Documentation:Streaming HowTo/Advanced Streaming Using the Command Line

Structure of stream output

Stream output is the name of the feature of VLC that allows to output any stream read by VLC to a file or as a network stream instead of displaying it. Different kind of processing can be applied to the stream during this process (transcoding, re-scaling, filters, re-muxing...). Stream output includes different modules, each of them having different capabilities. You can chain modules to enhance the possibilities.

Here is the list of the modules currently available:

standard allows to send the stream via an access output module: for example, UDP, file, HTTP, ... You will probably want to use this module at the end of your chains.

transcode is used to transcode (decode and re-encode the stream using a different codec and/or bitrate) the audio and the video of the input stream. If the input or output access method doesn't allow pace control (network, capture devices), this done "on the fly", in real time. This can require quite a lot of CPU power, depending on the parameters set. Other streams, such as files and disks are transcoded as fast as the system allows it.

duplicate allows you to create a second chain, where the stream will be handled in an independent way.

display allows you to display the input stream, as VLC would normally do. Used with the duplicate module, this allows you to monitor the stream while processing it.

rtp streams over RTP (one UDP port for each elementary stream). This module also allows RTSP support.

es allows you to make separate Elementary Streams (ES) out of an input stream. This can be used to save audio and video streams to separate files, for instance.

bridge-out TODO
bridge-in TODO
mosaic-bridge TODO

Each of these modules may take options. Here is the syntax that you must use:

% vlc input_stream --sout "#module1{option1=parameter1{parameter-option1},option2=parameter2}:module2{option1=...,option2=...}:..."

NOTE: Some of the module options (option1 in the example) have to be set, others are optional. Option parameters (parameter-option1 in the example) are always optional. These option parameters are also often very advanced settings. If you don't understand their description, this certainly means that you don't need them.

You may also use the following syntax :

% vlc input_stream --sout-module1-option1=... --sout-module1-option2=... --sout-module2-option1=... --sout-module2-option2=... ...

For example, to transcode a stream and send it, use:

% vlc input_stream --sout '#transcode{options}:standard{options}'

Description of the modules

standard (alias std)

This module saves the stream to a file or sends it over a network, after having muxed it.

The available options are:

access

This option allows to set the medium used to save or send the stream. This is a compulsory option. Available options are:

file: saves the stream to a file.

Use the append option to append the stream to an existing file instead of replacing it.

udp: streams to a UDP unicast or multicast address.

Item options are:

- caching= to set the time VLC should buffer data before sending it;
- ttl=<ttl> to set the ttl of the sent udp packets;
- group=<amount of packets> to sent packets by burst instead of one by one;
- late= to drop packets that arrive too late at this stage of the chain;
- raw if you don't want to wait until the MTU is filled before sending the packet.

http: streams over HTTP.

Item options are:

- user=<user name> to enable HTTP basic authentication and set the user;
- pwd=<password> to set the basic authentication password;
- mime=<mime type> to set the mime type returned by the server.

https: streams over HTTP, using a secured SSL connection.

Item options are the same as for http and:

- cert=<path to certificate>to set the certificate to use;
- key=<path to key> to set the private key file the server should use for the SSL connection;
- ca=<path to certificate> to set the path to the root CA certificates to use for SSL;
- crl=<path to certificate> to set the revocation certificate to use for the SSL connection.

mmsh: streams using the Microsoft MMS protocol. This protocol is used as transport method by many Microsoft's softwares.

Note that only a small part of the MMS protocol is supported (MMS encapsulated in HTTP). Item options are the same as for the http module.

rtp: streams over RTP. Please note that this is a legacy module. It doesn't allow RTSP and can only be used to stream TS streams. Please have a look at the rtp stream output module for a more compliant and complete RTP streamer.

Options are the same as for the udp setting.

