParticipantList);
    const int highestFreq =
        (participantListFrequency > anonymousListFrequency) ?
            participantListFrequency : anonymousListFrequency;
    // Check if the user specified a lowest mixing frequency.
    if(_minimumMixingFreq != kLowestPossible) {
        if(_minimumMixingFreq > highestFreq) {
            return _minimumMixingFreq;
        }
    }
    return highestFreq;
}

int32_t AudioConferenceMixerImpl::GetLowestMixingFrequencyFromList(
    MixerParticipantList* mixList) {
    int32_t highestFreq = 8000;
    for (MixerParticipantList::iterator iter = mixList->begin();
         iter != mixList->end();
         ++iter) {
        const int32_t neededFrequency = (*iter)->NeededFrequency(_id);
        if(neededFrequency > highestFreq) {
            highestFreq = neededFrequency;
        }
    }
    return highestFreq;
}

void AudioConferenceMixerImpl::UpdateToMix(
    AudioFrameList* mixList,
    AudioFrameList* rampOutList,
    std::map<int, MixerParticipant*>* mixParticipantList,
    size_t& maxAudioFrameCounter) {
    WEBRTC_TRACE(kTraceStream, kTraceAudioMixerServer, _id,
                 "UpdateToMix(mixList,rampOutList,mixParticipantList,%d)",
                 maxAudioFrameCounter);
    const size_t mixListStartSize = mixList->size();
    AudioFrameList activeList;
    // Struct needed by the passive lists to keep track of which AudioFrame
    // belongs to which MixerParticipant.
    ParticipantFramePairList passiveWasNotMixedList;
    ParticipantFramePairList passiveWasMixedList;
    for (MixerParticipantList::iterator participant = _participantList.begin();
         participant != _participantList.end();
         ++participant) {
        // Stop keeping track of passive participants if there are already
        // enough participants available (they wont be mixed anyway).
        bool mustAddToPassiveList = (maxAudioFrameCounter >
                                    (activeList.size() +
                                     passiveWasMixedList.size() +
                                     passiveWasNotMixedList.size()));

        bool wasMixed = false;
        (*participant)->_mixHistory->WasMixed(wasMixed);
        AudioFrame* audioFrame = NULL;
        if(_audioFramePool->PopMemory(audioFrame) == -1) {
            WEBRTC_TRACE(kTraceMemory, kTraceAudioMixerServer, _id,
                         "failed PopMemory() call");
            assert(false);
            return;
        }
        audioFrame->sample_rate_hz_ = _outputFrequency;

        if((*participant)->GetAudioFrame(_id,*audioFrame) != 0) {
            WEBRTC_TRACE(kTraceWarning, kTraceAudioMixerServer, _id,
                         "failed to GetAudioFrame() from participant");
            _audioFramePool->PushMemory(audioFrame);
            continue;
        }
        if (_participantList.size() != 1) {
          // TODO(wu): Issue 3390, add support for multiple participants case.
          audioFrame->ntp_time_ms_ = -1;
        }

        // TODO(henrike): this assert triggers in some test cases where SRTP is
        // used which prevents NetEQ from making a VAD. Temporarily disable this
        // assert until the problem is fixed on a higher level.
        // assert(audioFrame->vad_activity_ != AudioFrame::kVadUnknown);
        if (audioFrame->vad_activity_ == AudioFrame::kVadUnknown) {
            WEBRTC_TRACE(kTraceWarning, kTraceAudioMixerServer, _id,
                         "invalid VAD state from participant");
        }

        if(audioFrame->vad_activity_ == AudioFrame::kVadActive) {
            if(!wasMixed) {
                RampIn(*audioFrame);
            }

            if(activeList.size() >= maxAudioFrameCounter) {
                // There are already more active participants than should be
                // mixed. Only keep the ones with the highest energy.
                AudioFrameList::iterator replaceItem;
                CalculateEnergy(*audioFrame);
                uint32_t lowestEnergy = audioFrame->energy_;

                bool found_replace_item = false;
                for (AudioFrameList::iterator iter = activeList.begin();
                     iter != activeList.end();
                     ++iter) {
                    CalculateEnergy(**iter);
                    if((*iter)->energy_ < lowestEnergy) {
                        replaceItem = iter;
                        lowestEnergy = (*iter)->energy_;
                        found_replace_item = true;
                    }
                }
                if(found_replace_item) {
                    AudioFrame* replaceFrame = *replaceItem;

                    bool replaceWasMixed = false;
                    std::map<int, MixerParticipant*>::iterator it =
                        mixParticipantList->find(replaceFrame->id_);

