update WebRTC to #23794, support mix unit less then 10ms

README.md

@@ -131,14 +131,13 @@ For more detailed info, please refer to [the source code](https://github.com/Pia
 ## Dependencies
 
 + FFmpeg: 3.4.2
-+ WebRTC: #23505
++ WebRTC: #23794
 
 ## Development
 
 + Before run Android demo, push mp3 to sdcard: `adb push mp3 /sdcard/`
 + Generate sources: `./run_djinni.sh`
 + Extract libs: `./extract_libs.sh`
-+ Extract WebRTC ADM libs: `./extract_webrtc_objc_libs.sh`
 
 ## Caveat
 

extract_libs.sh

@@ -1,6 +1,6 @@
 #!/bin/bash
 
-TAG=libs-20180610
+TAG=libs-20180702
 
 wget https://github.com/Piasy/AudioMixer/releases/download/${TAG}/${TAG}.zip && \
 rm -rf libs && \

ios_project/AudioMixer.xcodeproj/project.pbxproj

@@ -10,10 +10,7 @@
 		8634187320D3E89400C144CB /* AudioMixerTestController.mm in Sources */ = {isa = PBXBuildFile; fileRef = 86D17F9920D3E4CA00DFFEAC /* AudioMixerTestController.mm */; };
 		866A956620D164D600C7694D /* AudioMixer.h in Headers */ = {isa = PBXBuildFile; fileRef = 866A956520D164D600C7694D /* AudioMixer.h */; settings = {ATTRIBUTES = (Public, ); }; };
 		866A956D20D165E600C7694D /* morning-mono-16k.mp3 in Resources */ = {isa = PBXBuildFile; fileRef = 866A956820D165E500C7694D /* morning-mono-16k.mp3 */; };
-		866A956E20D165E600C7694D /* morning.mp3 in Resources */ = {isa = PBXBuildFile; fileRef = 866A956920D165E500C7694D /* morning.mp3 */; };
-		866A956F20D165E600C7694D /* iamyou.mp3 in Resources */ = {isa = PBXBuildFile; fileRef = 866A956A20D165E500C7694D /* iamyou.mp3 */; };
 		866A957020D165E600C7694D /* morning.raw in Resources */ = {isa = PBXBuildFile; fileRef = 866A956B20D165E500C7694D /* morning.raw */; };
-		866A957120D165E600C7694D /* lion.mp3 in Resources */ = {isa = PBXBuildFile; fileRef = 866A956C20D165E500C7694D /* lion.mp3 */; };
 		869959A320CF447A009946FC /* audio_file_decoder.h in Headers */ = {isa = PBXBuildFile; fileRef = 8699595B20CF4479009946FC /* audio_file_decoder.h */; };
 		869959A420CF447A009946FC /* audio_source.h in Headers */ = {isa = PBXBuildFile; fileRef = 8699595C20CF4479009946FC /* audio_source.h */; };
 		869959A520CF447A009946FC /* audio_resampler.h in Headers */ = {isa = PBXBuildFile; fileRef = 8699595D20CF4479009946FC /* audio_resampler.h */; };
@@ -59,6 +56,9 @@
 		869959E720CF7D9F009946FC /* PYAAudioBuffer.mm in Sources */ = {isa = PBXBuildFile; fileRef = 869959E520CF7D9F009946FC /* PYAAudioBuffer.mm */; };
 		869959E920CF85CB009946FC /* audio_mixer_api.cpp in Sources */ = {isa = PBXBuildFile; fileRef = 869959E820CF85CB009946FC /* audio_mixer_api.cpp */; };
 		869959EB20CF85E7009946FC /* PYAAudioMixerApi.mm in Sources */ = {isa = PBXBuildFile; fileRef = 869959EA20CF85E7009946FC /* PYAAudioMixerApi.mm */; };
+		86E1F3CD20E8E9A2001C31C2 /* iamyou-48k.mp3 in Resources */ = {isa = PBXBuildFile; fileRef = 86E1F3CA20E8E9A2001C31C2 /* iamyou-48k.mp3 */; };
+		86E1F3CE20E8E9A2001C31C2 /* lion-48k.mp3 in Resources */ = {isa = PBXBuildFile; fileRef = 86E1F3CB20E8E9A2001C31C2 /* lion-48k.mp3 */; };
+		86E1F3CF20E8E9A2001C31C2 /* morning-48k.mp3 in Resources */ = {isa = PBXBuildFile; fileRef = 86E1F3CC20E8E9A2001C31C2 /* morning-48k.mp3 */; };
 		BF_208404374351 /* Assets.xcassets in Resources */ = {isa = PBXBuildFile; fileRef = FR_316246524774 /* Assets.xcassets */; };
 		BF_232288806286 /* AudioMixer.framework in Frameworks */ = {isa = PBXBuildFile; fileRef = FR_262413649868 /* AudioMixer.framework */; };
