Dependencies¶

To install Livestreamer from source you will need these dependencies.

Installing without root permissions¶

If you do not wish to install Livestreamer globally on your system it’s recommended to use virtualenv to create a user owned Python environment instead.

Creating an environment
$ virtualenv ~/myenv

Activating the environment
$ source ~/myenv/bin/activate

Installing livestreamer into the environment
(myenv)$ pip install livestreamer

Using livestreamer in the enviroment
(myenv)$ livestreamer ...

Deactivating the enviroment
(myenv)$ deactivate

Using livestreamer without activating the environment
$ ~/myenv/bin/livestreamer ...

Installer¶

This is a installer which contains:

• A compiled version of Livestreamer that does not require an existing Python installation
• RTMPDump for viewing RTMP streams

and performs the following tasks:

• Generates a default configuration file
• Adds Livestreamer to your $PATH (making it possible to use livestreamer directly from the command prompt without specifying its directory)

Zip archive¶

This is minimal zip archive containing a compiled version of Livestreamer that does not require an existing Python installation.

Nightly build¶

This is an automatically generated build of the latest development code from the git repo.

Syntax¶

The config file is a simple text file and should contain one command-line option (omitting the dashes) per line in the format:

option=value

or for a option without value:

option

Example¶

# Player options
player=mpv --cache 2048
player-no-close

# Authenticate with Twitch
twitch-oauth-token=mytoken

Examples¶

Have a look at the list of plugins to see the name of each built-in plugin.

Authenticating with Twitch¶

It’s possible to access subscription content on Twitch by giving Livestreamer access to your account.

Authentication is done by creating an OAuth token that Livestreamer will use to access your account. It’s done like this:

$ livestreamer --twitch-oauth-authenticate

This will open a web browser where Twitch will ask you if you want to give Livestreamer permission to access your account, then forwards you to a page with further instructions on how to use it.

Authenticating with Crunchyroll¶

Crunchyroll requires authenticating with a premium account to access some of their content. To do so, the plugin provides a couple of options to input your information, --crunchyroll-username and --crunchyroll-password.

You can login like this:

$ livestreamer --crunchyroll-username=xxxx --crunchyroll-password=xxx http://crunchyroll.com/a-crunchyroll-episode-link

Once logged in, the plugin makes sure to save the session credentials to avoid asking your username and password again.

Neverthless, these credentials are valid for a limited amount of time, so it might be a good idea to save your username and password in your configuration file anyway.

HTTP proxy with Crunchyroll¶

You can use the --http-proxy and --https-proxy options (you need both since the plugin uses both protocols) to access the Crunchyroll servers through a proxy to be able to stream region locked content.

When doing this, it’s very probable that you will get denied to access the stream; this occurs because the session and credentials used by the plugin where obtained when logged from your own region, and the server still assumes you’re in that region.

For this, the plugin provides the --crunchyroll-purge-credentials option, which removes your saved session and credentials and tries to log in again using your username and password.

Supported streaming protocols¶

Positional arguments¶

[URL]

A URL to attempt to extract streams from.

If it’s a HTTP URL then “http://” can be omitted.

[STREAM]

Stream to play.

Use “best” or “worst” for highest or lowest quality available.

Fallback streams can be specified by using a comma-separated list:

"720p,480p,best"

If no stream is specified and --default-stream is not used then a list of available streams will be printed.

General options¶

-h, --help¶

Show this help message and exit.

-V, --version¶

Show version number and exit.

--plugins¶

Print a list of all currently installed plugins.

--can-handle-url URL¶

Check if Livestreamer has a plugin that can handle the specified URL.

Returns status code 1 for false and 0 for true.

Useful for external scripting.

--config FILENAME¶

Load options from this config file.

Can be repeated to load multiple files, in which case the options are merged on top of each other where the last config has highest priority.

-l LEVEL, --loglevel LEVEL¶

Set the log message threshold.

Valid levels are: none, error, warning, info, debug

-Q, --quiet¶

Hide all log output.

Alias for “--loglevel none”.

-j, --json¶

Output JSON representations instead of the normal text output.

Useful for external scripting.

--no-version-check¶

Do not check for new Livestreamer releases.

--version-check¶

Runs a version check and exits.

Player options¶

-p COMMAND, --player COMMAND¶

Player to feed stream data to. This is a shell-like syntax to support passing options to the player. For example:

"vlc --file-caching=5000"

To use a player that is located in a path with spaces you must quote the path:

"'/path/with spaces/vlc' --file-caching=5000"

By default VLC will be used if it can be found in its default location.

