/pulseaudio-dlna

A lightweight streaming server which brings DLNA / UPNP and Chromecast support to PulseAudio and Linux

Primary LanguagePythonGNU General Public License v3.0GPL-3.0

About

This is _pulseaudio-dlna_. A lightweight streaming server which brings DLNA / UPNP and Chromecast support to PulseAudio and Linux. It can stream your current PulseAudio playback to different UPNP devices (UPNP Media Renderers) or Chromecasts in your network. It's main goals are: easy to use, no configuration hassle, no big dependencies.

UPNP renderers in your network will show up as pulseaudio sinks.

Image of pulseaudio-dlna

License

pulseaudio-dlna is licensed under GPLv3.

pulseaudio-dlna is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

pulseaudio-dlna is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with pulseaudio-dlna.  If not, see <http://www.gnu.org/licenses/>.

Donation

Image of pulseaudio-dlna If I could help you or if you like my work, you can buy me a coffee, a beer or pizza.

Changelog

  • 0.4.7 - (2015-11-18)

    • The application can now co-exist with other applications which are using the port 1900/udp (thanks to klaernie)
    • Fixed the daemon mode to support psutil 1.x and 2.x (thanks to klaernie)
    • HTML entities in device descriptions are now converted automatically
    • Faster and more reliable device discovery (new dependency python-netifaces)
    • Added the --cover-mode option, one mode requires (optional) dependencies gtk, cairo, rsvg
    • L16 codecs are now selected better (e.g. needed for XBox 360)
    • Fixed a bug where sometimes it was tried to remove sinks twice on cleanup
    • Added the --update-device-config flag
    • Added the --ssdp-ttl, --ssdp-mx, --ssdp-amount options
    • Added the --msearch-port option
  • 0.4.6 - (2015-10-17)

    • Added support for Google Chromecast Audio (thanks to leonhandreke)
    • Fixed a bug where devices which does not specifiy control urls made the application crash
    • Added the --disable-device-stop flag
    • Added the --request-timeout option
    • You can now also add rules to renderers (e.g. DISABLE_DEVICE_STOP, REQUEST_TIMEOUT)
    • Fixed a bug where stream urls where not parsed correctly
    • Fixed a bug which made a Chomecast Audio throwing exceptions while stopping
    • Fixed a bug where the system's default encoding could not be determined when piping the applications output
  • 0.4.5.2 - (2015-09-21)

    • Fixed a bug where the encoding of SSDP headers was not detected correctly (new dependency: python-chardet)
  • 0.4.5.1 - (2015-09-20)

    • Added a missing dependency python-concurrent.futures (thanks to Takkat-Nebuk)
  • 0.4.5 - (2015-09-20)

    • Exceptions while updating sink and device information from pulseaudio are now handled better
    • Changed --fake-http10-content-length flag to --fake-http-content-length to also support HTTP 1.1 requests
    • Fixed a bug where the supported device mime types could not get parsed correctly
    • Fixed a bug where the device UUID was not parsed correctly
    • Fixed a bug where just mime types beginning with audio/ where accepted, but not e.g. application/ogg
    • The stream server will now respond with 206 when receiving requests with range header
    • UPNP control commands have now a timeout of 10 seconds
    • Fixed a bug where the wrong stream was removed from the stream manager
    • Fixed several bugs caused by purely relying on stopping actions for the devices idle state
    • Added L16 Encoder
    • The encoder option can now handle multiple options separated by comma
    • Added the --create-device-config flag
    • Fixed a bug where the dbus session was bound from the wrong process
    • Fix a bug where the wrong device UDN was retrieved from XML documents containing multiple devices
  • 0.4.4 - (2015-08-07)

    • Added --disable-ssdp-listener option
    • Fixed a bug with applications which remove and re-add streams all the time
    • Added a missing dependency python-psutil
  • 0.4.3 - (2015-08-02)

    • Fixed a bug when trying to terminate an encoder process
    • Catch exceptions when trying to update pulseaudio sinks
    • Fixed a timing issue where the streamserver was not ready but devices were already instructed to play
  • 0.4.2 - (2015-08-02)

    • The mp3 encoder is now prioritize over wav
    • Added --disable-switchback option
    • Wav encoders do not longer share their encoder process
  • 0.4.1 - (2015-07-27)

    • Fixed Makefile for launchpad
  • 0.4.0 - (2015-07-27)

