webrtc opus 设置与编码

webrtc opus 设置与编码

webrtc opus bg57iv3 扩展头 格式 audio encoder

Received session description :{

   "sdp" : "v=0

o=- 7489544636758395528 2 IN IP4 127.0.0.1

s=-

t=0 0

a=group:BUNDLE 0 1

a=msid-semantic: WMS stream_id

m=audio 9 RTP/AVPF 111 103 104 9 102 0 8 106 105 13 110 112 113 126

c=IN IP4 0.0.0.0

a=rtcp:9 IN IP4 0.0.0.0

a=ice-ufrag:xUFn

a=ice-pwd:Zq3MQKxIqnkSsr0ESYN8gEFV

a=ice-options:trickle

a=mid:0

a=extmap:1 urn:ietf:params:rtp-hdrext:ssrc-audio-level

a=extmap:2 

a=extmap:3 urn:ietf:params:rtp-hdrext:sdes:mid

a=extmap:4 urn:ietf:params:rtp-hdrext:sdes:rtp-stream-id

a=extmap:5 urn:ietf:params:rtp-hdrext:sdes:repaired-rtp-stream-id

a=sendrecv

a=msid:stream_id audio_label

a=rtcp-mux

a=rtpmap:111 opus/48000/2

a=rtcp-fb:111 transport-cc

a=fmtp:111 minptime=10;useinbandfec=1

a=rtpmap:103 ISAC/16000

a=rtpmap:104 ISAC/32000

a=rtpmap:9 G722/8000

a=rtpmap:102 ILBC/8000

a=rtpmap:0 PCMU/8000

a=rtpmap:8 PCMA/8000

a=rtpmap:106 CN/32000

a=rtpmap:105 CN/16000

a=rtpmap:13 CN/8000

a=rtpmap:110 telephone-event/48000

a=rtpmap:112 telephone-event/32000

a=rtpmap:113 telephone-event/16000

a=rtpmap:126 telephone-event/8000

a=ssrc:1527975052 cname:WiREXAcYmJ5rkNaW

m=video 9 RTP/AVPF 96 97 98 99 100 101 127 124 125

c=IN IP4 0.0.0.0

a=rtcp:9 IN IP4 0.0.0.0

a=ice-ufrag:xUFn

a=ice-pwd:Zq3MQKxIqnkSsr0ESYN8gEFV

a=ice-options:trickle

a=mid:1

a=extmap:14 urn:ietf:params:rtp-hdrext:toffset

a=extmap:13 

a=extmap:12 urn:3gpp:video-orientation

a=extmap:2 

a=extmap:11 http://www.webrtc/experiments/rtp-hdrext/playout-delay

a=extmap:6 

a=extmap:7 

a=extmap:8 

a=extmap:9 

a=extmap:3 urn:ietf:params:rtp-hdrext:sdes:mid

a=extmap:4 urn:ietf:params:rtp-hdrext:sdes:rtp-stream-id

a=extmap:5 urn:ietf:params:rtp-hdrext:sdes:repaired-rtp-stream-id

a=recvonly

a=rtcp-mux

a=rtcp-rsize

a=rtpmap:96 VP8/90000

a=rtcp-fb:96 goog-remb

a=rtcp-fb:96 transport-cc

a=rtcp-fb:96 ccm fir

a=rtcp-fb:96 nack

a=rtcp-fb:96 nack pli

a=rtpmap:97 rtx/90000

a=fmtp:97 apt=96

a=rtpmap:98 VP9/90000

a=rtcp-fb:98 goog-remb

a=rtcp-fb:98 transport-cc

a=rtcp-fb:98 ccm fir

a=rtcp-fb:98 nack

a=rtcp-fb:98 nack pli

a=fmtp:98 profile-id=0

a=rtpmap:99 rtx/90000

a=fmtp:99 apt=98

a=rtpmap:100 VP9/90000

a=rtcp-fb:100 goog-remb

a=rtcp-fb:100 transport-cc

a=rtcp-fb:100 ccm fir

a=rtcp-fb:100 nack

a=rtcp-fb:100 nack pli

a=fmtp:100 profile-id=2

a=rtpmap:101 rtx/90000

a=fmtp:101 apt=100

a=rtpmap:127 red/90000

a=rtpmap:124 rtx/90000

a=fmtp:124 apt=127

a=rtpmap:125 ulpfec/90000

",

   "type" : "answer"

