henrike@webrtc.org | 28e2075 | 2013-07-10 00:45:36 +0000 | [diff] [blame^] | 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 | #ifndef TALK_MEDIA_BASE_MEDIACHANNEL_H_ |
| 29 | #define TALK_MEDIA_BASE_MEDIACHANNEL_H_ |
| 30 | |
| 31 | #include <string> |
| 32 | #include <vector> |
| 33 | |
| 34 | #include "talk/base/basictypes.h" |
| 35 | #include "talk/base/buffer.h" |
| 36 | #include "talk/base/logging.h" |
| 37 | #include "talk/base/sigslot.h" |
| 38 | #include "talk/base/socket.h" |
| 39 | #include "talk/base/window.h" |
| 40 | #include "talk/media/base/codec.h" |
| 41 | #include "talk/media/base/constants.h" |
| 42 | #include "talk/media/base/streamparams.h" |
| 43 | // TODO(juberti): re-evaluate this include |
| 44 | #include "talk/session/media/audiomonitor.h" |
| 45 | |
| 46 | namespace talk_base { |
| 47 | class Buffer; |
| 48 | class RateLimiter; |
| 49 | class Timing; |
| 50 | } |
| 51 | |
| 52 | namespace cricket { |
| 53 | |
| 54 | class AudioRenderer; |
| 55 | struct RtpHeader; |
| 56 | class ScreencastId; |
| 57 | struct VideoFormat; |
| 58 | class VideoCapturer; |
| 59 | class VideoRenderer; |
| 60 | |
| 61 | const int kMinRtpHeaderExtensionId = 1; |
| 62 | const int kMaxRtpHeaderExtensionId = 255; |
| 63 | const int kScreencastDefaultFps = 5; |
| 64 | |
| 65 | // Used in AudioOptions and VideoOptions to signify "unset" values. |
| 66 | template <class T> |
| 67 | class Settable { |
| 68 | public: |
| 69 | Settable() : set_(false), val_() {} |
| 70 | explicit Settable(T val) : set_(true), val_(val) {} |
| 71 | |
| 72 | bool IsSet() const { |
| 73 | return set_; |
| 74 | } |
| 75 | |
| 76 | bool Get(T* out) const { |
| 77 | *out = val_; |
| 78 | return set_; |
| 79 | } |
| 80 | |
| 81 | T GetWithDefaultIfUnset(const T& default_value) const { |
| 82 | return set_ ? val_ : default_value; |
| 83 | } |
| 84 | |
| 85 | virtual void Set(T val) { |
| 86 | set_ = true; |
| 87 | val_ = val; |
| 88 | } |
| 89 | |
| 90 | void Clear() { |
| 91 | Set(T()); |
| 92 | set_ = false; |
| 93 | } |
| 94 | |
| 95 | void SetFrom(const Settable<T>& o) { |
| 96 | // Set this value based on the value of o, iff o is set. If this value is |
| 97 | // set and o is unset, the current value will be unchanged. |
| 98 | T val; |
| 99 | if (o.Get(&val)) { |
| 100 | Set(val); |
| 101 | } |
| 102 | } |
| 103 | |
| 104 | std::string ToString() const { |
| 105 | return set_ ? talk_base::ToString(val_) : ""; |
| 106 | } |
| 107 | |
| 108 | bool operator==(const Settable<T>& o) const { |
| 109 | // Equal if both are unset with any value or both set with the same value. |
| 110 | return (set_ == o.set_) && (!set_ || (val_ == o.val_)); |
| 111 | } |
| 112 | |
| 113 | bool operator!=(const Settable<T>& o) const { |
| 114 | return !operator==(o); |
| 115 | } |
| 116 | |
| 117 | protected: |
| 118 | void InitializeValue(const T &val) { |
| 119 | val_ = val; |
| 120 | } |
| 121 | |
| 122 | private: |
| 123 | bool set_; |
| 124 | T val_; |
| 125 | }; |
| 126 | |
| 127 | class SettablePercent : public Settable<float> { |
| 128 | public: |
| 129 | virtual void Set(float val) { |
| 130 | if (val < 0) { |
| 131 | val = 0; |
| 132 | } |
| 133 | if (val > 1.0) { |
| 134 | val = 1.0; |
| 135 | } |
| 136 | Settable<float>::Set(val); |
| 137 | } |
| 138 | }; |
| 139 | |
| 140 | template <class T> |
| 141 | static std::string ToStringIfSet(const char* key, const Settable<T>& val) { |
| 142 | std::string str; |
| 143 | if (val.IsSet()) { |
| 144 | str = key; |
| 145 | str += ": "; |
| 146 | str += val.ToString(); |
| 147 | str += ", "; |
| 148 | } |
| 149 | return str; |
| 150 | } |
| 151 | |
| 152 | // Options that can be applied to a VoiceMediaChannel or a VoiceMediaEngine. |
| 153 | // Used to be flags, but that makes it hard to selectively apply options. |
| 154 | // We are moving all of the setting of options to structs like this, |
| 155 | // but some things currently still use flags. |
| 156 | struct AudioOptions { |
| 157 | void SetAll(const AudioOptions& change) { |
| 158 | echo_cancellation.SetFrom(change.echo_cancellation); |
| 159 | auto_gain_control.SetFrom(change.auto_gain_control); |
| 160 | noise_suppression.SetFrom(change.noise_suppression); |
| 161 | highpass_filter.SetFrom(change.highpass_filter); |
| 162 | stereo_swapping.SetFrom(change.stereo_swapping); |
| 163 | typing_detection.SetFrom(change.typing_detection); |
| 164 | conference_mode.SetFrom(change.conference_mode); |
| 165 | adjust_agc_delta.SetFrom(change.adjust_agc_delta); |
| 166 | experimental_agc.SetFrom(change.experimental_agc); |
| 167 | experimental_aec.SetFrom(change.experimental_aec); |
| 168 | aec_dump.SetFrom(change.aec_dump); |
| 169 | } |