    • Added the --fake-http10-content-length option
    • The application can now run as root
    • Catch pulseaudio exceptions for streams, sinks and modules when those are gone
    • Fixed a bug where a missing ssdp header field made the application crash
    • New devices are added now during runtime (thanks to coder-hugo)
    • Rewrite of the streaming server
    • Upnp devices can now request their audio format based on their capabilities
    • Added AAC encoder
    • If a device stops playing, the streams currently playing on the corresponding sink are switched back to the default sink
    • If a device failes to start playing, streams currently playing on the corresponding sink are switched back to the default sink
    • Added Chromecast support (new dependency: python-protobuf)
    • Fixed a bug where the application crashed when there was no suitable encoder found
    • Added the --bit-rate option
    • Added additional headers for DLNA devices
    • Added switch back mode also for sinks, not just for streams (new dependency: python-notify2)
    • Added better logging
    • Validate encoders for installed dependencies
  • 0.3.5 - (2015-04-09)

    • Fixed a bug where Sonos description XML could not get parsed correctly
  • 0.3.4 - (2015-03-22)

    • Fixed Makefile for launchpad
  • 0.3.3 - (2015-03-22)

    • Added the --filter-device option
    • Send 2 SSDP packets by default for better UPNP device discovery
    • Added virtualenv for local installation
  • 0.3.2 - (2015-03-14)

    • Added the Opus Encoder (new dependency: opus-tools) (thanks to MobiusHorizons)
    • Fixed a bug where an empty UPNP device name made the application crash
    • Added a missing dependency (python-gobject)
  • 0.3.1 - (2015-02-13)

  • 0.3.0 - (2015-02-01)

    • Added debian packaging
    • Added proper signal handlers (new dependency: python-setproctitle)
    • Fixed a bug where binding to an already used port made the application crash
    • HTTP charset encoding is now specified correctly
  • 0.2.4 - (2015-01-25)

    • Stream changes are now handled correctly (thanks to Takkat-Nebuk)
  • 0.2.3 - (2015-01-21)

    • Fixed a timing bug where the pulseaudio module was not loaded fast enough (thanks to Takkat-Nebuk)
  • 0.2.2 - (2015-01-18)

    • Fixed encoding issues
    • Try to load the DBus module if it is not loaded before (thanks to Takkat-Nebuk)
  • 0.2.1 - (2015-01-11)

    • TTL changed to 10 and timeout to 5 for UDP broadcasting
    • Added the --renderer-urls option to manually add UPNP devices via their control url
    • Added the --debug flag
    • The host ip address is now discovered automatically, no need to specifiy --host anymore

Installation via PPA

Supported Ubuntu releases:

  • 15.10 (Wily Werewolf)
  • 15.04 (Vivid Vervet)
  • 14.04.2 LTS (Trusty Tahr)

Ubuntu users can install pulseaudio-dlna via the following repository.

sudo apt-add-repository ppa:qos/pulseaudio-dlna
sudo apt-get update
sudo apt-get install pulseaudio-dlna

Starting

After that you can start pulseaudio-dlna via:

pulseaudio-dlna

Head over the the using section for further instructions.

Installation for other distributions

Some community members are providing packages for others distributions. Keep in mind that since i am not using those, i can hardly support them!

Installation via git

Other linux users can clone this git repository, make sure you have all the dependencies installed and the PulseAudio DBus module is loaded.

Basic requirements

These are the requirements pulseaudio-dlna acutally needs to run. These dependencies will get installed if you install it via the PPA.

  • python2.7
  • python-pip
  • python-setuptools
  • python-dbus
  • python-beautifulsoup
  • python-docopt
  • python-requests
  • python-setproctitle
  • python-gobject
  • python-protobuf
  • python-notify2
  • python-psutil
  • python-concurrent.futures
  • python-chardet
  • python-netifaces
  • vorbis-tools
  • sox
  • lame
  • flac
  • faac
  • opus-tools

You can install all the dependencies in Ubuntu via:

sudo apt-get install python2.7 python-pip python-setuptools python-dbus python-beautifulsoup python-docopt python-requests python-setproctitle python-gobject python-protobuf python-notify2 python-psutil python-concurrent.futures python-chardet python-netifaces vorbis-tools sox lame flac faac opus-tools

PulseAudio DBus module

Since version 0.2.2 the DBus module should be loaded automatically, if it was not loaded before. It that does not work, you can load the DBus module in Ubuntu via the following command. Note that you have to do this every time you restart PulseAudio (or your computer).

pacmd load-module module-dbus-protocol

Or to make changes persistant edit the file /etc/pulse/default.pa with your favorite editor and append the following line:

load-module module-dbus-protocol

Install it local

The recommend method of using pulseaudio-dlna is to install it local to a python virtualenv. In that way you will keep your system clean. If you don't like it anymore, just delete the folder. For that method you need some additional dependencies.

virtualenv requirements

  • python-virtualenv (Ubuntu <= 14.04 Trusty LTS)
  • virtualenv (Ubuntu >= 14.10 Utopic)
  • python-dev

So all Ubuntu versions prior to 14.10 Utopic need to install:

sudo apt-get install python-virtualenv python-dev

All Ubuntu versions above install:

sudo apt-get install virtualenv python-dev

Installing & starting

Change to the project root folder and start the installation via:

make

After that you can start pulseaudio-dlna via:

bin/pulseaudio-dlna

Install it to your system

Since some people like it more to install software globally, you can do that too. In many software projects this is the default installation method.