 		BF_387680161997 /* AudioMixer.framework in CopyFiles */ = {isa = PBXBuildFile; fileRef = FR_262413649868 /* AudioMixer.framework */; settings = {ATTRIBUTES = (CodeSignOnCopy, RemoveHeadersOnCopy, ); }; };
@@ -93,10 +93,7 @@
 /* Begin PBXFileReference section */
 		866A956520D164D600C7694D /* AudioMixer.h */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.c.h; path = AudioMixer.h; sourceTree = "<group>"; };
 		866A956820D165E500C7694D /* morning-mono-16k.mp3 */ = {isa = PBXFileReference; lastKnownFileType = audio.mp3; path = "morning-mono-16k.mp3"; sourceTree = "<group>"; };
-		866A956920D165E500C7694D /* morning.mp3 */ = {isa = PBXFileReference; lastKnownFileType = audio.mp3; path = morning.mp3; sourceTree = "<group>"; };
-		866A956A20D165E500C7694D /* iamyou.mp3 */ = {isa = PBXFileReference; lastKnownFileType = audio.mp3; path = iamyou.mp3; sourceTree = "<group>"; };
 		866A956B20D165E500C7694D /* morning.raw */ = {isa = PBXFileReference; lastKnownFileType = file; path = morning.raw; sourceTree = "<group>"; };
-		866A956C20D165E500C7694D /* lion.mp3 */ = {isa = PBXFileReference; lastKnownFileType = audio.mp3; path = lion.mp3; sourceTree = "<group>"; };
 		8699595B20CF4479009946FC /* audio_file_decoder.h */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.c.h; path = audio_file_decoder.h; sourceTree = "<group>"; };
 		8699595C20CF4479009946FC /* audio_source.h */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.c.h; path = audio_source.h; sourceTree = "<group>"; };
 		8699595D20CF4479009946FC /* audio_resampler.h */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.c.h; path = audio_resampler.h; sourceTree = "<group>"; };
@@ -144,6 +141,9 @@
 		869959EA20CF85E7009946FC /* PYAAudioMixerApi.mm */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.cpp.objcpp; path = PYAAudioMixerApi.mm; sourceTree = "<group>"; };
 		86D17F9820D3E4CA00DFFEAC /* AudioMixerTestController.h */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.c.h; path = AudioMixerTestController.h; sourceTree = "<group>"; };
 		86D17F9920D3E4CA00DFFEAC /* AudioMixerTestController.mm */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.cpp.objcpp; path = AudioMixerTestController.mm; sourceTree = "<group>"; };
+		86E1F3CA20E8E9A2001C31C2 /* iamyou-48k.mp3 */ = {isa = PBXFileReference; lastKnownFileType = audio.mp3; path = "iamyou-48k.mp3"; sourceTree = "<group>"; };
+		86E1F3CB20E8E9A2001C31C2 /* lion-48k.mp3 */ = {isa = PBXFileReference; lastKnownFileType = audio.mp3; path = "lion-48k.mp3"; sourceTree = "<group>"; };
+		86E1F3CC20E8E9A2001C31C2 /* morning-48k.mp3 */ = {isa = PBXFileReference; lastKnownFileType = audio.mp3; path = "morning-48k.mp3"; sourceTree = "<group>"; };
 		FR_166489251693 /* main.m */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.c.objc; path = main.m; sourceTree = "<group>"; };
 		FR_262413649868 /* AudioMixer.framework */ = {isa = PBXFileReference; explicitFileType = wrapper.framework; includeInIndex = 0; path = AudioMixer.framework; sourceTree = BUILT_PRODUCTS_DIR; };
 		FR_316246524774 /* Assets.xcassets */ = {isa = PBXFileReference; lastKnownFileType = folder.assetcatalog; path = Assets.xcassets; sourceTree = "<group>"; };
@@ -170,11 +170,11 @@
 		866A956720D165E500C7694D /* mp3 */ = {
 			isa = PBXGroup;
 			children = (
+				86E1F3CA20E8E9A2001C31C2 /* iamyou-48k.mp3 */,
+				86E1F3CB20E8E9A2001C31C2 /* lion-48k.mp3 */,
+				86E1F3CC20E8E9A2001C31C2 /* morning-48k.mp3 */,
 				866A956820D165E500C7694D /* morning-mono-16k.mp3 */,
-				866A956920D165E500C7694D /* morning.mp3 */,