</itemizedlist> </sect3> <sect3><title>mux=</title>

This option allows you to set the encapsulation method used for the resulting stream. This option has to be set. Available options are:

<listitem>ts: the MPEG2/TS muxer. This the

standard muxer used to stream MPEG 2. This muxer can be used with any access method. Supported codecs are MPEG 1/2/4, MJPEG, H263, H264, I263, WMV 1/2 and theora for video, MPEG audio, AAC and

a52 for the audio stream. Item options are: pid-video=<pid>

to set the PID of the video track, pid-audio=<pid> to set the PID of the audio track, pid-spu=<pid> to set the PID of the subtitle track, pid-pmt=<pid> to set the PID of the PMT (Program Map Table), tsid=<id> to set the ID of the resulting TS stream, shaping=<shaping delay in ms> to set the minimum interval during which the bitrate of the stream will remain constant, for variable bitrate streams, use-key-frames uses I frames as limits for the shaping intervals, pcr=<PCR interval in ms> allows to set at which interval Program Clock References will be sent, dts-delay=<delay in ms> allows to delay PTS (Presentation Time Stamps) from the DTS (Decoding Time Stamp) from the given time, crypt-audio allows to enable encryption of the audio track using the CSA algorithm, csa-ck=<key as a 16 character word> allows to set the key used for CSA

encryption.</listitem> <listitem>ps: the MPEG2/PS muxer. This the standard muxer for MPEG

2 files(.mpg). It can be used with the file and http output methods. Supported codecs are MPEG 1/2 and MJPEG for video, MPEG audio and a52 for

audio streams. The only available item option is dst-delay=<delay

in ms>. It allows to delay PTS (Presentation Time Stamps) from the DTS (Decoding Time Stamp) from the given

time.</listitem> <listitem>mpeg1: the standard MPEG 1 muxer. This muxer should be

used instead of ps with MPEG 1 video streams, when saved to a file or

streamed over HTTP. Supported codecs are MPEG 1 and MPEG audio. Items options are the same as for the PS muxer.</listitem> <listitem>ogg: the ogg muxer. This is the muxer from the Xiph

project. It can be used with the HTTP and file output methods. Supported codecs are MPEG 1/2/4, MJPEG WMV 1/2 and Theora, audio streams can be vorbis,

flac, speex, a52 or MPEG audio. There is no item option for this muxer.</listitem> <listitem>asf: the Microsoft ASF muxer. This is the standard muxer

used for streaming by Microsoft's softwares. Is also used as container for WMA audio files. This muxer can be used with the file and HTTP output methods. Supported codecs are MPEG 4, MJPEG, WMV 1/2 for video,

MPEG audio, a52 for audio streams. Item options are:

title=<title>, autor=<author>, copyright=<copyright message>, comment=<comments>,

rating=<rating> allow you to set what will be displayed in the according field of the stream comments.</listitem> <listitem>asfh: this is a special version of the ASF muxer, that

should be used for MMSH streaming. MMSH is the only supported output

method. Supported codecs are the same as for ASF. Item options are the same as for ASF.</listitem> <listitem>avi: the Microsoft AVI muxer. This is very common

encapsulation format for MPEG 4 files. The only supported output method is file. Supported codecs are MPEG 1/2/4, H263, H264 and I263 for video,

MPEG audio and a52 for audio streams. There is not item option for this muxer.</listitem> <listitem>mpjpeg: the multipart jpeg muxer.

This encapsulation format is mostly used on surveillance video cameras with an integrated web-server. Such streams are usually embedded in web-pages and seen with standard Internet browsers, as they are seen as a succession of jpeg images. The only supported output method is HTTP. The only usable codec is MJPEG. No sound track can be muxed

in such streams. No item option is available for this muxer.</listitem>

</itemizedlist> </sect3>

This option allows to give informations about the location where the stream should actually be saved or sent. Here is the meaning of the dst option

depending on the parameter used for the access

option.

<listitem>If the file output method is used, dst is the complete path where the file should be saved.</listitem> <listitem>If the udp or rtp

output method is used, dst is the unicast or multicast destination address and, optionally, UDP port, in the form

address:port.</listitem> <listitem>If the http,

https or mmsh output method is chosen, dst is the address, port and path of the local network interface on which the server should listen for requests. If no address is given, VLC will listen on all the network interfaces. These information have to be supplied using the address:port/path

syntax.</listitem>

</itemizedlist>

</sect3>

Use this option if you want VLC to send SAP (Session Announcement

Protocol) announces. SAP is a service discovery protocol, that uses a special multicast address to send a list of available streams on a

server. <note>This option can only be enabled with the udp output method.</note>

</sect3>

<sect3><title>group=</title>

This option allows to specify the name of an optional group of streams. A VLC used as a client will use this field to classify the stream. <note>This option uses a private extension of the SAP protocol. VLC will be the only client able to read this field. This option can only be used it the sap option has been enabled.</note>

</sect3>

Use this option if you want the SAP announces to be sent using the IPv6 protocol instead of IPv4. <note>This option can only be used it the sap option has been enabled.</note>

</sect3>

SLP stands for Service Location Protocol. It is an alternative to SAP for session announcement. Use this option if you want to send such announces.

