RTSPMediaSourceMTS Class Reference

IMediaSource for an MPEG transport stream, e.g. live video and audio muxed by MuxerMTS: the muxer writes its output to this object (it is a Print), and an RTSPMediaStreamer sends it as one RTP track (MP2T, RFC 2250 - see RTSPFormatMTS) to the RTSP client, which demuxes video and audio itself. More...

#include <RTSPMediaSourceMTS.h>

Inheritance diagram for RTSPMediaSourceMTS:

Public Member Functions
	RTSPMediaSourceMTS (int bufferPackets=128, float sendRate=100)
virtual int	availableForWrite ()
virtual int	availableForWrite ()
uint32_t	droppedPackets ()
	Number of TS packets dropped because the buffer was full.
uint32_t	droppedVideoPackets ()
	Number of video TS packets dropped to keep the audio going.
virtual void	flush ()
virtual void	flush ()
RTSPFormat &	getFormat () override
	Get the media format configuration.
int	packetSize () override
	Size of the next RTP payload: up to kPacketsPerFragment TS packets.
int	readBytes (void *dest, int maxBytes) override
	Provides whole TS packets only.
void	start () override
void	stop () override
	Called when the client stops streaming.
size_t	write (const uint8_t *data, size_t len) override
	Accepts the transport stream, in any chunk sizes.
size_t	write (uint8_t c) override

Static Public Attributes
static constexpr int	kPacketSize = 188
static constexpr int	kPacketsPerFragment = 7

Protected Member Functions
void	addPacket ()
bool	isVideoPes ()

Protected Attributes
int	_timeout = 10
RingBufferSPSC< uint8_t >	buffer
int	capacity = 0
uint32_t	dropped = 0
uint32_t	dropped_video = 0
bool	dropping_video = false
RTSPFormatMTS	format
volatile bool	is_active = false
uint8_t	packet [kPacketSize]
size_t	packet_len = 0
int	video_pid = -1

Detailed Description

IMediaSource for an MPEG transport stream, e.g. live video and audio muxed by MuxerMTS: the muxer writes its output to this object (it is a Print), and an RTSPMediaStreamer sends it as one RTP track (MP2T, RFC 2250 - see RTSPFormatMTS) to the RTSP client, which demuxes video and audio itself.

Unlike RTSPOutput's generic byte stream, this source is meant for a live stream written from other tasks (e.g. by VideoMuxerWithTasks):

• the data is kept as complete 188 byte TS packets in a fixed size, lock-free single producer/single consumer ring buffer (RingBufferSPSC), and every RTP packet contains whole TS packets (up to 7: 1316 bytes, below the usual MTU);
• it is a packetized source (packetSize()), so the streamer sends all queued packets at each timer tick instead of a fixed amount: the send rate adapts to the bitrate of the stream;
• while no client is streaming, the data is dropped, so a new client doesn't get old data;
• if the network is slower than the stream, video is dropped rather than audio: when the buffer is filled to 3/4, the video packets are dropped from the start of the next video frame on, until the buffer is half empty again and the next keyframe starts (MuxerMTS writes the PAT before each keyframe). So the picture freezes for a moment while the sound continues. Audio and the other packets are only dropped if the buffer is completely full. Audio and video are recognized by the stream id of their PES packets, so this works with any transport stream. See droppedPackets() and droppedVideoPackets(). A client starts decoding at the next keyframe: MuxerMTS repeats the stream description (PAT/PMT) before each one.

The data can be written from any task, but not from several tasks at the same time - e.g. by a MuxerMTS driven by VideoMuxerWithTasks, which serializes the writes of its audio and video task with its mutex.