                    // When a frame is pushed to |activeList| it is also pushed
                    // to mixParticipantList with the frame's id. This means
                    // that the Find call above should never fail.
                    assert(it != mixParticipantList->end());
                    it->second->_mixHistory->WasMixed(replaceWasMixed);

                    mixParticipantList->erase(replaceFrame->id_);
                    activeList.erase(replaceItem);

                    activeList.push_front(audioFrame);
                    (*mixParticipantList)[audioFrame->id_] = *participant;
                    assert(mixParticipantList->size() <=
                           kMaximumAmountOfMixedParticipants);

                    if (replaceWasMixed) {
                      RampOut(*replaceFrame);
                      rampOutList->push_back(replaceFrame);
                      assert(rampOutList->size() <=
                             kMaximumAmountOfMixedParticipants);
                    } else {
                      _audioFramePool->PushMemory(replaceFrame);
                    }
                } else {
                    if(wasMixed) {
                        RampOut(*audioFrame);
                        rampOutList->push_back(audioFrame);
                        assert(rampOutList->size() <=
                               kMaximumAmountOfMixedParticipants);
                    } else {
                        _audioFramePool->PushMemory(audioFrame);
                    }
                }
            } else {
                activeList.push_front(audioFrame);
                (*mixParticipantList)[audioFrame->id_] = *participant;
                assert(mixParticipantList->size() <=
                       kMaximumAmountOfMixedParticipants);
            }
        } else {
            if(wasMixed) {
                ParticipantFramePair* pair = new ParticipantFramePair;
                pair->audioFrame  = audioFrame;
                pair->participant = *participant;
                passiveWasMixedList.push_back(pair);
            } else if(mustAddToPassiveList) {
                RampIn(*audioFrame);
                ParticipantFramePair* pair = new ParticipantFramePair;
                pair->audioFrame  = audioFrame;
                pair->participant = *participant;
                passiveWasNotMixedList.push_back(pair);
            } else {
                _audioFramePool->PushMemory(audioFrame);
            }
        }
    }
    assert(activeList.size() <= maxAudioFrameCounter);
    // At this point it is known which participants should be mixed. Transfer
    // this information to this functions output parameters.
    for (AudioFrameList::iterator iter = activeList.begin();
         iter != activeList.end();
         ++iter) {
        mixList->push_back(*iter);
    }
    activeList.clear();
    // Always mix a constant number of AudioFrames. If there aren't enough
    // active participants mix passive ones. Starting with those that was mixed
    // last iteration.
    for (ParticipantFramePairList::iterator iter = passiveWasMixedList.begin();
         iter != passiveWasMixedList.end();
         ++iter) {
        if(mixList->size() < maxAudioFrameCounter + mixListStartSize) {
            mixList->push_back((*iter)->audioFrame);
            (*mixParticipantList)[(*iter)->audioFrame->id_] =
                (*iter)->participant;
            assert(mixParticipantList->size() <=
                   kMaximumAmountOfMixedParticipants);
        } else {
            _audioFramePool->PushMemory((*iter)->audioFrame);
        }
        delete *iter;
    }
    // And finally the ones that have not been mixed for a while.
    for (ParticipantFramePairList::iterator iter =
             passiveWasNotMixedList.begin();
         iter != passiveWasNotMixedList.end();
         ++iter) {
        if(mixList->size() <  maxAudioFrameCounter + mixListStartSize) {
          mixList->push_back((*iter)->audioFrame);
            (*mixParticipantList)[(*iter)->audioFrame->id_] =
                (*iter)->participant;
            assert(mixParticipantList->size() <=
                   kMaximumAmountOfMixedParticipants);
        } else {
            _audioFramePool->PushMemory((*iter)->audioFrame);
        }
        delete *iter;
    }
    assert(maxAudioFrameCounter + mixListStartSize >= mixList->size());
    maxAudioFrameCounter += mixListStartSize - mixList->size();
}