-				866A956A20D165E500C7694D /* iamyou.mp3 */,
 				866A956B20D165E500C7694D /* morning.raw */,
-				866A956C20D165E500C7694D /* lion.mp3 */,
 			);
 			name = mp3;
 			path = ../mp3;
@@ -441,12 +441,12 @@
 			isa = PBXResourcesBuildPhase;
 			buildActionMask = 2147483647;
 			files = (
-				866A956F20D165E600C7694D /* iamyou.mp3 in Resources */,
+				86E1F3CD20E8E9A2001C31C2 /* iamyou-48k.mp3 in Resources */,
 				BF_208404374351 /* Assets.xcassets in Resources */,
-				866A957120D165E600C7694D /* lion.mp3 in Resources */,
 				866A957020D165E600C7694D /* morning.raw in Resources */,
-				866A956E20D165E600C7694D /* morning.mp3 in Resources */,
+				86E1F3CE20E8E9A2001C31C2 /* lion-48k.mp3 in Resources */,
 				866A956D20D165E600C7694D /* morning-mono-16k.mp3 in Resources */,
+				86E1F3CF20E8E9A2001C31C2 /* morning-48k.mp3 in Resources */,
 			);
 			runOnlyForDeploymentPostprocessing = 0;
 		};
@@ -554,7 +554,7 @@
 				SKIP_INSTALL = YES;
 				SYSTEM_HEADER_SEARCH_PATHS = "$(inherited) $(PROJECT_DIR)/../libs/ffmpeg/include $(PROJECT_DIR)/../libs/webrtc/include";
 				TARGETED_DEVICE_FAMILY = "1,2";
-				USER_HEADER_SEARCH_PATHS = "$(inherited) $(PROJECT_DIR)/../libs/ffmpeg/include $(PROJECT_DIR)/../libs/webrtc/include $(PROJECT_DIR)/../djinni/support-lib $(PROJECT_DIR)/../djinni/support-lib/objc $(PROJECT_DIR)/../generated_src/cpp $(PROJECT_DIR)/../generated_src/objc $(PROJECT_DIR)/../src";
+				USER_HEADER_SEARCH_PATHS = "$(inherited) $(PROJECT_DIR)/../libs/ffmpeg/include $(PROJECT_DIR)/../libs/webrtc/include $(PROJECT_DIR)/../libs/webrtc/include/third_party/abseil-cpp $(PROJECT_DIR)/../djinni/support-lib $(PROJECT_DIR)/../djinni/support-lib/objc $(PROJECT_DIR)/../generated_src/cpp $(PROJECT_DIR)/../generated_src/objc $(PROJECT_DIR)/../src";
 				VERSIONING_SYSTEM = "apple-generic";
 			};
 			name = Release;
@@ -670,7 +670,7 @@
 				SKIP_INSTALL = YES;
 				SYSTEM_HEADER_SEARCH_PATHS = "$(inherited) $(PROJECT_DIR)/../libs/ffmpeg/include $(PROJECT_DIR)/../libs/webrtc/include";
 				TARGETED_DEVICE_FAMILY = "1,2";
-				USER_HEADER_SEARCH_PATHS = "$(inherited) $(PROJECT_DIR)/../libs/ffmpeg/include $(PROJECT_DIR)/../libs/webrtc/include $(PROJECT_DIR)/../djinni/support-lib $(PROJECT_DIR)/../djinni/support-lib/objc $(PROJECT_DIR)/../generated_src/cpp $(PROJECT_DIR)/../generated_src/objc $(PROJECT_DIR)/../src";
+				USER_HEADER_SEARCH_PATHS = "$(inherited) $(PROJECT_DIR)/../libs/ffmpeg/include $(PROJECT_DIR)/../libs/webrtc/include $(PROJECT_DIR)/../libs/webrtc/include/third_party/abseil-cpp $(PROJECT_DIR)/../djinni/support-lib $(PROJECT_DIR)/../djinni/support-lib/objc $(PROJECT_DIR)/../generated_src/cpp $(PROJECT_DIR)/../generated_src/objc $(PROJECT_DIR)/../src";
 				VERSIONING_SYSTEM = "apple-generic";
 			};
 			name = Debug;
@@ -764,6 +764,7 @@
 				SDKROOT = iphoneos;
 				SYSTEM_HEADER_SEARCH_PATHS = "$(PROJECT_DIR)/../libs/ffmpeg/include $(PROJECT_DIR)/../libs/webrtc/include";
 				TARGETED_DEVICE_FAMILY = "1,2";
+				USER_HEADER_SEARCH_PATHS = "$(PROJECT_DIR)/../libs/webrtc/include/third_party/abseil-cpp";
 				VALID_ARCHS = "arm64 armv7 armv7s";
 			};
 			name = Debug;
@@ -807,6 +808,7 @@
 				SDKROOT = iphoneos;
 				SYSTEM_HEADER_SEARCH_PATHS = "$(PROJECT_DIR)/../libs/ffmpeg/include $(PROJECT_DIR)/../libs/webrtc/include";
 				TARGETED_DEVICE_FAMILY = "1,2";
+				USER_HEADER_SEARCH_PATHS = "$(PROJECT_DIR)/../libs/webrtc/include/third_party/abseil-cpp";
 				VALID_ARCHS = "arm64 armv7 armv7s";
 			};
 			name = Release;