Author

Phil Schatzmann

Copyright

GPLv3

Constructor & Destructor Documentation

RTSPMediaSourceMTS()

RTSPMediaSourceMTS ( int bufferPackets = 128, float sendRate = 100 )

inline

Parameters

bufferPackets	size of the buffer in TS packets
sendRate	number of times per second the queued packets are sent

Member Function Documentation

addPacket()

void addPacket ( )

inlineprotected

availableForWrite() [1/2]

virtual int availableForWrite ( )

inlinevirtualinherited

Reimplemented in A2DPStream, AbstractURLStream, AdapterAudioStreamToAudioOutput, AnalogAudioStream, AdapterAudioStreamToAudioOutput, AudioOutput, AudioOutput, AudioStreamWrapper, AudioStreamWrapper, BaseStream, BaseStream, ConverterStream< T >, ConverterStream< T >, CsvOutput< T >, CsvOutput< T >, DynamicMemoryStream, DynamicMemoryStream, EncodedAudioOutput, EncodedAudioStream, FilteredStream< T, TF >, FilteredStream< T, TF >, FilteredStream< int16_t, int16_t >, MemoryOutput, MemoryOutput, MemoryStream, MemoryStream, OutputMixer< T >, OutputMixer< T >, ProgressStream, ProgressStream, QueueStream< T >, QueueStream< T >, QueueStream< uint8_t >, ReformatBaseStream, RingBufferStream, RingBufferStream, Throttle, Throttle, TimedStream, AudioBLEClient, AudioBLEClient, AudioBLEServer, AudioBLEServer, AudioEffectStream, AudioEffectStreamT< T >, AudioESP32ULP, AudioFFTBase, AudioLoRa, AudioOutput, AudioServerEx, AudioStreamWrapper, AudioSyncWriter, AutomaticGainControlStream, BaseStream, ChannelFormatConverterStream, ChannelFormatConverterStreamT< T >, ConverterStream< T >, CsvOutput< T >, DynamicMemoryStream, EncodedAudioOutput, EncodedAudioStream, Equalizer3BandsPerChannelT< T >, Equalizer3BandsPerChannelT< float >, Equalizer3BandsT< T >, Equalizer3BandsT< float >, ESPNowStream, FadeStream, FaustStream< DSP >, FileOutput, FilteredStream< T, TF >, FrequencyDetectorAutoCorrelation, FrequencyDetectorZeroCrossing, HammingFEC< bytecount, block_t >, I2SCodecStream, I2SStream, MeasuringStream, MemoryOutput, MemoryStream, MiniAudioStream, NRF24Stream, NumberFormatConverterStream, NumberFormatConverterStreamT< TFrom, TTo >, OutputMixer< T >, Pipeline, ProgressStream, PWMAudioOutput< PWMDriverT >, QualityAnalysisStream, QueueStream< T >, ReadioHeadStream, ReedSolomonFEC< bytecount, additional_bytes >, ReformatBaseStream, RingBufferStream, RTSPOutput< Platform >, SPIAudioMaster, SPIAudioSlave, StdioStream, SupportedRatesStream, SynchronizedStream, Throttle, TimedStream, UDPStream, URLStreamESP32, USBAudioStreamBase, VBANStream, and VolumeStream.

availableForWrite() [2/2]

virtual int availableForWrite ( )