}

Opus 格式设置为：

 WebRtcVoiceMediaChannel::SetSendParameters: {codecs: [AudioCodec[111:opus:48000:0:2], AudioCodec[103:ISAC:16000:0:1], AudioCodec[104:ISAC:32000:0:1], AudioCodec[9:G722:8000:0:1], AudioCodec[102:ILBC:8000:0:1], AudioCodec[0:PCMU:8000:0:1], AudioCodec[8:PCMA:8000:0:1], AudioCodec[106:CN:32000:0:1], AudioCodec[105:CN:16000:0:1], AudioCodec[13:CN:8000:0:1], AudioCodec[110:telephone-event:48000:0:1], AudioCodec[112:telephone-event:32000:0:1], AudioCodec[113:telephone-event:16000:0:1], AudioCodec[126:telephone-event:8000:0:1]], extensions: [{uri: urn:ietf:params:rtp-hdrext:ssrc-audio-level, id: 1}, {uri: , id: 2}, {uri: urn:ietf:params:rtp-hdrext:sdes:mid, id: 3}, {uri: urn:ietf:params:rtp-hdrext:sdes:rtp-stream-id, id: 4}, {uri: urn:ietf:params:rtp-hdrext:sdes:repaired-rtp-stream-id, id: 5}], extmap-allow-mixed: false, max_bandwidth_bps: -1, mid: 0, options: AudioOptions {}}

(audio_send_stream:258): AudioSendStream::ConfigureStream: {rtp: {ssrc: 1708860063, extmap-allow-mixed: false, extensions: [{uri: , id: 2}, {uri: urn:ietf:params:rtp-hdrext:sdes:mid, id: 3}, {uri: urn:ietf:params:rtp-hdrext:sdes:repaired-rtp-stream-id, id: 5}, {uri: urn:ietf:params:rtp-hdrext:sdes:rtp-stream-id, id: 4}, {uri: urn:ietf:params:rtp-hdrext:ssrc-audio-level, id: 1}], c_name: +zQ1fzmm8MR/paky}, rtcp_report_interval_ms: 5000, send_transport: (Transport), media_transport: null, min_bitrate_bps: 32000, max_bitrate_bps: 32000, send_codec_spec: {nack_enabled: false, transport_cc_enabled: true, cng_payload_type: <unset>, payload_type: 111, format: {name: opus, clockrate_hz: 48000, num_channels: 2, parameters: {minptime: 10, useinbandfec: 1}}}}

fc7ff708 11debac8 77258033 5019e027 51327245 0653a6ed 4faf51d4 d254a135 2f6f78f4 d269413d 52fa12e4 cdd88727 05fb302f 9739b367 bda9c26b de98514a 982d93c5 d359ddb0 fbfeaecc 19c9e690 e645e35f fa8fd12e fbc131cf 6913aa0a 4facf341 9f2021a3 8fb620

68222a50 0b25ad5f 9c19543b c2d15e0f a1b3170e 0728546a f0ae3e04 2bf4e1ba 98e98220 3f8fd24e 4ba1b6d3 0260ee36 9d91ffee b4e6b429 4dedee9b 5d7bfbbb fa34df8c 67a205cb 36d7b761 44e778e4 cb4685a5 bd

780be749 feef07fc d847401c f2af5d2b 191eec21 018489d0 eecb817c 2c4c8835 658022dd 51639426 c1a19e0b 7f4dce2f a4ce

一些 OPUS 相关定义：

#define OPUS_SIGNAL_VOICE                    3001 /**< Signal being encoded is voice */

#define OPUS_SIGNAL_MUSIC                    3002 /**< Signal being encoded is music */

#define OPUS_BANDWIDTH_NARROWBAND            1101 /**< 4 kHz bandpass @hideinitializer*/

#define OPUS_BANDWIDTH_MEDIUMBAND            1102 /**< 6 kHz bandpass @hideinitializer*/