void AudioConferenceMixerImpl::GetAdditionalAudio(
    AudioFrameList* additionalFramesList) {
    WEBRTC_TRACE(kTraceStream, kTraceAudioMixerServer, _id,
                 "GetAdditionalAudio(additionalFramesList)");
    // The GetAudioFrame() callback may result in the participant being removed
    // from additionalParticipantList_. If that happens it will invalidate any
    // iterators. Create a copy of the participants list such that the list of
    // participants can be traversed safely.
    MixerParticipantList additionalParticipantList;
    additionalParticipantList.insert(additionalParticipantList.begin(),
                                     _additionalParticipantList.begin(),
                                     _additionalParticipantList.end());

    for (MixerParticipantList::iterator participant =
             additionalParticipantList.begin();
         participant != additionalParticipantList.end();
         ++participant) {
        AudioFrame* audioFrame = NULL;
        if(_audioFramePool->PopMemory(audioFrame) == -1) {
            WEBRTC_TRACE(kTraceMemory, kTraceAudioMixerServer, _id,
                         "failed PopMemory() call");
            assert(false);
            return;
        }
        audioFrame->sample_rate_hz_ = _outputFrequency;
        if((*participant)->GetAudioFrame(_id, *audioFrame) != 0) {
            WEBRTC_TRACE(kTraceWarning, kTraceAudioMixerServer, _id,
                         "failed to GetAudioFrame() from participant");
            _audioFramePool->PushMemory(audioFrame);
            continue;
        }
        if(audioFrame->samples_per_channel_ == 0) {
            // Empty frame. Don't use it.
            _audioFramePool->PushMemory(audioFrame);
            continue;
        }
        additionalFramesList->push_back(audioFrame);
    }
}

void AudioConferenceMixerImpl::UpdateMixedStatus(
    std::map<int, MixerParticipant*>& mixedParticipantsMap) {
    WEBRTC_TRACE(kTraceStream, kTraceAudioMixerServer, _id,
                 "UpdateMixedStatus(mixedParticipantsMap)");
    assert(mixedParticipantsMap.size() <= kMaximumAmountOfMixedParticipants);

    // Loop through all participants. If they are in the mix map they
    // were mixed.
    for (MixerParticipantList::iterator participant = _participantList.begin();
         participant != _participantList.end();
         ++participant) {
        bool isMixed = false;
        for (std::map<int, MixerParticipant*>::iterator it =
                 mixedParticipantsMap.begin();
             it != mixedParticipantsMap.end();
             ++it) {
          if (it->second == *participant) {
            isMixed = true;
            break;
          }
        }
        (*participant)->_mixHistory->SetIsMixed(isMixed);
    }
}

void AudioConferenceMixerImpl::ClearAudioFrameList(
    AudioFrameList* audioFrameList) {
    WEBRTC_TRACE(kTraceStream, kTraceAudioMixerServer, _id,
                 "ClearAudioFrameList(audioFrameList)");
    for (AudioFrameList::iterator iter = audioFrameList->begin();
         iter != audioFrameList->end();
         ++iter) {
        _audioFramePool->PushMemory(*iter);
    }
    audioFrameList->clear();
}

void AudioConferenceMixerImpl::UpdateVADPositiveParticipants(
    AudioFrameList* mixList) {
    WEBRTC_TRACE(kTraceStream, kTraceAudioMixerServer, _id,
                 "UpdateVADPositiveParticipants(mixList)");

    for (AudioFrameList::iterator iter = mixList->begin();
         iter != mixList->end();
         ++iter) {
        CalculateEnergy(**iter);
        if((*iter)->vad_activity_ == AudioFrame::kVadActive) {
            _scratchVadPositiveParticipants[
                _scratchVadPositiveParticipantsAmount].participant =
                (*iter)->id_;
            // TODO(andrew): to what should this be set?
            _scratchVadPositiveParticipants[
                _scratchVadPositiveParticipantsAmount].level = 0;
            _scratchVadPositiveParticipantsAmount++;
        }
    }
}

bool AudioConferenceMixerImpl::IsParticipantInList(
    MixerParticipant& participant,
    MixerParticipantList* participantList) const {
    WEBRTC_TRACE(kTraceStream, kTraceAudioMixerServer, _id,
                 "IsParticipantInList(participant,participantList)");
    for (MixerParticipantList::const_iterator iter = participantList->begin();
         iter != participantList->end();
         ++iter) {
        if(&participant == *iter) {
            return true;
        }
    }
    return false;
}