| 170 | |
| 171 | bool operator==(const AudioOptions& o) const { |
| 172 | return echo_cancellation == o.echo_cancellation && |
| 173 | auto_gain_control == o.auto_gain_control && |
| 174 | noise_suppression == o.noise_suppression && |
| 175 | highpass_filter == o.highpass_filter && |
| 176 | stereo_swapping == o.stereo_swapping && |
| 177 | typing_detection == o.typing_detection && |
| 178 | conference_mode == o.conference_mode && |
| 179 | experimental_agc == o.experimental_agc && |
| 180 | experimental_aec == o.experimental_aec && |
| 181 | adjust_agc_delta == o.adjust_agc_delta && |
| 182 | aec_dump == o.aec_dump; |
| 183 | } |
| 184 | |
| 185 | std::string ToString() const { |
| 186 | std::ostringstream ost; |
| 187 | ost << "AudioOptions {"; |
| 188 | ost << ToStringIfSet("aec", echo_cancellation); |
| 189 | ost << ToStringIfSet("agc", auto_gain_control); |
| 190 | ost << ToStringIfSet("ns", noise_suppression); |
| 191 | ost << ToStringIfSet("hf", highpass_filter); |
| 192 | ost << ToStringIfSet("swap", stereo_swapping); |
| 193 | ost << ToStringIfSet("typing", typing_detection); |
| 194 | ost << ToStringIfSet("conference", conference_mode); |
| 195 | ost << ToStringIfSet("agc_delta", adjust_agc_delta); |
| 196 | ost << ToStringIfSet("experimental_agc", experimental_agc); |
| 197 | ost << ToStringIfSet("experimental_aec", experimental_aec); |
| 198 | ost << ToStringIfSet("aec_dump", aec_dump); |
| 199 | ost << "}"; |
| 200 | return ost.str(); |
| 201 | } |
| 202 | |
| 203 | // Audio processing that attempts to filter away the output signal from |
| 204 | // later inbound pickup. |
| 205 | Settable<bool> echo_cancellation; |
| 206 | // Audio processing to adjust the sensitivity of the local mic dynamically. |
| 207 | Settable<bool> auto_gain_control; |
| 208 | // Audio processing to filter out background noise. |
| 209 | Settable<bool> noise_suppression; |
| 210 | // Audio processing to remove background noise of lower frequencies. |
| 211 | Settable<bool> highpass_filter; |
| 212 | // Audio processing to swap the left and right channels. |
| 213 | Settable<bool> stereo_swapping; |
| 214 | // Audio processing to detect typing. |
| 215 | Settable<bool> typing_detection; |
| 216 | Settable<bool> conference_mode; |
| 217 | Settable<int> adjust_agc_delta; |
| 218 | Settable<bool> experimental_agc; |
| 219 | Settable<bool> experimental_aec; |
| 220 | Settable<bool> aec_dump; |
| 221 | }; |
| 222 | |
| 223 | // Options that can be applied to a VideoMediaChannel or a VideoMediaEngine. |
| 224 | // Used to be flags, but that makes it hard to selectively apply options. |
| 225 | // We are moving all of the setting of options to structs like this, |
| 226 | // but some things currently still use flags. |
| 227 | struct VideoOptions { |
| 228 | VideoOptions() { |
| 229 | process_adaptation_threshhold.Set(kProcessCpuThreshold); |
| 230 | system_low_adaptation_threshhold.Set(kLowSystemCpuThreshold); |
| 231 | system_high_adaptation_threshhold.Set(kHighSystemCpuThreshold); |
| 232 | } |
| 233 | |
| 234 | void SetAll(const VideoOptions& change) { |
| 235 | adapt_input_to_encoder.SetFrom(change.adapt_input_to_encoder); |
| 236 | adapt_input_to_cpu_usage.SetFrom(change.adapt_input_to_cpu_usage); |
| 237 | adapt_view_switch.SetFrom(change.adapt_view_switch); |
| 238 | video_noise_reduction.SetFrom(change.video_noise_reduction); |
| 239 | video_three_layers.SetFrom(change.video_three_layers); |
| 240 | video_enable_camera_list.SetFrom(change.video_enable_camera_list); |
| 241 | video_one_layer_screencast.SetFrom(change.video_one_layer_screencast); |
| 242 | video_high_bitrate.SetFrom(change.video_high_bitrate); |
| 243 | video_watermark.SetFrom(change.video_watermark); |
| 244 | video_temporal_layer_screencast.SetFrom( |
| 245 | change.video_temporal_layer_screencast); |
| 246 | video_leaky_bucket.SetFrom(change.video_leaky_bucket); |
| 247 | conference_mode.SetFrom(change.conference_mode); |
| 248 | process_adaptation_threshhold.SetFrom(change.process_adaptation_threshhold); |
| 249 | system_low_adaptation_threshhold.SetFrom( |
| 250 | change.system_low_adaptation_threshhold); |
| 251 | system_high_adaptation_threshhold.SetFrom( |
| 252 | change.system_high_adaptation_threshhold); |
| 253 | buffered_mode_latency.SetFrom(change.buffered_mode_latency); |
| 254 | } |
| 255 | |
| 256 | bool operator==(const VideoOptions& o) const { |
| 257 | return adapt_input_to_encoder == o.adapt_input_to_encoder && |
| 258 | adapt_input_to_cpu_usage == o.adapt_input_to_cpu_usage && |
| 259 | adapt_view_switch == o.adapt_view_switch && |
| 260 | video_noise_reduction == o.video_noise_reduction && |
| 261 | video_three_layers == o.video_three_layers && |