#define OPUS_BANDWIDTH_WIDEBAND              1103 /**< 8 kHz bandpass @hideinitializer*/

#define OPUS_BANDWIDTH_SUPERWIDEBAND         1104 /**<12 kHz bandpass @hideinitializer*/

#define OPUS_BANDWIDTH_FULLBAND              1105 /**<20 kHz bandpass @hideinitializer*/

 

#define MODE_SILK_ONLY          1000

#define MODE_HYBRID             1001

#define MODE_CELT_ONLY          1002

 

	Name	Value	Type
◢	this	0x000000000c24df60 {config_={frame_size_ms=20 num_channels=1 application=kVoip (0) ...} payload_type_=...}	webrtc::AudioEncoderOpusImpl *
	◢ webrtc::AudioEncoder	{...}	webrtc::AudioEncoder
	▶ __vfptr	0x0000000142125800 {!void(* webrtc::AudioEncoderOpusImpl::`vftable'[27])()} {0x0000000140a09dd5 {!webrtc::AudioEncoderOpusImpl::`vector deleting destructor'(unsigned int)}, ...}	void * *
	◢ config_	{frame_size_ms=20 num_channels=1 application=kVoip (0) ...}	webrtc::AudioEncoderOpusConfig
	frame_size_ms	20	int
	num_channels	1	unsigned __int64
	application	kVoip (0)	webrtc::AudioEncoderOpusConfig::ApplicationMode
	◢ bitrate_bps	{...}	absl::optional<int>
	▶ absl::optional_internal::optional_data<int,1>	{...}	absl::optional_internal::optional_data<int,1>
	absl::optional_internal::optional_ctor_base<0>	{...}	absl::optional_internal::optional_ctor_base<0>
	absl::optional_internal::optional_assign_base<0>	{...}	absl::optional_internal::optional_assign_base<0>
	fec_enabled	true	bool
	cbr_enabled	false	bool
	max_playback_rate_hz	48000	int
	complexity	9	int
	low_rate_complexity	9	int
	complexity_threshold_bps	12500	int
	complexity_threshold_window_bps	1500	int
	dtx_enabled	false	bool
	▶ supported_frame_lengths_ms	{ size=3 }	std::vector<int,std::allocator<int> >
	uplink_bandwidth_update_interval_ms	200	int
	payload_type	-1	int
	payload_type_	111	const int
	send_side_bwe_with_overhead_	false	const bool
	use_link_capacity_for_adaptation_	false	const bool
	adjust_bandwidth_	false	const bool
	bitrate_changed_	true	bool
	packet_loss_rate_	0.000000000	float
	min_packet_loss_rate_	0.000000000	const float
	▶ new_packet_loss_optimizer_	empty	const std::unique_ptr<webrtc::AudioEncoderOpusImpl::NewPacketLossRateOptimizer,std::default_delete<webrtc::AudioEncoderOpusImpl::NewPacketLossRateOptimizer> >
	▶ input_buffer_	{ size=960 }	std::vector<short,std::allocator<short> >
	▶ inst_	0x000000000c1fa2e0 {encoder=0x000000000c2a8070 {celt_enc_offset=38288 silk_enc_offset=18040 silk_mode=...} ...}	WebRtcOpusEncInst *
	first_timestamp_in_buffer_	3840	unsigned int
	num_channels_to_encode_	1	unsigned __int64
	next_frame_length_ms_	20	int
	complexity_	9	int
	▶ packet_loss_fraction_smoother_	unique_ptr {last_sample_time_ms_=175498943 smoother_={alpha_=0.999899983 filtered_=-1.00000000 max_=-1.00000000 } }	std::unique_ptr<webrtc::AudioEncoderOpusImpl::PacketLossFractionSmoother,std::default_delete<webrtc::AudioEncoderOpusImpl::PacketLossFractionSmoother> >
	▶ audio_network_adaptor_creator_	std::unique_ptr<webrtc::AudioNetworkAdaptor,std::default_delete<webrtc::AudioNetworkAdaptor> > <lambda>(const std::basic_string<char,std::char_traits<char>,std::allocator<char> > & config_string, webrtc::RtcEventLog * event_log){...}	const std::function<std::unique_ptr<webrtc::AudioNetworkAdaptor,std::default_delete<webrtc::AudioNetworkAdaptor> > __cdecl(std::basic_string<char,std::char_traits<char>,std::allocator<char> > const &,webrtc::RtcEventLog *)>
	▶ audio_network_adaptor_	empty	std::unique_ptr<webrtc::AudioNetworkAdaptor,std::default_delete<webrtc::AudioNetworkAdaptor> >
	▶ overhead_bytes_per_packet_	{...}	absl::optional<unsigned __int64>
	▶ bitrate_smoother_	unique_ptr {init_time_ms_=5000 init_factor_=0.998297989 init_const_=8.50878906 ...}	const std::unique_ptr<webrtc::SmoothingFilter,std::default_delete<webrtc::SmoothingFilter> >
	▶ bitrate_smoother_last_update_time_	{...}	absl::optional<__int64>
	▶ link_capacity_allocation_bps_	{...}	absl::optional<__int64>
	consecutive_dtx_frames_	0	int