</sect3>

Use this option to specify the name of the stream that will be sent in SAP and SLP announces. <note>This option can only be used it the sap or slp option has been enabled.</note>

</sect3>

</sect2>

<sect2><title>display</title>

This module can be used to display the stream. This is

particularly useful in a duplicate chain, in order

to monitor a stream while it is being saved or streamed. available options are:

<sect3><title>novideo</title>

You can use this option to disable video in the displayed stream.

</sect3>

<sect3><title>noaudio</title>

You can use this option to disable audio in the displayed stream.

</sect3>

<sect3><title>delay=</title>

You can use this option to introduce a delay in the display of the stream. Delay has to be given in ms.

</sect3>

</sect2>

This module can be used to send a stream using the RTP (Real Time Protocol) protocol (see RFC 3550). <note>Although use of RTSP is possible using

this module, it won't allow you to make Video On demand. Please have a

look at the description of the VLM module for that.</note> The different available options are:

This option allow to give the destination UDP address. This can

be the address of a host or a multicast group. This option has to be given, unless the sdp=rtsp://option is given (see below). In the latter case, the stream will be sent to the host doing the

RTSP request.

</sect3>

This option allows to set the UDP port used to send the first elementary stream. This port has to be even. Other streams will be streamed using even ports directly above this one.

</sect3>

<sect3><title>port-video=</title>

This option allows to set the UDP port used to send the first

video elementary stream. This port has to be

even.

</sect3>

<sect3><title>port-audio=</title>

This option allows to set the UDP port used to send the first

audio elementary stream. This port has to be

even.

</sect3>

This option allows to set the way the SDP (Session Description Protocol) file corresponding the the stream should be made available. Options are:

<listitem>file://<path to the file>, to export the SDP as a local file</listitem> <listitem>http://<local interface

IP:port/path>, to make the file available using the

integrated HTTP server of VLC. <note>The local interface IP argument

is optional. If not given, VLC will listen on all available

interfaces.</note></listitem> <listitem>rtsp://<local interface

IP:port/path>, to make the SDP file available using the

RTSP protocol (see RFC 2326). <note>The local interface IP argument

is optional. If not given, VLC will listen on all available

interfaces.</note></listitem> <listitem>sap, to export the SDP using the SAP (Session Announcement Protocol, see RFC 2974).</listitem>

</itemizedlist>

</sect3>

This option can be used to set the TTL (Time to Live) of the sent UDP packets.

</sect3>

This option allows to set the encapsulation method used to

send the stream. See mux= options of the standard module for a description of the available method. Only ts and raw (no encpasulation) are possible values for RTP

streams. Default is RAW.

</sect3>

This option can be used to set the name that will be displayed on the client receiving the stream.

</sect3>

<sect3><title>description=</title>

This option can be used to give an additional description of the stream.

</sect3>

This option allows to give the address of a website with additional informations about the stream.

</sect3>

<sect3><title>email=</title>

This options allows to give a contact e-mail address

</sect3>

</sect2>

The es module can be used to separate the

different elementary streams from a stream, and save each of them in a different file or send it to a separate

destination. The available parameters are:

<sect3><title>access-video=</title>

Use this option to set the medium used to save or send the video

elementary streams. Possible values and item options are the same as for the access option of

the standard module (see above).

</sect3>

<sect3><title>access-audio=</title>

Use this option to set the medium used to save or send the audio

elementary streams. Possible values and item options are the same than for the access option of

the standard module (see above).

</sect3>

<sect3><title>access=</title>

This option can be used instead of both

access-video and access-audio

options, when they share the same setting.

</sect3>

<sect3><title>mux-video=</title>

Use this option to set the encapsulation method used for the

video elementary streams. Possible values and item options are the same as for the mux option of

the standard module (see above).