ios_project/example/AudioMixerTestController.mm

@@ -236,7 +236,7 @@ - (void)doDecodeAny {
     _remainingData = 0;
 
     _source = new audio_mixer::AudioFileSource(
-        1, [[self pathForFileName:@"morning.mp3"]
+        1, [[self pathForFileName:@"morning-48k.mp3"]
                cStringUsingEncoding:NSUTF8StringEncoding],
         _sampleRate, _channelNum, 10, 1);
     _buffer = malloc(7680);
@@ -285,26 +285,26 @@ - (void)doMix {
             mixerSourceWithType:PYAMixerSourceTypeFile
                            ssrc:1
                          volume:1
-                           path:[self pathForFileName:@"morning.mp3"]
+                           path:[self pathForFileName:@"morning-48k.mp3"]
                      sampleRate:0
                      channelNum:0],
         [PYAMixerSource mixerSourceWithType:PYAMixerSourceTypeFile
                                        ssrc:2
                                      volume:1
-                                       path:[self pathForFileName:@"iamyou.mp3"]
+                                       path:[self pathForFileName:@"iamyou-48k.mp3"]
                                  sampleRate:0
                                  channelNum:0],
         [PYAMixerSource mixerSourceWithType:PYAMixerSourceTypeFile
                                        ssrc:3
                                      volume:1
-                                       path:[self pathForFileName:@"lion.mp3"]
+                                       path:[self pathForFileName:@"lion-48k.mp3"]
                                  sampleRate:0
                                  channelNum:0],
     ];
     PYAMixerConfig* config = [PYAMixerConfig mixerConfigWithSources:mixerSources
                                                    outputSampleRate:_sampleRate
                                                    outputChannelNum:_channelNum
-                                                    frameDurationMs:10];
+                                                    frameDurationMs:3];
     _mixer = [[PYAAudioMixer alloc] initWithConfig:config];
 
