File: EchoCanceller.cpp

package info (click to toggle)
libtgvoip 2.4.2-1
  • links: PTS, VCS
  • area: main
  • in suites: buster
  • size: 9,004 kB
  • sloc: cpp: 61,331; ansic: 24,725; sh: 4,150; makefile: 749; asm: 242
file content (235 lines) | stat: -rwxr-xr-x 5,951 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
//
// libtgvoip is free and unencumbered public domain software.
// For more information, see http://unlicense.org or the UNLICENSE file
// you should have received with this source code distribution.
//

#ifndef TGVOIP_NO_DSP
#include "webrtc_dsp/modules/audio_processing/include/audio_processing.h"
#include "webrtc_dsp/api/audio/audio_frame.h"
#endif

#include "EchoCanceller.h"
#include "audio/AudioOutput.h"
#include "audio/AudioInput.h"
#include "logging.h"
#include "VoIPServerConfig.h"
#include <string.h>
#include <stdio.h>

using namespace tgvoip;

EchoCanceller::EchoCanceller(bool enableAEC, bool enableNS, bool enableAGC){
#ifndef TGVOIP_NO_DSP
	this->enableAEC=enableAEC;
	this->enableAGC=enableAGC;
	this->enableNS=enableNS;
	isOn=true;

	webrtc::Config extraConfig;
#ifdef TGVOIP_USE_DESKTOP_DSP
	extraConfig.Set(new webrtc::DelayAgnostic(true));
#endif

	apm=webrtc::AudioProcessingBuilder().Create(extraConfig);

	webrtc::AudioProcessing::Config config;
	config.echo_canceller.enabled = enableAEC;
#ifndef TGVOIP_USE_DESKTOP_DSP
	config.echo_canceller.mobile_mode = true;
#else
	config.echo_canceller.mobile_mode = false;
#endif
	config.high_pass_filter.enabled = enableAEC;
	config.gain_controller2.enabled = enableAGC;
	apm->ApplyConfig(config);
	
	webrtc::NoiseSuppression::Level nsLevel;
#ifdef __APPLE__
	switch(ServerConfig::GetSharedInstance()->GetInt("webrtc_ns_level_vpio", 0)){
#else
	switch(ServerConfig::GetSharedInstance()->GetInt("webrtc_ns_level", 2)){
#endif
		case 0:
			nsLevel=webrtc::NoiseSuppression::Level::kLow;
			break;
		case 1:
			nsLevel=webrtc::NoiseSuppression::Level::kModerate;
			break;
		case 3:
			nsLevel=webrtc::NoiseSuppression::Level::kVeryHigh;
			break;
		case 2:
		default:
			nsLevel=webrtc::NoiseSuppression::Level::kHigh;
			break;
	}
	apm->noise_suppression()->set_level(nsLevel);
	apm->noise_suppression()->Enable(enableNS);
	if(enableAGC){
		apm->gain_control()->set_mode(webrtc::GainControl::Mode::kAdaptiveDigital);
		apm->gain_control()->set_target_level_dbfs(ServerConfig::GetSharedInstance()->GetInt("webrtc_agc_target_level", 9));
		apm->gain_control()->enable_limiter(ServerConfig::GetSharedInstance()->GetBoolean("webrtc_agc_enable_limiter", true));
		apm->gain_control()->set_compression_gain_db(ServerConfig::GetSharedInstance()->GetInt("webrtc_agc_compression_gain", 20));
	}
	apm->voice_detection()->set_likelihood(webrtc::VoiceDetection::Likelihood::kVeryLowLikelihood);

	audioFrame=new webrtc::AudioFrame();
	audioFrame->samples_per_channel_=480;
	audioFrame->sample_rate_hz_=48000;
	audioFrame->num_channels_=1;

	farendQueue=new BlockingQueue<int16_t*>(11);
	farendBufferPool=new BufferPool(960*2, 10);
	running=true;
	bufferFarendThread=new Thread(std::bind(&EchoCanceller::RunBufferFarendThread, this));
	bufferFarendThread->Start();

#else
	this->enableAEC=this->enableAGC=enableAGC=this->enableNS=enableNS=false;
	isOn=true;
#endif
}