</sect3>

<sect3><title>mux-audio=</title>

Use this option to set the encapsulation method used for the

audio elementary streams. Possible values and item options are the same than for the mux option of the

standard module (see above).

</sect3>

This option can be used instead of both

mux-video and mux-audio

options, when they share the same setting.

</sect3>

<sect3><title>dst-video=</title>

Use this option to set the location where the video

elementary streams should be saved, sent, or made available. The exact meaning of this option depends on the value of the access-video option and is the same as for the url option of the standard

module (see above). <note>If you use the %d string in the url

field, VLC will replace it by the number of the audio or video track considered. The %c string will be replaced by the

name (FOURCC) of the codec of the track.</note>

</sect3>

<sect3><title>dst-audio=</title>

Use this option to set the location where the audio

elementary streams should be saved, sent, or made available. The exact meaning of this option depends on the value of the access-audio option and is the same as for the url option of the standard

module (see above). <note>If you use the %d string in the url

field, VLC will replace it by the number of the audio or video track considered. The %c string will be replaced by the

name (FOURCC) of the codec of the track.</note>

</sect3>

This option can be used instead of both

dst-video and dst-audio

options, when they share the same setting.

</sect3>

</sect2>

<sect2><title>transcode</title>

You can use this module to transcode a stream, i.e. to change its

codecs or the encoding bitrates. Some additional processing can be done during this process, such as re-scaling, deinterlacing, resampling,

etc. <note>Depending on the bitrate of the original stream and of

the options chosen, transcoding can be a very CPU intensive task. As a consequence, streaming of a real time transcoded stream can lead to dropped frames or a jerky image and sound in some cases, when running

out of resources.</note> Available options are:

<sect3><title>vcodec=</title>

This options allows to specify the codec the video tracks of the input stream should be transcoded to. List of available codecs can be found on the <ulink url="/streaming/features.html">streaming features page</ulink>.

</sect3>

This option allows to set the bitrate of the transcoded video stream, in kbit/s

</sect3>

This allows to set the encoder to use to encode the videos stream. Available options are:

<listitem>ffmpeg: this is the libavcodec encoding

module. It handles a large variety of different codecs (the list can be found on the <ulink url="/streaming/features.html">streaming features

page</ulink>. Item options are: keyint=<number of

frames> allows to set the maximal amount of frames between 2 key frames, hurry-up allows the encoder to decrease the quality of the stream if the CPU can't keep up with the encoding rate, interlace allows to improve the quality of the encoding of interlaced streams, noise-reduction=<noise reduction factor> enables a noise reduction algorithm (will decrease required bitrate at the cost of details in the image), vt=<bitrate tolerance in kbit/s> allows to set a tolerance for the bitrate of the outputted video stream, bframes=<amount of frames> allows to set the amount of B frames between 2 key frames, qmin=<quantizer> allows to set the minimum quantizer scale, qmax=<quantizer> allows to set the maximum quantizer scale, qscale=<quantizer scale> allows to specify a fixed quantizer scale for VBR encodings, i-quant-factor=<quantization factor> allows to set the quantization factor of I frames, compared to P frames, hq=<quality> allows to choose the quality level for the encoding of the motion vectors (arguments are simple, rd or bits, default is simple

FIXME*), strict=<level of compliance>

allows to force a stricter standard compliance (possible values are -1, 0 and 1, default is 0), strict-rc enables a strict rate control algorithm, rc-buffer-size=<size of the buffer in bits> allows to choose the size of the buffer used for rate control (bigger means more efficient rate control), rc-buffer-aggressivity=<float representing the aggressiveness> allows to set the rate control buffer aggressiveness *FIXME*, pre-me allows to enable pre motion estimation, mpeg4-matrix enable use of the MPEG4 quantization matrix with MPEG2 streams, improving quality while keeping compatibility with MPEG2 decoders, trellis enables trelli quantization (better quality, but slower

processing).</listitem> <listitem>theora: The Xiph.org theora encoder.