     [self doStartTest];
@@ -350,7 +350,7 @@ - (void)doRecordAndMix {
             mixerSourceWithType:PYAMixerSourceTypeFile
                            ssrc:1
                          volume:1
-                           path:[self pathForFileName:@"morning.mp3"]
+                           path:[self pathForFileName:@"morning-48k.mp3"]
                      sampleRate:0
                      channelNum:0],
         [PYAMixerSource mixerSourceWithType:PYAMixerSourceTypeRecord

libs/webrtc/include/api/array_view.h

@@ -11,8 +11,8 @@
 #ifndef API_ARRAY_VIEW_H_
 #define API_ARRAY_VIEW_H_
 
-#include <array>
 #include <algorithm>
+#include <array>
 #include <type_traits>
 
 #include "rtc_base/checks.h"

libs/webrtc/include/api/audio/audio_frame.h

@@ -43,11 +43,7 @@ class AudioFrame {
     kMaxDataSizeBytes = kMaxDataSizeSamples * sizeof(int16_t),
   };
 
-  enum VADActivity {
-    kVadActive = 0,
-    kVadPassive = 1,
-    kVadUnknown = 2
-  };
+  enum VADActivity { kVadActive = 0, kVadPassive = 1, kVadUnknown = 2 };
   enum SpeechType {
     kNormalSpeech = 0,
     kPLC = 1,
@@ -66,9 +62,12 @@ class AudioFrame {
   // ResetWithoutMuting() to skip this wasteful zeroing.
   void ResetWithoutMuting();
 
-  void UpdateFrame(uint32_t timestamp, const int16_t* data,
-                   size_t samples_per_channel, int sample_rate_hz,
-                   SpeechType speech_type, VADActivity vad_activity,
+  void UpdateFrame(uint32_t timestamp,
+                   const int16_t* data,
+                   size_t samples_per_channel,
+                   int sample_rate_hz,
+                   SpeechType speech_type,
+                   VADActivity vad_activity,
                    size_t num_channels = 1);
 
   void CopyFrom(const AudioFrame& src);
@@ -111,7 +110,7 @@ class AudioFrame {
   // Monotonically increasing timestamp intended for profiling of audio frames.
   // Typically used for measuring elapsed time between two different points in
   // the audio path. No lock is used to save resources and we are thread safe
-  // by design. Also, rtc::Optional is not used since it will cause a "complex
+  // by design. Also, absl::optional is not used since it will cause a "complex
   // class/struct needs an explicit out-of-line destructor" build error.
   int64_t profile_timestamp_ms_ = 0;
 

libs/webrtc/include/api/audio/echo_canceller3_config.h

@@ -21,8 +21,8 @@ struct EchoCanceller3Config {
   EchoCanceller3Config(const EchoCanceller3Config& e);
   struct Delay {
     size_t default_delay = 5;
-    size_t down_sampling_factor = 8;
-    size_t num_filters = 5;
+    size_t down_sampling_factor = 4;
+    size_t num_filters = 6;
     size_t api_call_jitter_blocks = 26;
     size_t min_echo_path_delay_blocks = 0;
     size_t delay_headroom_blocks = 2;
@@ -65,7 +65,8 @@ struct EchoCanceller3Config {
     float lf = 1.f;
     float mf = 1.f;
     float hf = 1.f;
-    float default_len = 0.f;
+    float default_len = 0.88f;
+    bool reverb_based_on_render = true;
     bool echo_can_saturate = true;
     bool bounded_erl = false;
   } ep_strength;
@@ -156,6 +157,7 @@ struct EchoCanceller3Config {
 
   struct Suppressor {
     size_t bands_with_reliable_coherence = 5;
+    size_t nearend_average_blocks = 4;
   } suppressor;
 };
 }  // namespace webrtc

libs/webrtc/include/api/audio_codecs/L16/audio_decoder_L16.h

@@ -14,10 +14,10 @@
 #include <memory>
 #include <vector>
 
+#include "absl/types/optional.h"
 #include "api/audio_codecs/audio_codec_pair_id.h"
 #include "api/audio_codecs/audio_decoder.h"
 #include "api/audio_codecs/audio_format.h"
-#include "api/optional.h"
 