 

 

	Name	Value	Type
◢	st	0x000000000c2a8070 {celt_enc_offset=38288 silk_enc_offset=18040 silk_mode={nChannelsAPI=1 nChannelsInternal=...} ...}	OpusEncoder *
	celt_enc_offset	38288	int
	silk_enc_offset	18040	int
	◢ silk_mode	{nChannelsAPI=1 nChannelsInternal=1 API_sampleRate=48000 ...}	silk_EncControlStruct
	nChannelsAPI	1	int
	nChannelsInternal	1	int
	API_sampleRate	48000	int
	maxInternalSampleRate	16000	int
	minInternalSampleRate	16000	int
	desiredInternalSampleRate	16000	int
	payloadSize_ms	20	int
	bitRate	22100	int
	packetLossPercentage	0	int
	complexity	9	int
	useInBandFEC	1	int
	LBRR_coded	0	int
	useDTX	0	int
	useCBR	0	int
	maxBits	922	int
	toMono	0	int
	opusCanSwitch	0	int
	reducedDependency	0	int
	internalSampleRate	16000	int
	allowBandwidthSwitch	1	int
	inWBmodeWithoutVariableLP	1	int
	stereoWidth_Q14	15360	int
	switchReady	0	int
	signalType	0	int
	offset	240	int
	application	2048	int
	channels	1	int
	delay_compensation	192	int
	force_channels	-1000	int
	signal_type	-1000	int
	user_bandwidth	-1000	int
	max_bandwidth	1105	int
	user_forced_mode	-1000	int
	voice_ratio	82	int
	Fs	48000	int
	use_vbr	1	int
	vbr_constraint	1	int
	variable_duration	5000	int
	bitrate_bps	32000	int
	user_bitrate_bps	32000	int
	lsb_depth	24	int
	encoder_buffer	480	int
	lfe	0	int
	arch	4	int
	use_dtx	0	int
	◢ analysis	{arch=4 application=2048 Fs=48000 ...}	TonalityAnalysisState
	arch	4	int
	application	2048	int
	Fs	48000	int
	▶ angle	0x000000000c2a8138 {0.000000000, 0.371065110, -0.178583264, 0.169607878, -0.336950421, -0.0553655699, ...}	float[240]
	▶ d_angle	0x000000000c2a84f8 {0.000000000, 0.211459041, 0.108894110, -0.0403590798, -0.0484426916, -0.0585320257, ...}	float[240]
	▶ d2_angle	0x000000000c2a88b8 {0.000000000, 0.00185681810, 0.000212006868, 0.000139717609, 0.000239213259, 0.000370358146, ...}	float[240]
	▶ inmem	0x000000000c2a8c78 {-0.00359834707, -0.00321650039, -0.00343828206, -0.00407011248, -0.00430583814, ...}	float[720]
	mem_fill	240	int
	▶ prev_band_tonality	0x000000000c2a97bc {0.0707340315, 0.256839573, 0.0646021217, 0.0535893366, 4.82304631e-06, 0.00638849102, ...}	float[18]
	prev_tonality	0.0744439214	float
	prev_bandwidth	16	int
	▶ E	0x000000000c2a980c {0x000000000c2a980c {2.55750088e-08, 9.36143074e-09, 2.76515277e-09, 1.35677691e-09, ...}, ...}	float[8][18]
	▶ logE	0x000000000c2a9a4c {0x000000000c2a9a4c {-17.4777470, -18.4760418, -19.6706448, -20.3470402, -21.3746700, ...}, ...}	float[8][18]