The module is used to produce theora streams. Theora is a free patent

and royalties free video codec. The only available item option is quality=<quality

level>. This option allows to create a VBR stream, overriding vb setting. the quality level must be an

integer between 1 and 10. Higher is better.</listitem> <listitem>x264. x264 is a free open-source h264 encoder. h264 (or MPEG4-AVC) is a quite recent high quality video codec. Item options are: keyint=<number of

frames> allows to set the maximal amount of frames between 2 key frames, idrint=<number of frames> allows to set the maximal amount of frames between 2 IDR frames, bframes=<amount of frames> allows to set the amount of B frames between an I and a P frame, qp=<quantizer parameter> allows to specify a fixed quantizer (between 1 and 51), qp-max=<quantizer parameter> allows to set the maximum value for the quantizer, qp-min=<quantizer parameter> allows to set the minimum value for the quantizer, cabac enables the CABAC (Context-Adaptive Binary Arithmetic Coding) algorithm (slower, but enhances quality), loopfilter enables deblocking loop filter, analyse enables the analyze mode, frameref=<amount of frames> allows to set the number of previous frames used as predictors, scenecut=<sensibility< allows to control how aggressively the encoder should insert extra I-frame, on scene

change.</listitem>

</itemizedlist>

</sect3>

This options allows to set the framerate of the transcoded

video, in frame per second. reducing the framerate of a video can help

decreasing its bitrate.

</sect3>

<sect3><title>deinterlace</title>

This option allows to enable deinterlacing of interlaced video streams before encoding.

</sect3>

<sect3><title>croptop=</title>

This option allows to crop the upper part of the source video while

transcoding. The argument is the number of lines the video should be

cropped.

</sect3>

<sect3><title>cropbottom=</title>

This option allows to crop the lower part of the source video. The argument is the Y coordinate of the first line to be cropped.

</sect3>

<sect3><title>cropleft</title>

This option allows to crop the left part of the source video while

transcoding. The argument is the number of columns the video should be

cropped.

</sect3>

<sect3><title>cropright=</title>

This option allows to crop the right part of the source video. The argument is the X coordinate of the first column to be cropped.

</sect3>

<sect3><title>scale=</title>

This option allows the give the ratio from which the video should

be rescaled while being transcoded. This option can be particularly

useful to help reduce the bitrate of a stream.

</sect3>

<sect3><title>width=</title>

This options allows to give the width of the transcoded video in pixels.

</sect3>

<sect3><title>height</title>

This options allows to give the height of the transcoded video, in pixels.

</sect3>

<sect3><title>acodec=</title>

This options allows to specify the codec the audio tracks of the input stream should be transcoded to. List of available codecs can be found on the <ulink url="/streaming/features.html">streaming features page</ulink>.

</sect3>

This option allows to set the bitrate of the transcoded audio stream, in kbit/s

</sect3>

This allows to set the encoder to use to encode the audio stream. Available options are:

<listitem>ffmpeg: this is the libavcodec encoding

module. It handles a large variety of different codecs (the list can be found on the <ulink url="/streaming/features.html">streaming features

page</ulink>.</listitem> <listitem>vorbis. This module uses the vorbis

encoder from the Xiph.org project. Vorbis is a free, open, license-free

lossy audio codec. Item options are: quality=<quality

level> allows to use VBR (variable bitrate) encoding instead of the default CBR (constant bitrate), and to set the quality level (between 1 and 10, higher is better), max-bitrate=<bitrate in kbit/s> allows to set the maximum bitrate, for vbr encoding, min-bitrate=<bitrate in kbit/s> allows to set the minimum bitrate, for vbr encoding, cbr allows to force cbr

encoding.</listitem> <listitem>speex. This module uses the speex

encoder from the Xiph.org project. Speex is a lossy audio codec, best fit for very low bitrates (around 10 kbit/s) and particularly video

conference.</listitem>

</itemizedlist>

</sect3>

<sect3><title>samplerate=</title>

This option allows to set the samplerate of the transcoded audio

stream, in Hz. Reducing the samplerate is be a way to lower the bitrate

of the resulting audio stream.

</sect3>

<sect3><title>channels=</title>

This option allows to set the number of channels of the resulting

audio stream. This is useful for codecs that don't have support for more

than 2 channels, of to lower the bitrate of an audio stream.

</sect3>

<sect3><title>scodec=</title>

This options allows to specify subtitle format the subtitles tracks of the input stream should be converted to.

List of available codecs can be found on the <ulink url="/streaming/features.html">streaming features page</ulink>.

</sect3>

This allows to set the converter to use to encode the subtitle stream.