| 262 | video_enable_camera_list == o.video_enable_camera_list && |
| 263 | video_one_layer_screencast == o.video_one_layer_screencast && |
| 264 | video_high_bitrate == o.video_high_bitrate && |
| 265 | video_watermark == o.video_watermark && |
| 266 | video_temporal_layer_screencast == o.video_temporal_layer_screencast && |
| 267 | video_leaky_bucket == o.video_leaky_bucket && |
| 268 | conference_mode == o.conference_mode && |
| 269 | process_adaptation_threshhold == o.process_adaptation_threshhold && |
| 270 | system_low_adaptation_threshhold == |
| 271 | o.system_low_adaptation_threshhold && |
| 272 | system_high_adaptation_threshhold == |
| 273 | o.system_high_adaptation_threshhold && |
| 274 | buffered_mode_latency == o.buffered_mode_latency; |
| 275 | } |
| 276 | |
| 277 | std::string ToString() const { |
| 278 | std::ostringstream ost; |
| 279 | ost << "VideoOptions {"; |
| 280 | ost << ToStringIfSet("encoder adaption", adapt_input_to_encoder); |
| 281 | ost << ToStringIfSet("cpu adaption", adapt_input_to_cpu_usage); |
| 282 | ost << ToStringIfSet("adapt view switch", adapt_view_switch); |
| 283 | ost << ToStringIfSet("noise reduction", video_noise_reduction); |
| 284 | ost << ToStringIfSet("3 layers", video_three_layers); |
| 285 | ost << ToStringIfSet("camera list", video_enable_camera_list); |
| 286 | ost << ToStringIfSet("1 layer screencast", |
| 287 | video_one_layer_screencast); |
| 288 | ost << ToStringIfSet("high bitrate", video_high_bitrate); |
| 289 | ost << ToStringIfSet("watermark", video_watermark); |
| 290 | ost << ToStringIfSet("video temporal layer screencast", |
| 291 | video_temporal_layer_screencast); |
| 292 | ost << ToStringIfSet("leaky bucket", video_leaky_bucket); |
| 293 | ost << ToStringIfSet("conference mode", conference_mode); |
| 294 | ost << ToStringIfSet("process", process_adaptation_threshhold); |
| 295 | ost << ToStringIfSet("low", system_low_adaptation_threshhold); |
| 296 | ost << ToStringIfSet("high", system_high_adaptation_threshhold); |
| 297 | ost << ToStringIfSet("buffered mode latency", buffered_mode_latency); |
| 298 | ost << "}"; |
| 299 | return ost.str(); |
| 300 | } |
| 301 | |
| 302 | // Encoder adaption, which is the gd callback in LMI, and TBA in WebRTC. |
| 303 | Settable<bool> adapt_input_to_encoder; |
| 304 | // Enable CPU adaptation? |
| 305 | Settable<bool> adapt_input_to_cpu_usage; |
| 306 | // Enable Adapt View Switch? |
| 307 | Settable<bool> adapt_view_switch; |
| 308 | // Enable denoising? |
| 309 | Settable<bool> video_noise_reduction; |
| 310 | // Experimental: Enable multi layer? |
| 311 | Settable<bool> video_three_layers; |
| 312 | // Experimental: Enable camera list? |
| 313 | Settable<bool> video_enable_camera_list; |
| 314 | // Experimental: Enable one layer screencast? |
| 315 | Settable<bool> video_one_layer_screencast; |
| 316 | // Experimental: Enable WebRtc higher bitrate? |
| 317 | Settable<bool> video_high_bitrate; |
| 318 | // Experimental: Add watermark to the rendered video image. |
| 319 | Settable<bool> video_watermark; |
| 320 | // Experimental: Enable WebRTC layered screencast. |
| 321 | Settable<bool> video_temporal_layer_screencast; |
| 322 | // Enable WebRTC leaky bucket when sending media packets. |
| 323 | Settable<bool> video_leaky_bucket; |
| 324 | // Use conference mode? |
| 325 | Settable<bool> conference_mode; |
| 326 | // Threshhold for process cpu adaptation. (Process limit) |
| 327 | SettablePercent process_adaptation_threshhold; |
| 328 | // Low threshhold for cpu adaptation. (Adapt up) |
| 329 | SettablePercent system_low_adaptation_threshhold; |
| 330 | // High threshhold for cpu adaptation. (Adapt down) |
| 331 | SettablePercent system_high_adaptation_threshhold; |
| 332 | // Specify buffered mode latency in milliseconds. |
| 333 | Settable<int> buffered_mode_latency; |
| 334 | }; |
| 335 | |
| 336 | // A class for playing out soundclips. |
| 337 | class SoundclipMedia { |
| 338 | public: |
| 339 | enum SoundclipFlags { |
| 340 | SF_LOOP = 1, |
| 341 | }; |
| 342 | |
| 343 | virtual ~SoundclipMedia() {} |
| 344 | |
| 345 | // Plays a sound out to the speakers with the given audio stream. The stream |
| 346 | // must be 16-bit little-endian 16 kHz PCM. If a stream is already playing |
| 347 | // on this SoundclipMedia, it is stopped. If clip is NULL, nothing is played. |
| 348 | // Returns whether it was successful. |
| 349 | virtual bool PlaySound(const char *clip, int len, int flags) = 0; |
| 350 | }; |
| 351 | |
| 352 | struct RtpHeaderExtension { |
| 353 | RtpHeaderExtension() : id(0) {} |
| 354 | RtpHeaderExtension(const std::string& u, int i) : uri(u), id(i) {} |