	▶ lowE	0x000000000c2a9c8c {-17.4777470, -18.4760418, -19.6706448, -20.3470402, -21.3746700, -20.4640789, -20.0291424, ...}	float[18]
	▶ highE	0x000000000c2a9cd4 {-10.1126461, -13.4705744, -15.9786558, -17.1679306, -17.9192219, -16.4453983, -14.3947802, ...}	float[18]
	▶ meanE	0x000000000c2a9d1c {1.69691320e-05, 1.21011362e-06, 3.28179901e-08, 9.97130112e-09, 4.68893901e-09, ...}	float[19]
	▶ mem	0x000000000c2a9d68 {-75.0991974, 4.48027182, 1.93349600, 4.91965246, 1.66545582, 4.41493130, -0.158232212, ...}	float[32]
	▶ cmean	0x000000000c2a9de8 {-73.6217041, 5.10436296, 1.28940845, 4.50601101, 0.000000000, 0.000000000, 0.000000000, ...}	float[8]
	▶ std	0x000000000c2a9e08 {0.282320261, 0.187619641, 0.852573156, 0.488151670, 1.46251786, 0.275115967, 0.825375319, ...}	float[9]
	music_prob	0.102564953	float
	vad_prob	0.0822941810	float
	Etracker	-0.0210000016	float
	lowECount	1.00000000	float
	E_count	7	int
	last_music	0	int
	count	7	int
	analysis_offset	0	int
	▶ pspeech	0x000000000c2a9e4c {0.896234274, 0.000000000, 0.000000000, 0.000000000, 0.000000000, 0.000000000, 0.000000000, ...}	float[100]
	▶ pmusic	0x000000000c2a9fdc {0.103052884, 0.000000000, 0.000000000, 0.000000000, 0.000000000, 0.000000000, 0.000000000, ...}	float[100]
	speech_confidence	0.100000001	float
	music_confidence	0.899999976	float
	speech_confidence_count	0	int
	music_confidence_count	0	int
	write_pos	7	int
	read_pos	7	int
	read_subframe	0	int
	hp_ener_accum	0.000000000	float
	▶ downmix_state	0x000000000c2aa18c {-0.00205630972, -0.00201843097, 0.00201843097}	float[3]
	▶ info	0x000000000c2aa198 {{valid=1 tonality=0.0630559996 tonality_slope=0.0293915570 ...}, {valid=1 tonality=...}, ...}	AnalysisInfo[100]
	stream_channels	1	int
	hybrid_stereo_width_Q14	16384	short
	variable_HP_smth2_Q15	193536	int
	prev_HB_gain	0.998388469	float
	▶ hp_mem	0x000000000c2ab788 {0.000429835636, -0.000430019340, 0.000000000, 0.000000000}	float[4]
	mode	1001	int
	prev_mode	1001	int
	prev_channels	1	int
	prev_framesize	960	int
	bandwidth	1104	int
	auto_bandwidth	1105	int
	silk_bw_switch	0	int
	first	0	int
	▶ energy_masking	0x0000000000000000 {???}	float *
	▶ width_mem	{XX=0.000000000 XY=0.000000000 YY=0.000000000 ...}	StereoWidthState
	▶ delay_buffer	0x000000000c2ab7d4 {-0.00312746083, -0.00290988642, -0.00281588663, -0.00290483702, -0.00314440276, ...}	float[960]
	detected_bandwidth	1104	int
	nb_no_activity_frames	0	int
	peak_signal_energy	0.000117333788	float
	nonfinal_frame	0	int
	rangeFinal	0	unsigned int