bool AudioConferenceMixerImpl::AddParticipantToList(
    MixerParticipant& participant,
    MixerParticipantList* participantList) {
    WEBRTC_TRACE(kTraceStream, kTraceAudioMixerServer, _id,
                 "AddParticipantToList(participant, participantList)");
    participantList->push_back(&participant);
    // Make sure that the mixed status is correct for new MixerParticipant.
    participant._mixHistory->ResetMixedStatus();
    return true;
}

bool AudioConferenceMixerImpl::RemoveParticipantFromList(
    MixerParticipant& participant,
    MixerParticipantList* participantList) {
    WEBRTC_TRACE(kTraceStream, kTraceAudioMixerServer, _id,
                 "RemoveParticipantFromList(participant, participantList)");
    for (MixerParticipantList::iterator iter = participantList->begin();
         iter != participantList->end();
         ++iter) {
        if(*iter == &participant) {
            participantList->erase(iter);
            // Participant is no longer mixed, reset to default.
            participant._mixHistory->ResetMixedStatus();
            return true;
        }
    }
    return false;
}

int32_t AudioConferenceMixerImpl::MixFromList(
    AudioFrame& mixedAudio,
    const AudioFrameList* audioFrameList) {
    WEBRTC_TRACE(kTraceStream, kTraceAudioMixerServer, _id,
                 "MixFromList(mixedAudio, audioFrameList)");
    if(audioFrameList->empty()) return 0;

    uint32_t position = 0;

    if (_numMixedParticipants == 1) {
      mixedAudio.timestamp_ = audioFrameList->front()->timestamp_;
      mixedAudio.elapsed_time_ms_ = audioFrameList->front()->elapsed_time_ms_;
    } else {
      // TODO(wu): Issue 3390.
      // Audio frame timestamp is only supported in one channel case.
      mixedAudio.timestamp_ = 0;
      mixedAudio.elapsed_time_ms_ = -1;
    }

    for (AudioFrameList::const_iterator iter = audioFrameList->begin();
         iter != audioFrameList->end();
         ++iter) {
        if(position >= kMaximumAmountOfMixedParticipants) {
            WEBRTC_TRACE(
                kTraceMemory,
                kTraceAudioMixerServer,
                _id,
                "Trying to mix more than max amount of mixed participants:%d!",
                kMaximumAmountOfMixedParticipants);
            // Assert and avoid crash
            assert(false);
            position = 0;
        }
        MixFrames(&mixedAudio, (*iter), use_limiter_);

        SetParticipantStatistics(&_scratchMixedParticipants[position],
                                 **iter);

        position++;
    }

    return 0;
}

// TODO(andrew): consolidate this function with MixFromList.
int32_t AudioConferenceMixerImpl::MixAnonomouslyFromList(
    AudioFrame& mixedAudio,
    const AudioFrameList* audioFrameList) {
    WEBRTC_TRACE(kTraceStream, kTraceAudioMixerServer, _id,
                 "MixAnonomouslyFromList(mixedAudio, audioFrameList)");

    if(audioFrameList->empty()) return 0;

    for (AudioFrameList::const_iterator iter = audioFrameList->begin();
         iter != audioFrameList->end();
         ++iter) {
        MixFrames(&mixedAudio, *iter, use_limiter_);
    }
    return 0;
}

bool AudioConferenceMixerImpl::LimitMixedAudio(AudioFrame& mixedAudio) {
    if (!use_limiter_) {
      return true;
    }

    // Smoothly limit the mixed frame.
    const int error = _limiter->ProcessStream(&mixedAudio);

    // And now we can safely restore the level. This procedure results in
    // some loss of resolution, deemed acceptable.
    //
    // It's possible to apply the gain in the AGC (with a target level of 0 dbFS
    // and compression gain of 6 dB). However, in the transition frame when this
    // is enabled (moving from one to two participants) it has the potential to
    // create discontinuities in the mixed frame.
    //
    // Instead we double the frame (with addition since left-shifting a
    // negative value is undefined).
    mixedAudio += mixedAudio;

    if(error != _limiter->kNoError) {
        WEBRTC_TRACE(kTraceError, kTraceAudioMixerServer, _id,
                     "Error from AudioProcessing: %d", error);
        assert(false);
        return false;
    }
    return true;
}
}  // namespace webrtc