-a ARGUMENTS, --player-args ARGUMENTS¶

This option allows you to customize the default arguments which are put together with the value of --player to create a command to execute.

This value can contain formatting variables surrounded by curly braces, { and }. If you need to include a brace character, it can be escaped by doubling, e.g. {{ and }}.

Formatting variables available:

filename

This is the filename that the player will use. It’s usually “-” (stdin), but can also be a URL or a file depending on the options used.

It’s usually enough to use --player instead of this unless you need to add arguments after the filename.

Default is: “{filename}”.

-v, --verbose-player¶

Allow the player to display its console output.

-n, --player-fifo, --fifo¶

Make the player read the stream through a named pipe instead of the stdin pipe.

--player-http¶

Make the player read the stream through HTTP instead of the stdin pipe.

--player-continuous-http¶

Make the player read the stream through HTTP, but unlike --player-http it will continuously try to open the stream if the player requests it.

This makes it possible to handle stream disconnects if your player is capable of reconnecting to a HTTP stream. This is usually done by setting your player to a “repeat mode”.

--player-external-http¶

Serve stream data through HTTP without running any player. This is useful to allow external devices like smartphones or streaming boxes to watch streams they wouldn’t be able to otherwise.

Behavior will be similar to the continuous HTTP option, but no player program will be started, and the server will listen on all available connections instead of just in the local (loopback) interface.

The URLs that can be used to access the stream will be printed to the console, and the server can be interrupted using CTRL-C.

--player-external-http-port PORT¶

A fixed port to use for the external HTTP server if that mode is enabled. Omit or set to 0 to use a random high (>1024) port.

--player-passthrough TYPES¶

A comma-delimited list of stream types to pass to the player as a URL to let it handle the transport of the stream instead.

Stream types that can be converted into a playable URL are:

• hls
• http
• rtmp

Make sure your player can handle the stream type when using this.

--player-no-close¶

By default Livestreamer will close the player when the stream ends. This is to avoid “dead” GUI players lingering after a stream ends.

It does however have the side-effect of sometimes closing a player before it has played back all of its cached data.

This option will instead let the player decide when to exit.

File output options¶

-o FILENAME, --output FILENAME¶

Write stream data to FILENAME instead of playing it.

You will be prompted if the file already exists.

-f, --force¶

When using -o, always write to file even if it already exists.

-O, --stdout¶

Write stream data to stdout instead of playing it.

Stream options¶

--default-stream STREAM¶

Open this stream when no stream argument is specified, e.g. “best”.

--retry-streams DELAY¶

Will retry fetching streams until streams are found while waiting DELAY (seconds) between each attempt.

--retry-open ATTEMPTS¶

Will try ATTEMPTS times to open the stream until giving up.

Default is: 1.

--stream-types TYPES, --stream-priority TYPES¶

A comma-delimited list of stream types to allow.

The order will be used to separate streams when there are multiple streams with the same name but different stream types.

Default is: “rtmp,hls,hds,http,akamaihd”.

--stream-sorting-excludes STREAMS¶

Fine tune best/worst synonyms by excluding unwanted streams.

Uses a filter expression in the format:

[operator]<value>

Valid operators are >, >=, < and <=. If no operator is specified then equality is tested.

For example this will exclude streams ranked higher than “480p”:

">480p"

Multiple filters can be used by separating each expression with a comma.

For example this will exclude streams from two quality types:

">480p,>medium"

Stream transport options¶

--hds-live-edge SECONDS¶

The time live HDS streams will start from the edge of stream.

Default is: 10.0.

--hds-segment-attempts ATTEMPTS¶

How many attempts should be done to download each HDS segment before giving up.

Default is: 3.

--hds-segment-threads THREADS¶

The size of the thread pool used to download HDS segments. Minimum value is 1 and maximum is 10.

Default is: 1.

--hds-segment-timeout TIMEOUT¶

HDS segment connect and read timeout.

Default is: 10.0.

--hds-timeout TIMEOUT¶

Timeout for reading data from HDS streams.

Default is: 60.0.

--hls-live-edge SEGMENTS¶

How many segments from the end to start live HLS streams on.

The lower the value the lower latency from the source you will be, but also increases the chance of buffering.

Default is: 3.