| 355 | std::string uri; |
| 356 | int id; |
| 357 | // TODO(juberti): SendRecv direction; |
| 358 | |
| 359 | bool operator==(const RtpHeaderExtension& ext) const { |
| 360 | // id is a reserved word in objective-c. Therefore the id attribute has to |
| 361 | // be a fully qualified name in order to compile on IOS. |
| 362 | return this->id == ext.id && |
| 363 | uri == ext.uri; |
| 364 | } |
| 365 | }; |
| 366 | |
| 367 | // Returns the named header extension if found among all extensions, NULL |
| 368 | // otherwise. |
| 369 | inline const RtpHeaderExtension* FindHeaderExtension( |
| 370 | const std::vector<RtpHeaderExtension>& extensions, |
| 371 | const std::string& name) { |
| 372 | for (std::vector<RtpHeaderExtension>::const_iterator it = extensions.begin(); |
| 373 | it != extensions.end(); ++it) { |
| 374 | if (it->uri == name) |
| 375 | return &(*it); |
| 376 | } |
| 377 | return NULL; |
| 378 | } |
| 379 | |
| 380 | enum MediaChannelOptions { |
| 381 | // Tune the stream for conference mode. |
| 382 | OPT_CONFERENCE = 0x0001 |
| 383 | }; |
| 384 | |
| 385 | enum VoiceMediaChannelOptions { |
| 386 | // Tune the audio stream for vcs with different target levels. |
| 387 | OPT_AGC_MINUS_10DB = 0x80000000 |
| 388 | }; |
| 389 | |
| 390 | // DTMF flags to control if a DTMF tone should be played and/or sent. |
| 391 | enum DtmfFlags { |
| 392 | DF_PLAY = 0x01, |
| 393 | DF_SEND = 0x02, |
| 394 | }; |
| 395 | |
| 396 | // Special purpose DTMF event code used by the VoiceMediaChannel::InsertDtmf. |
| 397 | const int kDtmfDelay = -1; // Insert a delay to the end of the DTMF queue. |
| 398 | const int kDtmfReset = -2; // Reset the DTMF queue. |
| 399 | // The delay in ms when the InsertDtmf is called with kDtmfDelay. |
| 400 | const int kDtmfDelayInMs = 2000; |
| 401 | |
| 402 | class MediaChannel : public sigslot::has_slots<> { |
| 403 | public: |
| 404 | class NetworkInterface { |
| 405 | public: |
| 406 | enum SocketType { ST_RTP, ST_RTCP }; |
| 407 | virtual bool SendPacket(talk_base::Buffer* packet) = 0; |
| 408 | virtual bool SendRtcp(talk_base::Buffer* packet) = 0; |
| 409 | virtual int SetOption(SocketType type, talk_base::Socket::Option opt, |
| 410 | int option) = 0; |
| 411 | virtual ~NetworkInterface() {} |
| 412 | }; |
| 413 | |
| 414 | MediaChannel() : network_interface_(NULL) {} |
| 415 | virtual ~MediaChannel() {} |
| 416 | |
| 417 | // Gets/sets the abstract inteface class for sending RTP/RTCP data. |
| 418 | NetworkInterface *network_interface() { return network_interface_; } |
| 419 | virtual void SetInterface(NetworkInterface *iface) { |
| 420 | network_interface_ = iface; |
| 421 | } |
| 422 | |
| 423 | // Called when a RTP packet is received. |
| 424 | virtual void OnPacketReceived(talk_base::Buffer* packet) = 0; |
| 425 | // Called when a RTCP packet is received. |
| 426 | virtual void OnRtcpReceived(talk_base::Buffer* packet) = 0; |
| 427 | // Called when the socket's ability to send has changed. |
| 428 | virtual void OnReadyToSend(bool ready) = 0; |
| 429 | // Creates a new outgoing media stream with SSRCs and CNAME as described |
| 430 | // by sp. |
| 431 | virtual bool AddSendStream(const StreamParams& sp) = 0; |
| 432 | // Removes an outgoing media stream. |
| 433 | // ssrc must be the first SSRC of the media stream if the stream uses |
| 434 | // multiple SSRCs. |
| 435 | virtual bool RemoveSendStream(uint32 ssrc) = 0; |
| 436 | // Creates a new incoming media stream with SSRCs and CNAME as described |
| 437 | // by sp. |
| 438 | virtual bool AddRecvStream(const StreamParams& sp) = 0; |
| 439 | // Removes an incoming media stream. |
| 440 | // ssrc must be the first SSRC of the media stream if the stream uses |
| 441 | // multiple SSRCs. |
| 442 | virtual bool RemoveRecvStream(uint32 ssrc) = 0; |
| 443 | |
| 444 | // Mutes the channel. |
| 445 | virtual bool MuteStream(uint32 ssrc, bool on) = 0; |
| 446 | |
| 447 | // Sets the RTP extension headers and IDs to use when sending RTP. |
| 448 | virtual bool SetRecvRtpHeaderExtensions( |
| 449 | const std::vector<RtpHeaderExtension>& extensions) = 0; |
| 450 | virtual bool SetSendRtpHeaderExtensions( |
| 451 | const std::vector<RtpHeaderExtension>& extensions) = 0; |
| 452 | // Sets the rate control to use when sending data. |
| 453 | virtual bool SetSendBandwidth(bool autobw, int bps) = 0; |
| 454 | |
| 455 | protected: |
| 456 | NetworkInterface *network_interface_; |
| 457 | }; |
| 458 | |
| 459 | enum SendFlags { |
| 460 | SEND_NOTHING, |
| 461 | SEND_RINGBACKTONE, |
| 462 | SEND_MICROPHONE |
| 463 | }; |
| 464 | |
| 465 | struct VoiceSenderInfo { |
| 466 | VoiceSenderInfo() |
| 467 | : ssrc(0), |
| 468 | bytes_sent(0), |