inlinevirtualinherited

Reimplemented in A2DPStream, AbstractURLStream, AdapterAudioStreamToAudioOutput, AnalogAudioStream, AdapterAudioStreamToAudioOutput, AudioOutput, AudioOutput, AudioStreamWrapper, AudioStreamWrapper, BaseStream, BaseStream, ConverterStream< T >, ConverterStream< T >, CsvOutput< T >, CsvOutput< T >, DynamicMemoryStream, DynamicMemoryStream, EncodedAudioOutput, EncodedAudioStream, FilteredStream< T, TF >, FilteredStream< T, TF >, FilteredStream< int16_t, int16_t >, MemoryOutput, MemoryOutput, MemoryStream, MemoryStream, OutputMixer< T >, OutputMixer< T >, ProgressStream, ProgressStream, QueueStream< T >, QueueStream< T >, QueueStream< uint8_t >, ReformatBaseStream, RingBufferStream, RingBufferStream, Throttle, Throttle, TimedStream, AudioBLEClient, AudioBLEClient, AudioBLEServer, AudioBLEServer, AudioEffectStream, AudioEffectStreamT< T >, AudioESP32ULP, AudioFFTBase, AudioLoRa, AudioOutput, AudioServerEx, AudioStreamWrapper, AudioSyncWriter, AutomaticGainControlStream, BaseStream, ChannelFormatConverterStream, ChannelFormatConverterStreamT< T >, ConverterStream< T >, CsvOutput< T >, DynamicMemoryStream, EncodedAudioOutput, EncodedAudioStream, Equalizer3BandsPerChannelT< T >, Equalizer3BandsPerChannelT< float >, Equalizer3BandsT< T >, Equalizer3BandsT< float >, ESPNowStream, FadeStream, FaustStream< DSP >, FileOutput, FilteredStream< T, TF >, FrequencyDetectorAutoCorrelation, FrequencyDetectorZeroCrossing, HammingFEC< bytecount, block_t >, I2SCodecStream, I2SStream, MeasuringStream, MemoryOutput, MemoryStream, MiniAudioStream, NRF24Stream, NumberFormatConverterStream, NumberFormatConverterStreamT< TFrom, TTo >, OutputMixer< T >, Pipeline, ProgressStream, PWMAudioOutput< PWMDriverT >, QualityAnalysisStream, QueueStream< T >, ReadioHeadStream, ReedSolomonFEC< bytecount, additional_bytes >, ReformatBaseStream, RingBufferStream, RTSPOutput< Platform >, SPIAudioMaster, SPIAudioSlave, StdioStream, SupportedRatesStream, SynchronizedStream, Throttle, TimedStream, UDPStream, URLStreamESP32, USBAudioStreamBase, VBANStream, and VolumeStream.

droppedPackets()

uint32_t droppedPackets ( )

inline

Number of TS packets dropped because the buffer was full.

droppedVideoPackets()

uint32_t droppedVideoPackets ( )

inline

Number of video TS packets dropped to keep the audio going.

flush() [1/2]

virtual void flush ( )

inlinevirtualinherited

Reimplemented in AudioOutput, AudioOutput, AudioStreamWrapper, AudioStreamWrapper, BaseStream, BaseStream, BufferedStream, BufferedStream, EncodedAudioStream, FilteredStream< T, TF >, FilteredStream< T, TF >, FilteredStream< int16_t, int16_t >, GeneratedSoundStream< T >, GeneratedSoundStream< T >, GeneratedSoundStream< int16_t >, HexDumpOutput, HexDumpOutput, MemoryStream, MemoryStream, MultiOutput, QueueStream< T >, QueueStream< T >, QueueStream< uint8_t >, ReformatBaseStream, RingBufferStream, RingBufferStream, TimedStream, AudioEffectStream, AudioEffectStreamT< T >, AudioOutput, AudioStreamWrapper, AutomaticGainControlStream, BaseStream, BufferedStream, BufferedTaskStream, EncodedAudioStream, Equalizer3BandsPerChannelT< T >, Equalizer3BandsPerChannelT< float >, Equalizer3BandsT< T >, Equalizer3BandsT< float >, EqualizerNBands< SampleT, AccT, NUM_TAPS, NUM_BANDS >, FadeStream, FilteredStream< T, TF >, GeneratedSoundStream< T >, HDLCStream, HexDumpOutput, I2SStream, MeasuringStream, MemoryStream, MultiOutput, NamePrinter, Pipeline::ModifyingStreamAdapter, PureDataStream, QualityAnalysisStream, QueueStream< T >, ReformatBaseStream, ResampleStream, RingBufferStream, SupportedRatesStream, TimedStream, URLStream, URLStreamBufferedT< T >, URLStreamBufferedT< ICYStream >, URLStreamBufferedT< ICYStreamESP32 >, URLStreamBufferedT< URLStream >, URLStreamBufferedT< URLStreamESP32 >, VFSFile, VolumeStream, and ZephyrFile.