EchoCanceller::~EchoCanceller(){
#ifndef TGVOIP_NO_DSP
	delete apm;
	delete audioFrame;
#endif
}

void EchoCanceller::Start(){

}

void EchoCanceller::Stop(){

}


void EchoCanceller::SpeakerOutCallback(unsigned char* data, size_t len){
    if(len!=960*2 || !enableAEC || !isOn)
		return;
#ifndef TGVOIP_NO_DSP
	int16_t* buf=(int16_t*)farendBufferPool->Get();
	if(buf){
		memcpy(buf, data, 960*2);
		farendQueue->Put(buf);
	}
#endif
}

#ifndef TGVOIP_NO_DSP
void EchoCanceller::RunBufferFarendThread(){
	webrtc::AudioFrame frame;
	frame.num_channels_=1;
	frame.sample_rate_hz_=48000;
	frame.samples_per_channel_=480;
	while(running){
		int16_t* samplesIn=farendQueue->GetBlocking();
		if(samplesIn){
			memcpy(frame.mutable_data(), samplesIn, 480*2);
			apm->ProcessReverseStream(&frame);
			memcpy(frame.mutable_data(), samplesIn+480, 480*2);
			apm->ProcessReverseStream(&frame);
			didBufferFarend=true;
			farendBufferPool->Reuse(reinterpret_cast<unsigned char*>(samplesIn));
		}
	}
}
#endif

void EchoCanceller::Enable(bool enabled){
	isOn=enabled;
}

void EchoCanceller::ProcessInput(int16_t* inOut, size_t numSamples, bool& hasVoice){
	if(!isOn || (!enableAEC && !enableAGC && !enableNS)){
		return;
	}
	int delay=audio::AudioInput::GetEstimatedDelay()+audio::AudioOutput::GetEstimatedDelay();
	assert(numSamples==960);

	memcpy(audioFrame->mutable_data(), inOut, 480*2);
	if(enableAEC)
    	apm->set_stream_delay_ms(delay);
	apm->ProcessStream(audioFrame);
	if(enableVAD)
    	hasVoice=apm->voice_detection()->stream_has_voice();
	memcpy(inOut, audioFrame->data(), 480*2);
	memcpy(audioFrame->mutable_data(), inOut+480, 480*2);
	if(enableAEC)
    	apm->set_stream_delay_ms(delay);
	apm->ProcessStream(audioFrame);
	if(enableVAD){
    	hasVoice=hasVoice || apm->voice_detection()->stream_has_voice();
	}
	memcpy(inOut+480, audioFrame->data(), 480*2);
}

void EchoCanceller::SetAECStrength(int strength){
#ifndef TGVOIP_NO_DSP
	/*if(aec){
#ifndef TGVOIP_USE_DESKTOP_DSP
		AecmConfig cfg;
		cfg.cngMode=AecmFalse;
		cfg.echoMode=(int16_t) strength;
		WebRtcAecm_set_config(aec, cfg);
#endif
	}*/
#endif
}

void EchoCanceller::SetVoiceDetectionEnabled(bool enabled){
	enableVAD=enabled;
	apm->voice_detection()->Enable(enabled);
}

using namespace tgvoip::effects;

AudioEffect::~AudioEffect(){

}

void AudioEffect::SetPassThrough(bool passThrough){
	this->passThrough=passThrough;
}

Volume::Volume(){

}

Volume::~Volume(){

}

void Volume::Process(int16_t* inOut, size_t numSamples){
	if(level==1.0f || passThrough){
		return;
	}
	for(size_t i=0;i<numSamples;i++){
		float sample=(float)inOut[i]*multiplier;
		if(sample>32767.0f)
			inOut[i]=INT16_MAX;
		else if(sample<-32768.0f)
			inOut[i]=INT16_MIN;
		else
			inOut[i]=(int16_t)sample;
	}
}

void Volume::SetLevel(float level){
	this->level=level;
	float db;
	if(level<1.0f)
		db=-50.0f*(1.0f-level);
	else if(level>1.0f && level<=2.0f)
		db=10.0f*(level-1.0f);
	else
		db=0.0f;
	multiplier=expf(db/20.0f * logf(10.0f));
}

float Volume::GetLevel(){
	return level;
}