| 469 | packets_sent(0), |
| 470 | packets_lost(0), |
| 471 | fraction_lost(0.0), |
| 472 | ext_seqnum(0), |
| 473 | rtt_ms(0), |
| 474 | jitter_ms(0), |
| 475 | audio_level(0), |
| 476 | aec_quality_min(0.0), |
| 477 | echo_delay_median_ms(0), |
| 478 | echo_delay_std_ms(0), |
| 479 | echo_return_loss(0), |
| 480 | echo_return_loss_enhancement(0) { |
| 481 | } |
| 482 | |
| 483 | uint32 ssrc; |
| 484 | std::string codec_name; |
| 485 | int64 bytes_sent; |
| 486 | int packets_sent; |
| 487 | int packets_lost; |
| 488 | float fraction_lost; |
| 489 | int ext_seqnum; |
| 490 | int rtt_ms; |
| 491 | int jitter_ms; |
| 492 | int audio_level; |
| 493 | float aec_quality_min; |
| 494 | int echo_delay_median_ms; |
| 495 | int echo_delay_std_ms; |
| 496 | int echo_return_loss; |
| 497 | int echo_return_loss_enhancement; |
| 498 | }; |
| 499 | |
| 500 | struct VoiceReceiverInfo { |
| 501 | VoiceReceiverInfo() |
| 502 | : ssrc(0), |
| 503 | bytes_rcvd(0), |
| 504 | packets_rcvd(0), |
| 505 | packets_lost(0), |
| 506 | fraction_lost(0.0), |
| 507 | ext_seqnum(0), |
| 508 | jitter_ms(0), |
| 509 | jitter_buffer_ms(0), |
| 510 | jitter_buffer_preferred_ms(0), |
| 511 | delay_estimate_ms(0), |
| 512 | audio_level(0), |
| 513 | expand_rate(0) { |
| 514 | } |
| 515 | |
| 516 | uint32 ssrc; |
| 517 | int64 bytes_rcvd; |
| 518 | int packets_rcvd; |
| 519 | int packets_lost; |
| 520 | float fraction_lost; |
| 521 | int ext_seqnum; |
| 522 | int jitter_ms; |
| 523 | int jitter_buffer_ms; |
| 524 | int jitter_buffer_preferred_ms; |
| 525 | int delay_estimate_ms; |
| 526 | int audio_level; |
| 527 | // fraction of synthesized speech inserted through pre-emptive expansion |
| 528 | float expand_rate; |
| 529 | }; |
| 530 | |
| 531 | struct VideoSenderInfo { |
| 532 | VideoSenderInfo() |
| 533 | : bytes_sent(0), |
| 534 | packets_sent(0), |
| 535 | packets_cached(0), |
| 536 | packets_lost(0), |
| 537 | fraction_lost(0.0), |
| 538 | firs_rcvd(0), |
| 539 | nacks_rcvd(0), |
| 540 | rtt_ms(0), |
| 541 | frame_width(0), |
| 542 | frame_height(0), |
| 543 | framerate_input(0), |
| 544 | framerate_sent(0), |
| 545 | nominal_bitrate(0), |
| 546 | preferred_bitrate(0), |
| 547 | adapt_reason(0) { |
| 548 | } |
| 549 | |
| 550 | std::vector<uint32> ssrcs; |
| 551 | std::vector<SsrcGroup> ssrc_groups; |
| 552 | std::string codec_name; |
| 553 | int64 bytes_sent; |
| 554 | int packets_sent; |
| 555 | int packets_cached; |
| 556 | int packets_lost; |
| 557 | float fraction_lost; |
| 558 | int firs_rcvd; |
| 559 | int nacks_rcvd; |
| 560 | int rtt_ms; |
| 561 | int frame_width; |
| 562 | int frame_height; |
| 563 | int framerate_input; |
| 564 | int framerate_sent; |
| 565 | int nominal_bitrate; |
| 566 | int preferred_bitrate; |
| 567 | int adapt_reason; |
| 568 | }; |
| 569 | |
| 570 | struct VideoReceiverInfo { |
| 571 | VideoReceiverInfo() |
| 572 | : bytes_rcvd(0), |
| 573 | packets_rcvd(0), |
| 574 | packets_lost(0), |
| 575 | packets_concealed(0), |
| 576 | fraction_lost(0.0), |
| 577 | firs_sent(0), |
| 578 | nacks_sent(0), |
| 579 | frame_width(0), |
| 580 | frame_height(0), |
| 581 | framerate_rcvd(0), |
| 582 | framerate_decoded(0), |
| 583 | framerate_output(0), |
| 584 | framerate_render_input(0), |
| 585 | framerate_render_output(0) { |
| 586 | } |
| 587 | |
| 588 | std::vector<uint32> ssrcs; |
| 589 | std::vector<SsrcGroup> ssrc_groups; |
| 590 | int64 bytes_rcvd; |
| 591 | // vector<int> layer_bytes_rcvd; |
| 592 | int packets_rcvd; |
| 593 | int packets_lost; |
| 594 | int packets_concealed; |
| 595 | float fraction_lost; |
| 596 | int firs_sent; |
| 597 | int nacks_sent; |
| 598 | int frame_width; |
| 599 | int frame_height; |
| 600 | int framerate_rcvd; |
| 601 | int framerate_decoded; |
| 602 | int framerate_output; |
| 603 | // Framerate as sent to the renderer. |
| 604 | int framerate_render_input; |
| 605 | // Framerate that the renderer reports. |
| 606 | int framerate_render_output; |
| 607 | }; |
| 608 | |
| 609 | struct DataSenderInfo { |
| 610 | DataSenderInfo() |
| 611 | : ssrc(0), |
| 612 | bytes_sent(0), |
| 613 | packets_sent(0) { |
| 614 | } |
| 615 | |
| 616 | uint32 ssrc; |
| 617 | std::string codec_name; |
| 618 | int64 bytes_sent; |
| 619 | int packets_sent; |
| 620 | }; |
| 621 | |
| 622 | struct DataReceiverInfo { |
| 623 | DataReceiverInfo() |
| 624 | : ssrc(0), |
| 625 | bytes_rcvd(0), |
| 626 | packets_rcvd(0) { |
| 627 | } |
| 628 | |
| 629 | uint32 ssrc; |
| 630 | int64 bytes_rcvd; |
| 631 | int packets_rcvd; |
| 632 | }; |
| 633 | |
| 634 | struct BandwidthEstimationInfo { |
| 635 | BandwidthEstimationInfo() |
| 636 | : available_send_bandwidth(0), |
| 637 | available_recv_bandwidth(0), |