--hls-segment-attempts ATTEMPTS¶

How many attempts should be done to download each HLS segment before giving up.

Default is: 3.

--hls-segment-threads THREADS¶

The size of the thread pool used to download HLS segments. Minimum value is 1 and maximum is 10.

Default is: 1.

--hls-segment-timeout TIMEOUT¶

HLS segment connect and read timeout.

Default is: 10.0.

--hls-timeout TIMEOUT¶

Timeout for reading data from HLS streams.

Default is: 60.0.

--http-stream-timeout TIMEOUT¶

Timeout for reading data from HTTP streams.

Default is: 60.0.

--ringbuffer-size SIZE¶

The maximum size of ringbuffer. Add a M or K suffix to specify mega or kilo bytes instead of bytes.

The ringbuffer is used as a temporary storage between the stream and the player. This is to allows us to download the stream faster than the player wants to read it.

The smaller the size, the higher chance of the player buffering if there are download speed dips and the higher size the more data we can use as a storage to catch up from speed dips.

It also allows you to temporary pause as long as the ringbuffer doesn’t get full since we continue to download the stream in the background.

Note

A smaller size is recommended on lower end systems (such as Raspberry Pi) when playing stream types that require some extra processing (such as HDS) to avoid unnecessary background processing.

Default is: “16M”.

--rtmp-proxy PROXY, --rtmpdump-proxy PROXY¶

A SOCKS proxy that RTMP streams will use.

Example: 127.0.0.1:9050

--rtmp-rtmpdump FILENAME, --rtmpdump FILENAME, -r FILENAME¶

RTMPDump is used to access RTMP streams. You can specify the location of the rtmpdump executable if it is not in your PATH.

Example: “/usr/local/bin/rtmpdump”

--rtmp-timeout TIMEOUT¶

Timeout for reading data from RTMP streams.

Default is: 60.0.

--stream-segment-attempts ATTEMPTS¶

How many attempts should be done to download each segment before giving up.

This is generic option used by streams not covered by other options, such as stream protocols specific to plugins, e.g. UStream.

Default is: 3.

--stream-segment-threads THREADS¶

The size of the thread pool used to download segments. Minimum value is 1 and maximum is 10.

This is generic option used by streams not covered by other options, such as stream protocols specific to plugins, e.g. UStream.

Default is: 1.

--stream-segment-timeout TIMEOUT¶

Segment connect and read timeout.

This is generic option used by streams not covered by other options, such as stream protocols specific to plugins, e.g. UStream.

Default is: 10.0.

--stream-timeout TIMEOUT¶

Timeout for reading data from streams.

This is generic option used by streams not covered by other options, such as stream protocols specific to plugins, e.g. UStream.

Default is: 60.0.

--stream-url¶

If possible, translate the stream to a URL and print it.

--subprocess-cmdline, --cmdline, -c¶

Print command-line used internally to play stream.

This is only available on RTMP streams.

--subprocess-errorlog, --errorlog, -e¶

Log possible errors from internal subprocesses to a temporary file. The file will be saved in your systems temporary directory.

Useful when debugging rtmpdump related issues.

HTTP options¶

--http-proxy HTTP_PROXY¶

A HTTP proxy to use for all HTTP requests.

Example: http://hostname:port/

--https-proxy HTTPS_PROXY¶

A HTTPS capable proxy to use for all HTTPS requests.

Example: http://hostname:port/

--http-cookie KEY=VALUE¶

A cookie to add to each HTTP request.

Can be repeated to add multiple cookies.

--http-header KEY=VALUE¶

A header to add to each HTTP request.

Can be repeated to add multiple headers.

--http-query-param KEY=VALUE¶

A query parameter to add to each HTTP request.

Can be repeated to add multiple query parameters.

--http-ignore-env¶

Ignore HTTP settings set in the environment such as environment variables (HTTP_PROXY, etc) or ~/.netrc authentication.

--http-no-ssl-verify¶

Don’t attempt to verify SSL certificates.

Usually a bad idea, only use this if you know what you’re doing.

--http-ssl-cert FILENAME¶

SSL certificate to use.

Expects a .pem file.

--http-ssl-cert-crt-key CRT_FILENAME KEY_FILENAME¶

SSL certificate to use.

Expects a .crt and a .key file.

--http-timeout TIMEOUT¶

General timeout used by all HTTP requests except the ones covered by other options.

Default is: 20.0.

Plugin options¶

--plugin-dirs DIRECTORY¶

Attempts to load plugins from these directories.