The only subtitle encoder we have at this time is dvbsub.

</sect3>

<sect3><title>soverlay</title>

This option allow to render subtitles directly on the video,

while transcoding it. Do not confuse this option with senc/scodec that

transcode the subtitles and streams them

</sect3>

<sect3><title>sfilter=</title>

This option allows to render some images generated by a so called

subpicture filter (e.g. a logo, a text string,

etc) on top of the video. List of available subpicture filters can be

found on the <ulink url="http://www.videolan.org/streaming/features.html">streaming features

page</ulink>. The Item options of this modules can be found using the following command line:

<prompt>% </prompt><userinput>vlc -p --advanced <module name></userinput>

</sect3>

<sect3><title>threads=</title>

This options allows to set the amount of threads that should

be used to encode the streams. Increasing this number to the amount of processors on the computer, (or twice this number on Intel P4 HT

processors) should improve transcoding performance.

</sect3>

<sect3><title>audio-sync</title>

When this option is enabled, VLC will drop/duplicate video frames to synchronize the video track on the audio track.

</sect3>

</sect2>

<sect2><title>duplicate</title>

This module can be used to duplicate the stream, and so process it through several different chains. Available options are:

This options allows to give the chain through which the duplicated stream should be processed. <note>Several dst= options have to be used in the same duplicate block to actually duplicate the stream. Any of the stream output module described earlier can be used as parameter of this option.</note>

</sect3>

<sect3><title>select=</title>

This options can be used to duplicate only a part elementary streams of a complete stream. Several criteria can be given, by separating each of them with a comma. For criteria that need a parameter, such as

es and program, you can also specify a range, using the syntax

criteria=num_start-num_end. Available parameters are:

<listitem>program=: duplicate only

elementary streams belonging to the selected program (or SID). This option only works with MPEG2/TS

streams.</listitem> <listitem>noprogram=: do not duplicate

elementary streams belonging to the selected program (or PID). This option only works with MPEG2/TS

streams.</listitem> <listitem>es=: duplicate only the

elementary stream with the selected

id.</listitem> <listitem>noes=: do not duplicate

the elementary stream with the selected

id.</listitem> <listitem>video: duplicate only video elementary streams.</listitem> <listitem>novideo: do not duplicate video elementary streams.</listitem> <listitem>audio: duplicate only audio elementary streams.</listitem> <listitem>noaudio: do not duplicate audio elementary streams.</listitem> <listitem>spu: duplicate only subtitle elementary streams.</listitem> <listitem>nospu: do not duplicate subtitle elementary streams.</listitem>

</itemizedlist>

Example:

#duplicate{dst=std{...},select="program=100-200,novideo"}

This duplicate chain will only output the non

video elementary streams belonging to the programs

which PID are between 100 and 200.

</sect3>

</sect2>

<sect2><title> Miscellaneous </title>

Here are a few additional global options:

Enable streaming of all ES (default disabled). By default VLC will

only stream one audio ES and one video ES (the first ones). If you

enable sout-all, all ES (audio, video and SPU) will be streamed.

</sect3>

Keep sout open (default disabled) : use the same sout instance across the various playlist items, if possible.

</sect3>

<sect3><title>--no-sout-audio</title>

This options allows to disable audio in the outputted stream.

</sect3>

<sect3><title>--no-sout-video</title>

This options allows to disable video in the outputted stream.

</sect3>

</sect2>

<sect2><title>Simplified Syntax</title>

The stream output also offers a simplified syntax, with which you can only you use the standard module main options:

<prompt>% </prompt><userinput>vlc input_stream --sout access/mux:url</userinput>

where access,

mux and url are as defined in the options of the standard module.

</sect2>

</sect1>

<sect1><title> Examples </title>

To understand fully the complex syntax of VLC's stream output, please look at the examples in the next section.

</sect1>

</chapter>

This page is part of official VLC media player Documentation (User Guide • Streaming HowTo • Hacker Guide • Modules)
Please read the Documentation Editing Guidelines before you edit the documentation

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.

Documentation:Streaming HowTo/Advanced Streaming Using the Command Line

Contents

Structure of stream output

Description of the modules

standard (alias std)

access

Navigation menu

Personal tools

Namespaces

Variants

Views

More

Search

Help / Documentation

Development

VideoLAN wiki

Tools