| 638 | target_enc_bitrate(0), |
| 639 | actual_enc_bitrate(0), |
| 640 | retransmit_bitrate(0), |
| 641 | transmit_bitrate(0), |
| 642 | bucket_delay(0) { |
| 643 | } |
| 644 | |
| 645 | int available_send_bandwidth; |
| 646 | int available_recv_bandwidth; |
| 647 | int target_enc_bitrate; |
| 648 | int actual_enc_bitrate; |
| 649 | int retransmit_bitrate; |
| 650 | int transmit_bitrate; |
| 651 | int bucket_delay; |
| 652 | }; |
| 653 | |
| 654 | struct VoiceMediaInfo { |
| 655 | void Clear() { |
| 656 | senders.clear(); |
| 657 | receivers.clear(); |
| 658 | } |
| 659 | std::vector<VoiceSenderInfo> senders; |
| 660 | std::vector<VoiceReceiverInfo> receivers; |
| 661 | }; |
| 662 | |
| 663 | struct VideoMediaInfo { |
| 664 | void Clear() { |
| 665 | senders.clear(); |
| 666 | receivers.clear(); |
| 667 | bw_estimations.clear(); |
| 668 | } |
| 669 | std::vector<VideoSenderInfo> senders; |
| 670 | std::vector<VideoReceiverInfo> receivers; |
| 671 | std::vector<BandwidthEstimationInfo> bw_estimations; |
| 672 | }; |
| 673 | |
| 674 | struct DataMediaInfo { |
| 675 | void Clear() { |
| 676 | senders.clear(); |
| 677 | receivers.clear(); |
| 678 | } |
| 679 | std::vector<DataSenderInfo> senders; |
| 680 | std::vector<DataReceiverInfo> receivers; |
| 681 | }; |
| 682 | |
| 683 | class VoiceMediaChannel : public MediaChannel { |
| 684 | public: |
| 685 | enum Error { |
| 686 | ERROR_NONE = 0, // No error. |
| 687 | ERROR_OTHER, // Other errors. |
| 688 | ERROR_REC_DEVICE_OPEN_FAILED = 100, // Could not open mic. |
| 689 | ERROR_REC_DEVICE_MUTED, // Mic was muted by OS. |
| 690 | ERROR_REC_DEVICE_SILENT, // No background noise picked up. |
| 691 | ERROR_REC_DEVICE_SATURATION, // Mic input is clipping. |
| 692 | ERROR_REC_DEVICE_REMOVED, // Mic was removed while active. |
| 693 | ERROR_REC_RUNTIME_ERROR, // Processing is encountering errors. |
| 694 | ERROR_REC_SRTP_ERROR, // Generic SRTP failure. |
| 695 | ERROR_REC_SRTP_AUTH_FAILED, // Failed to authenticate packets. |
| 696 | ERROR_REC_TYPING_NOISE_DETECTED, // Typing noise is detected. |
| 697 | ERROR_PLAY_DEVICE_OPEN_FAILED = 200, // Could not open playout. |
| 698 | ERROR_PLAY_DEVICE_MUTED, // Playout muted by OS. |
| 699 | ERROR_PLAY_DEVICE_REMOVED, // Playout removed while active. |
| 700 | ERROR_PLAY_RUNTIME_ERROR, // Errors in voice processing. |
| 701 | ERROR_PLAY_SRTP_ERROR, // Generic SRTP failure. |
| 702 | ERROR_PLAY_SRTP_AUTH_FAILED, // Failed to authenticate packets. |
| 703 | ERROR_PLAY_SRTP_REPLAY, // Packet replay detected. |
| 704 | }; |
| 705 | |
| 706 | VoiceMediaChannel() {} |
| 707 | virtual ~VoiceMediaChannel() {} |
| 708 | // Sets the codecs/payload types to be used for incoming media. |
| 709 | virtual bool SetRecvCodecs(const std::vector<AudioCodec>& codecs) = 0; |
| 710 | // Sets the codecs/payload types to be used for outgoing media. |
| 711 | virtual bool SetSendCodecs(const std::vector<AudioCodec>& codecs) = 0; |
| 712 | // Starts or stops playout of received audio. |
| 713 | virtual bool SetPlayout(bool playout) = 0; |
| 714 | // Starts or stops sending (and potentially capture) of local audio. |
| 715 | virtual bool SetSend(SendFlags flag) = 0; |
| 716 | // Sets the renderer object to be used for the specified audio stream. |
| 717 | virtual bool SetRenderer(uint32 ssrc, AudioRenderer* renderer) = 0; |
| 718 | // Gets current energy levels for all incoming streams. |
| 719 | virtual bool GetActiveStreams(AudioInfo::StreamList* actives) = 0; |
| 720 | // Get the current energy level of the stream sent to the speaker. |
| 721 | virtual int GetOutputLevel() = 0; |
| 722 | // Get the time in milliseconds since last recorded keystroke, or negative. |
| 723 | virtual int GetTimeSinceLastTyping() = 0; |
| 724 | // Temporarily exposed field for tuning typing detect options. |
| 725 | virtual void SetTypingDetectionParameters(int time_window, |
| 726 | int cost_per_typing, int reporting_threshold, int penalty_decay, |
| 727 | int type_event_delay) = 0; |
| 728 | // Set left and right scale for speaker output volume of the specified ssrc. |
| 729 | virtual bool SetOutputScaling(uint32 ssrc, double left, double right) = 0; |
| 730 | // Get left and right scale for speaker output volume of the specified ssrc. |
| 731 | virtual bool GetOutputScaling(uint32 ssrc, double* left, double* right) = 0; |
| 732 | // Specifies a ringback tone to be played during call setup. |
| 733 | virtual bool SetRingbackTone(const char *buf, int len) = 0; |
| 734 | // Plays or stops the aforementioned ringback tone |
| 735 | virtual bool PlayRingbackTone(uint32 ssrc, bool play, bool loop) = 0; |
| 736 | // Returns if the telephone-event has been negotiated. |