Multiple directories can be used by separating them with a semi-colon.

--twitch-oauth-token TOKEN¶

An OAuth token to use for Twitch authentication. Use --twitch-oauth-authenticate to create a token.

--twitch-oauth-authenticate¶

Open a web browser where you can grant Livestreamer access to your Twitch account which creates a token for use with --twitch-oauth-token.

--twitch-cookie COOKIES¶

Twitch cookies to authenticate to allow access to subscription channels.

Example:

"_twitch_session_id=xxxxxx; persistent=xxxxx"

Note

This method is the old and clunky way of authenticating with Twitch, using --twitch-oauth-authenticate is the recommended and simpler way of doing it now.

--ustream-password PASSWORD¶

A password to access password protected UStream.tv channels.

--crunchyroll-username USERNAME¶

A Crunchyroll username to allow access to restricted streams.

--crunchyroll-password [PASSWORD]¶

A Crunchyroll password for use with --crunchyroll-username.

If left blank you will be prompted.

--crunchyroll-purge-credentials¶

Purge cached Crunchyroll credentials to initiate a new session and reauthenticate.

--crunchyroll-locale LOCALE¶

Indicate which locale to use for Crunchyroll subtitles.

The locale is formatted as [language_code]_[country_code], by default en_US is used.

--livestation-email EMAIL¶

A Livestation account email to access restricted or premium quality streams.

--livestation-password PASSWORD¶

A Livestation account password to use with --livestation-email.

MPC-HC reports “File not found”¶

Upgrading to version 1.7 or newer will solve this issue since reading data from standard input is not supported in version 1.6.x of MPC-HC.

MPC-HC only plays sound on Twitch streams¶

Twitch sometimes returns badly muxed streams which may confuse players. The following workaround was contributed by MPC-HC developer @kasper93:

To fix this problem go to options -> internal filters -> open splitter settings and increase “Stream Analysis Duration” this will let ffmpeg to properly detect all streams.

Using --player-passthrough hls has also been reported to work.

MPlayer tries to play Twitch streams at the wrong FPS¶

This is a bug in MPlayer, using the MPlayer fork mpv instead is recommended.

VLC hangs when buffering and no playback starts¶

Some versions of 64-bit VLC seem to be unable to read the stream created by rtmpdump. Using the 32-bit version of VLC might help.

Enable caching in your player¶

By default most players do not cache the data they receive from Livestreamer. Caching can reduce the amount of buffering you run into because the player will have some breathing room between receiving the data and playing it.

Use the --player option to pass these options to your player.

Multi-threaded streaming¶

On segmented streaming protocols (such as HLS and HDS) it’s possible to use multiple threads to potentially increase the throughput. Each stream type has it’s own option, these options are currently available:

Simple player¶

This example uses the PyGObject module to playback a stream using the GStreamer framework.

#!/usr/bin/env python

from __future__ import print_function

import sys

import gi

from gi.repository import GObject as gobject, Gst as gst
from livestreamer import Livestreamer, StreamError, PluginError, NoPluginError


def exit(msg):
    print(msg, file=sys.stderr)
    sys.exit()


class LivestreamerPlayer(object):
    def __init__(self):
        self.fd = None
        self.mainloop = gobject.MainLoop()

        # This creates a playbin pipeline and using the appsrc source
        # we can feed it our stream data
        self.pipeline = gst.ElementFactory.make("playbin", None)
        self.pipeline.set_property("uri", "appsrc://")

        # When the playbin creates the appsrc source it will call
        # this callback and allow us to configure it
        self.pipeline.connect("source-setup", self.on_source_setup)

        # Creates a bus and set callbacks to receive errors
        self.bus = self.pipeline.get_bus()
        self.bus.add_signal_watch()
        self.bus.connect("message::eos", self.on_eos)
        self.bus.connect("message::error", self.on_error)

    def exit(self, msg):
        self.stop()
        exit(msg)

    def stop(self):
        # Stop playback and exit mainloop
        self.pipeline.set_state(gst.State.NULL)
        self.mainloop.quit()

        # Close the stream
        if self.fd:
            self.fd.close()

    def play(self, stream):
        # Attempt to open the stream
        try:
            self.fd = stream.open()
        except StreamError as err:
            self.exit("Failed to open stream: {0}".format(err))