flush() [2/2]

virtual void flush ( )

inlinevirtualinherited

Reimplemented in AudioOutput, AudioOutput, AudioStreamWrapper, AudioStreamWrapper, BaseStream, BaseStream, BufferedStream, BufferedStream, EncodedAudioStream, FilteredStream< T, TF >, FilteredStream< T, TF >, FilteredStream< int16_t, int16_t >, GeneratedSoundStream< T >, GeneratedSoundStream< T >, GeneratedSoundStream< int16_t >, HexDumpOutput, HexDumpOutput, MemoryStream, MemoryStream, MultiOutput, QueueStream< T >, QueueStream< T >, QueueStream< uint8_t >, ReformatBaseStream, RingBufferStream, RingBufferStream, TimedStream, AudioEffectStream, AudioEffectStreamT< T >, AudioOutput, AudioStreamWrapper, AutomaticGainControlStream, BaseStream, BufferedStream, BufferedTaskStream, EncodedAudioStream, Equalizer3BandsPerChannelT< T >, Equalizer3BandsPerChannelT< float >, Equalizer3BandsT< T >, Equalizer3BandsT< float >, EqualizerNBands< SampleT, AccT, NUM_TAPS, NUM_BANDS >, FadeStream, FilteredStream< T, TF >, GeneratedSoundStream< T >, HDLCStream, HexDumpOutput, I2SStream, MeasuringStream, MemoryStream, MultiOutput, NamePrinter, Pipeline::ModifyingStreamAdapter, PureDataStream, QualityAnalysisStream, QueueStream< T >, ReformatBaseStream, ResampleStream, RingBufferStream, SupportedRatesStream, TimedStream, URLStream, URLStreamBufferedT< T >, URLStreamBufferedT< ICYStream >, URLStreamBufferedT< ICYStreamESP32 >, URLStreamBufferedT< URLStream >, URLStreamBufferedT< URLStreamESP32 >, VFSFile, VolumeStream, and ZephyrFile.

getFormat()

RTSPFormat & getFormat ( )

inlineoverridevirtual

Get the media format configuration.

Returns the RTSPFormat object that describes the media data characteristics including sample rate, bit depth, number of channels, and RTP packaging parameters. If no format has been explicitly set, creates a default 16-bit PCM format at 16kHz mono.

Returns

Pointer to RTSPFormat describing the media characteristics

Note

Default format: 16-bit PCM, 16000 Hz, 1 channel

See also

setFormat(), RTSPFormatPCM

Implements IMediaSource.

isVideoPes()

bool isVideoPes ( )

inlineprotected

True if the current packet starts a video PES packet (stream id 0xE0-0xEF)

packetSize()

int packetSize ( )

inlineoverridevirtual

Size of the next RTP payload: up to kPacketsPerFragment TS packets.

Reimplemented from IMediaSource.

readBytes()

int readBytes ( void * dest, int maxBytes )

inlineoverridevirtual

Provides whole TS packets only.

Implements IMediaSource.

start()

void start ( )

inlineoverridevirtual

Called when a client starts streaming, before the first readBytes(): discards the old data (on the consumer side - reset() would not be safe while the producer is writing)

Reimplemented from IMediaSource.

stop()

void stop ( )

inlineoverridevirtual

Called when the client stops streaming.

Reimplemented from IMediaSource.

write() [1/2]

size_t write ( const uint8_t * data, size_t len )

inlineoverridevirtual

Accepts the transport stream, in any chunk sizes.

Reimplemented from Print.

write() [2/2]

size_t write ( uint8_t c )

inlineoverride

Member Data Documentation

_timeout

int _timeout = 10

protectedinherited

buffer

RingBufferSPSC<uint8_t> buffer

protected

capacity

int capacity = 0

protected

dropped

uint32_t dropped = 0

protected

dropped_video

uint32_t dropped_video = 0

protected

dropping_video

bool dropping_video = false

protected

format

RTSPFormatMTS format

protected

is_active

volatile bool is_active = false

protected

kPacketSize

int kPacketSize = 188

staticconstexpr

kPacketsPerFragment

int kPacketsPerFragment = 7

staticconstexpr

packet

uint8_t packet[kPacketSize]

protected

packet_len

size_t packet_len = 0

protected

video_pid

int video_pid = -1

protected

---

The documentation for this class was generated from the following file:

• src/AudioTools/Communication/RTSP/RTSPMediaSourceMTS.h