| 737 | virtual bool CanInsertDtmf() { return false; } |
| 738 | // Send and/or play a DTMF |event| according to the |flags|. |
| 739 | // The DTMF out-of-band signal will be used on sending. |
| 740 | // The |ssrc| should be either 0 or a valid send stream ssrc. |
| 741 | // The valid value for the |event| are -2 to 15. |
| 742 | // kDtmfReset(-2) is used to reset the DTMF. |
| 743 | // kDtmfDelay(-1) is used to insert a delay to the end of the DTMF queue. |
| 744 | // 0 to 15 which corresponding to DTMF event 0-9, *, #, A-D. |
| 745 | virtual bool InsertDtmf(uint32 ssrc, int event, int duration, int flags) = 0; |
| 746 | // Gets quality stats for the channel. |
| 747 | virtual bool GetStats(VoiceMediaInfo* info) = 0; |
| 748 | // Gets last reported error for this media channel. |
| 749 | virtual void GetLastMediaError(uint32* ssrc, |
| 750 | VoiceMediaChannel::Error* error) { |
| 751 | ASSERT(error != NULL); |
| 752 | *error = ERROR_NONE; |
| 753 | } |
| 754 | // Sets the media options to use. |
| 755 | virtual bool SetOptions(const AudioOptions& options) = 0; |
| 756 | virtual bool GetOptions(AudioOptions* options) const = 0; |
| 757 | |
| 758 | // Signal errors from MediaChannel. Arguments are: |
| 759 | // ssrc(uint32), and error(VoiceMediaChannel::Error). |
| 760 | sigslot::signal2<uint32, VoiceMediaChannel::Error> SignalMediaError; |
| 761 | }; |
| 762 | |
| 763 | class VideoMediaChannel : public MediaChannel { |
| 764 | public: |
| 765 | enum Error { |
| 766 | ERROR_NONE = 0, // No error. |
| 767 | ERROR_OTHER, // Other errors. |
| 768 | ERROR_REC_DEVICE_OPEN_FAILED = 100, // Could not open camera. |
| 769 | ERROR_REC_DEVICE_NO_DEVICE, // No camera. |
| 770 | ERROR_REC_DEVICE_IN_USE, // Device is in already use. |
| 771 | ERROR_REC_DEVICE_REMOVED, // Device is removed. |
| 772 | ERROR_REC_SRTP_ERROR, // Generic sender SRTP failure. |
| 773 | ERROR_REC_SRTP_AUTH_FAILED, // Failed to authenticate packets. |
| 774 | ERROR_REC_CPU_MAX_CANT_DOWNGRADE, // Can't downgrade capture anymore. |
| 775 | ERROR_PLAY_SRTP_ERROR = 200, // Generic receiver SRTP failure. |
| 776 | ERROR_PLAY_SRTP_AUTH_FAILED, // Failed to authenticate packets. |
| 777 | ERROR_PLAY_SRTP_REPLAY, // Packet replay detected. |
| 778 | }; |
| 779 | |
| 780 | VideoMediaChannel() : renderer_(NULL) {} |
| 781 | virtual ~VideoMediaChannel() {} |
| 782 | // Sets the codecs/payload types to be used for incoming media. |
| 783 | virtual bool SetRecvCodecs(const std::vector<VideoCodec>& codecs) = 0; |
| 784 | // Sets the codecs/payload types to be used for outgoing media. |
| 785 | virtual bool SetSendCodecs(const std::vector<VideoCodec>& codecs) = 0; |
| 786 | // Gets the currently set codecs/payload types to be used for outgoing media. |
| 787 | virtual bool GetSendCodec(VideoCodec* send_codec) = 0; |
| 788 | // Sets the format of a specified outgoing stream. |
| 789 | virtual bool SetSendStreamFormat(uint32 ssrc, const VideoFormat& format) = 0; |
| 790 | // Starts or stops playout of received video. |
| 791 | virtual bool SetRender(bool render) = 0; |
| 792 | // Starts or stops transmission (and potentially capture) of local video. |
| 793 | virtual bool SetSend(bool send) = 0; |
| 794 | // Sets the renderer object to be used for the specified stream. |
| 795 | // If SSRC is 0, the renderer is used for the 'default' stream. |
| 796 | virtual bool SetRenderer(uint32 ssrc, VideoRenderer* renderer) = 0; |
| 797 | // If |ssrc| is 0, replace the default capturer (engine capturer) with |
| 798 | // |capturer|. If |ssrc| is non zero create a new stream with |ssrc| as SSRC. |
| 799 | virtual bool SetCapturer(uint32 ssrc, VideoCapturer* capturer) = 0; |
| 800 | // Gets quality stats for the channel. |
| 801 | virtual bool GetStats(VideoMediaInfo* info) = 0; |
| 802 | |
| 803 | // Send an intra frame to the receivers. |
| 804 | virtual bool SendIntraFrame() = 0; |
| 805 | // Reuqest each of the remote senders to send an intra frame. |
| 806 | virtual bool RequestIntraFrame() = 0; |
| 807 | // Sets the media options to use. |
| 808 | virtual bool SetOptions(const VideoOptions& options) = 0; |
| 809 | virtual bool GetOptions(VideoOptions* options) const = 0; |
| 810 | virtual void UpdateAspectRatio(int ratio_w, int ratio_h) = 0; |
| 811 | |
| 812 | // Signal errors from MediaChannel. Arguments are: |
| 813 | // ssrc(uint32), and error(VideoMediaChannel::Error). |
| 814 | sigslot::signal2<uint32, Error> SignalMediaError; |
| 815 | |
| 816 | protected: |
| 817 | VideoRenderer *renderer_; |
| 818 | }; |
| 819 | |
| 820 | enum DataMessageType { |
| 821 | // TODO(pthatcher): Make this enum match the SCTP PPIDs that WebRTC uses? |