        # Start playback
        self.pipeline.set_state(gst.State.PLAYING)
        self.mainloop.run()

    def on_source_setup(self, element, source):
        # When this callback is called the appsrc expects
        # us to feed it more data
        source.connect("need-data", self.on_source_need_data)

    def on_source_need_data(self, source, length):
        # Attempt to read data from the stream
        try:
            data = self.fd.read(length)
        except IOError as err:
            self.exit("Failed to read data from stream: {0}".format(err))

        # If data is empty it's the end of stream
        if not data:
            source.emit("end-of-stream")
            return

        # Convert the Python bytes into a GStreamer Buffer
        # and then push it to the appsrc
        buf = gst.Buffer.new_wrapped(data)
        source.emit("push-buffer", buf)

    def on_eos(self, bus, msg):
        # Stop playback on end of stream
        self.stop()

    def on_error(self, bus, msg):
        # Print error message and exit on error
        error = msg.parse_error()[1]
        self.exit(error)


def main():
    if len(sys.argv) < 3:
        exit("Usage: {0} <url> <quality>".format(sys.argv[0]))

    # Initialize and check GStreamer version
    gi.require_version("Gst", "1.0")
    gobject.threads_init()
    gst.init(None)

    # Collect arguments
    url = sys.argv[1]
    quality = sys.argv[2]

    # Create the Livestreamer session
    livestreamer = Livestreamer()

    # Enable logging
    livestreamer.set_loglevel("info")
    livestreamer.set_logoutput(sys.stdout)

    # Attempt to fetch streams
    try:
        streams = livestreamer.streams(url)
    except NoPluginError:
        exit("Livestreamer is unable to handle the URL '{0}'".format(url))
    except PluginError as err:
        exit("Plugin error: {0}".format(err))

    if not streams:
        exit("No streams found on URL '{0}'".format(url))

    # Look for specified stream
    if quality not in streams:
        exit("Unable to find '{0}' stream on URL '{1}'".format(quality, url))

    # We found the stream
    stream = streams[quality]

    # Create the player and start playback
    player = LivestreamerPlayer()

    # Blocks until playback is done
    player.play(stream)

if __name__ == "__main__":
    main()

Stream subclasses¶

You are able to inspect the parameters used by each stream, different properties are available depending on stream type.

class livestreamer.stream.AkamaiHDStream(session, url, swf=None, seek=None)¶

Implements the AkamaiHD Adaptive Streaming protocol

Attributes:

• url URL to the stream
• swf URL to a SWF used by the handshake protocol
• seek Position to seek to when opening the stream

class livestreamer.stream.HDSStream(session, baseurl, url, bootstrap, metadata=None, timeout=60, **request_params)¶

Implements the Adobe HTTP Dynamic Streaming protocol

Attributes:

• baseurl Base URL
• url Base path of the stream, joined with the base URL when fetching fragments
• bootstrap Either a URL pointing to the bootstrap or a bootstrap Box object used for initial information about the stream
• metadata Either None or a ScriptData object that contains metadata about the stream, such as height, width and bitrate

classmethod parse_manifest(session, url, timeout=60, pvswf=None, **request_params)¶

Parses a HDS manifest and returns its substreams.

Parameters:	url – The URL to the manifest. timeout – How long to wait for data to be returned from from the stream before raising an error. pvswf – URL of player SWF for Akamai HD player verification.

class livestreamer.stream.HLSStream(session_, url, **args)¶

Implementation of the Apple HTTP Live Streaming protocol

Attributes:

• url The URL to the HLS playlist.
• args A dict containing keyword arguments passed to requests.request(), such as headers and cookies.

Changed in version 1.7.0: Added args attribute.

classmethod parse_variant_playlist(session_, url, name_key='name', name_prefix='', check_streams=False, **request_params)¶

Attempts to parse a variant playlist and return its streams.

Parameters:	url – The URL of the variant playlist. name_key – Prefer to use this key as stream name, valid keys are: name, pixels, bitrate. name_prefix – Add this prefix to the stream names. check_streams – Only allow streams that are accesible.

class livestreamer.stream.HTTPStream(session_, url, buffered=True, **args)¶

A HTTP stream using the requests library.

Attributes:

• url The URL to the stream, prepared by requests.
• args A dict containing keyword arguments passed to requests.request(), such as headers and cookies.

class livestreamer.stream.RTMPStream(session, params, redirect=False)¶

RTMP stream using rtmpdump.

Attributes:

• params A dict containing parameters passed to rtmpdump