 namespace webrtc {
 
@@ -35,11 +35,11 @@ struct AudioDecoderL16 {
     int sample_rate_hz = 8000;
     int num_channels = 1;
   };
-  static rtc::Optional<Config> SdpToConfig(const SdpAudioFormat& audio_format);
+  static absl::optional<Config> SdpToConfig(const SdpAudioFormat& audio_format);
   static void AppendSupportedDecoders(std::vector<AudioCodecSpec>* specs);
   static std::unique_ptr<AudioDecoder> MakeAudioDecoder(
       const Config& config,
-      rtc::Optional<AudioCodecPairId> codec_pair_id = rtc::nullopt);
+      absl::optional<AudioCodecPairId> codec_pair_id = absl::nullopt);
 };
 
 }  // namespace webrtc

libs/webrtc/include/api/audio_codecs/L16/audio_encoder_L16.h

@@ -14,10 +14,10 @@
 #include <memory>
 #include <vector>
 
+#include "absl/types/optional.h"
 #include "api/audio_codecs/audio_codec_pair_id.h"
 #include "api/audio_codecs/audio_encoder.h"
 #include "api/audio_codecs/audio_format.h"
-#include "api/optional.h"
 
 namespace webrtc {
 
@@ -37,13 +37,13 @@ struct AudioEncoderL16 {
     int num_channels = 1;
     int frame_size_ms = 10;
   };
-  static rtc::Optional<Config> SdpToConfig(const SdpAudioFormat& audio_format);
+  static absl::optional<Config> SdpToConfig(const SdpAudioFormat& audio_format);
   static void AppendSupportedEncoders(std::vector<AudioCodecSpec>* specs);
   static AudioCodecInfo QueryAudioEncoder(const Config& config);
   static std::unique_ptr<AudioEncoder> MakeAudioEncoder(
       const Config& config,
       int payload_type,
-      rtc::Optional<AudioCodecPairId> codec_pair_id = rtc::nullopt);
+      absl::optional<AudioCodecPairId> codec_pair_id = absl::nullopt);
 };
 
 }  // namespace webrtc

libs/webrtc/include/api/audio_codecs/audio_decoder.h

@@ -14,8 +14,8 @@
 #include <memory>
 #include <vector>
 
+#include "absl/types/optional.h"
 #include "api/array_view.h"
-#include "api/optional.h"
 #include "rtc_base/buffer.h"
 #include "rtc_base/constructormagic.h"
 #include "typedefs.h"  // NOLINT(build/include)
@@ -53,11 +53,11 @@ class AudioDecoder {
 
     // Decodes this frame of audio and writes the result in |decoded|.
     // |decoded| must be large enough to store as many samples as indicated by a
-    // call to Duration() . On success, returns an rtc::Optional containing the
+    // call to Duration() . On success, returns an absl::optional containing the
     // total number of samples across all channels, as well as whether the
     // decoder produced comfort noise or speech. On failure, returns an empty
-    // rtc::Optional. Decode may be called at most once per frame object.
-    virtual rtc::Optional<DecodeResult> Decode(
+    // absl::optional. Decode may be called at most once per frame object.
+    virtual absl::optional<DecodeResult> Decode(
         rtc::ArrayView<int16_t> decoded) const = 0;
   };
 

libs/webrtc/include/api/audio_codecs/audio_decoder_factory.h

@@ -14,10 +14,10 @@
 #include <memory>
 #include <vector>
 
+#include "absl/types/optional.h"
 #include "api/audio_codecs/audio_codec_pair_id.h"
 #include "api/audio_codecs/audio_decoder.h"
 #include "api/audio_codecs/audio_format.h"
-#include "api/optional.h"
 #include "rtc_base/refcount.h"
 
 namespace webrtc {
@@ -41,7 +41,7 @@ class AudioDecoderFactory : public rtc::RefCountInterface {
   // work.
   virtual std::unique_ptr<AudioDecoder> MakeAudioDecoder(
       const SdpAudioFormat& format,
-      rtc::Optional<AudioCodecPairId> codec_pair_id) = 0;
+      absl::optional<AudioCodecPairId> codec_pair_id) = 0;
 };
 
 }  // namespace webrtc