| 822 | DMT_CONTROL = 0, |
| 823 | DMT_BINARY = 1, |
| 824 | DMT_TEXT = 2, |
| 825 | }; |
| 826 | |
| 827 | // Info about data received in DataMediaChannel. For use in |
| 828 | // DataMediaChannel::SignalDataReceived and in all of the signals that |
| 829 | // signal fires, on up the chain. |
| 830 | struct ReceiveDataParams { |
| 831 | // The in-packet stream indentifier. |
| 832 | // For SCTP, this is really SID, not SSRC. |
| 833 | uint32 ssrc; |
| 834 | // The type of message (binary, text, or control). |
| 835 | DataMessageType type; |
| 836 | // A per-stream value incremented per packet in the stream. |
| 837 | int seq_num; |
| 838 | // A per-stream value monotonically increasing with time. |
| 839 | int timestamp; |
| 840 | |
| 841 | ReceiveDataParams() : |
| 842 | ssrc(0), |
| 843 | type(DMT_TEXT), |
| 844 | seq_num(0), |
| 845 | timestamp(0) { |
| 846 | } |
| 847 | }; |
| 848 | |
| 849 | struct SendDataParams { |
| 850 | // The in-packet stream indentifier. |
| 851 | // For SCTP, this is really SID, not SSRC. |
| 852 | uint32 ssrc; |
| 853 | // The type of message (binary, text, or control). |
| 854 | DataMessageType type; |
| 855 | |
| 856 | // For SCTP, whether to send messages flagged as ordered or not. |
| 857 | // If false, messages can be received out of order. |
| 858 | bool ordered; |
| 859 | // For SCTP, whether the messages are sent reliably or not. |
| 860 | // If false, messages may be lost. |
| 861 | bool reliable; |
| 862 | // For SCTP, if reliable == false, provide partial reliability by |
| 863 | // resending up to this many times. Either count or millis |
| 864 | // is supported, not both at the same time. |
| 865 | int max_rtx_count; |
| 866 | // For SCTP, if reliable == false, provide partial reliability by |
| 867 | // resending for up to this many milliseconds. Either count or millis |
| 868 | // is supported, not both at the same time. |
| 869 | int max_rtx_ms; |
| 870 | |
| 871 | SendDataParams() : |
| 872 | ssrc(0), |
| 873 | type(DMT_TEXT), |
| 874 | // TODO(pthatcher): Make these true by default? |
| 875 | ordered(false), |
| 876 | reliable(false), |
| 877 | max_rtx_count(0), |
| 878 | max_rtx_ms(0) { |
| 879 | } |
| 880 | }; |
| 881 | |
| 882 | enum SendDataResult { SDR_SUCCESS, SDR_ERROR, SDR_BLOCK }; |
| 883 | |
| 884 | class DataMediaChannel : public MediaChannel { |
| 885 | public: |
| 886 | enum Error { |
| 887 | ERROR_NONE = 0, // No error. |
| 888 | ERROR_OTHER, // Other errors. |
| 889 | ERROR_SEND_SRTP_ERROR = 200, // Generic SRTP failure. |
| 890 | ERROR_SEND_SRTP_AUTH_FAILED, // Failed to authenticate packets. |
| 891 | ERROR_RECV_SRTP_ERROR, // Generic SRTP failure. |
| 892 | ERROR_RECV_SRTP_AUTH_FAILED, // Failed to authenticate packets. |
| 893 | ERROR_RECV_SRTP_REPLAY, // Packet replay detected. |
| 894 | }; |
| 895 | |
| 896 | virtual ~DataMediaChannel() {} |
| 897 | |
| 898 | virtual bool SetSendBandwidth(bool autobw, int bps) = 0; |
| 899 | virtual bool SetSendCodecs(const std::vector<DataCodec>& codecs) = 0; |
| 900 | virtual bool SetRecvCodecs(const std::vector<DataCodec>& codecs) = 0; |
| 901 | virtual bool SetRecvRtpHeaderExtensions( |
| 902 | const std::vector<RtpHeaderExtension>& extensions) = 0; |
| 903 | virtual bool SetSendRtpHeaderExtensions( |
| 904 | const std::vector<RtpHeaderExtension>& extensions) = 0; |
| 905 | virtual bool AddSendStream(const StreamParams& sp) = 0; |
| 906 | virtual bool RemoveSendStream(uint32 ssrc) = 0; |
| 907 | virtual bool AddRecvStream(const StreamParams& sp) = 0; |
| 908 | virtual bool RemoveRecvStream(uint32 ssrc) = 0; |
| 909 | virtual bool MuteStream(uint32 ssrc, bool on) { return false; } |
| 910 | // TODO(pthatcher): Implement this. |
| 911 | virtual bool GetStats(DataMediaInfo* info) { return true; } |
| 912 | |
| 913 | virtual bool SetSend(bool send) = 0; |
| 914 | virtual bool SetReceive(bool receive) = 0; |
| 915 | virtual void OnPacketReceived(talk_base::Buffer* packet) = 0; |
| 916 | virtual void OnRtcpReceived(talk_base::Buffer* packet) = 0; |
| 917 | |
| 918 | virtual bool SendData( |
| 919 | const SendDataParams& params, |
| 920 | const talk_base::Buffer& payload, |
| 921 | SendDataResult* result = NULL) = 0; |
| 922 | // Signals when data is received (params, data, len) |
| 923 | sigslot::signal3<const ReceiveDataParams&, |
| 924 | const char*, |
| 925 | size_t> SignalDataReceived; |
| 926 | // Signal errors from MediaChannel. Arguments are: |
| 927 | // ssrc(uint32), and error(DataMediaChannel::Error). |
| 928 | sigslot::signal2<uint32, DataMediaChannel::Error> SignalMediaError; |
| 929 | }; |
| 930 | |
| 931 | } // namespace cricket |
| 932 | |
| 933 | #endif // TALK_MEDIA_BASE_MEDIACHANNEL_H_ |