Installing & starting

Change to the root folder and start the installation via:

make install

After that you can start pulseaudio-dlna via:

pulseaudio-dlna

Using

pulseaudio-dlna should detect the ip address your computer is reachable within your local area network. If the detected ip address is not correct or there were no ips found, you still can specifiy them yourself via the host option (--host <your-ip>)

Right after startup it should start searching for UPNP devices in your LAN and add new PulseAudio sinks. After 5 seconds the progress is complete and you can select your UPNP renderers from the default audio control.

In case you just want to stream single audio streams to your UPNP devices you can do this via pavucontrol.

You can install pavucontrol in Ubuntu via the following command:

sudo apt-get install pavucontrol

Note that pulseaudio-dlna has to run all the time while you are listening to your music. If you stop pulseaudio-dlna it will cleanly remove the created UPNP devices from PulseAudio and your UPNP devices will stop playing.

Since 0.4, new devices are automatically discovered as they appear on the network.

CLI

Usage:
    pulseaudio-dlna [--host <host>] [--port <port>][--encoder <encoders>] [--bit-rate=<rate>]
                    [--filter-device=<filter-device>]
                    [--renderer-urls <urls>]
                    [--request-timeout <timeout>]
                    [--msearch-port=<msearch-port>] [--ssdp-mx <ssdp-mx>] [--ssdp-ttl <ssdp-ttl>] [--ssdp-amount <ssdp-amount>]
                    [--cover-mode <mode>]
                    [--debug]
                    [--fake-http10-content-length] [--fake-http-content-length]
                    [--disable-switchback] [--disable-ssdp-listener] [--disable-device-stop]
    pulseaudio-dlna [--host <host>] [--create-device-config] [--update-device-config]
                    [--msearch-port=<msearch-port>] [--ssdp-mx <ssdp-mx>] [--ssdp-ttl <ssdp-ttl>] [--ssdp-amount <ssdp-amount>]
    pulseaudio-dlna [-h | --help | --version]

Options:
    --create-device-config                 Discovers all devices in your network and write a config for them.
                                           That config can be editied manually to adjust various settings.
                                           You can set:
                                             - Device name
                                             - Codec order (The first one is used if the encoder binary is available on your system)
                                             - Various codec settings such as the mime type, specific rules or
                                               the bit rate (depends on the codec)
                                           A written config is loaded by default if the --encoder and --bit-rate options are not used.
    --update-device-config                 Same as --create-device-config but preserves your existing config from being overwritten
       --host=<host>                       Set the server ip.
    -p --port=<port>                       Set the server port [default: 8080].
    -e --encoder=<encoders>                Set the audio encoder.
                                           Possible encoders are:
                                             - mp3   MPEG Audio Layer III (MP3)
                                             - ogg   Ogg Vorbis (OGG)
                                             - flac  Free Lossless Audio Codec (FLAC)
                                             - wav   Waveform Audio File Format (WAV)
                                             - opus  Opus Interactive Audio Codec (OPUS)
                                             - aac   Advanced Audio Coding (AAC)
                                             - l16   Linear PCM (L16)
    -b --bit-rate=<rate>                   Set the audio encoder's bitrate.
    --filter-device=<filter-device>        Set a name filter for devices which should be added.
                                           Devices which get discovered, but won't match the
                                           filter text will be skipped.
    --renderer-urls=<urls>                 Set the renderer urls yourself. no discovery will commence.
    --request-timeout=<timeout>            Set the timeout for requests in seconds [default: 10].
    --ssdp-ttl=<ssdp-ttl>                  Set the SSDP socket's TTL [default: 10].
    --ssdp-mx=<ssdp-mx>                    Set the MX value of the SSDP discovery message [default: 3].
    --ssdp-amount=<ssdp-amount>            Set the amount of SSDP discovery messages being sent [default: 5].
    --msearch-port=<msearch-port>          Set the source port of the MSEARCH socket [default: random].
    --cover-mode=<mode>                    Set the cover mode [default: default].
                                           Possible modes are:
                                             - disabled       No icon is shown
                                             - default        The application icon is shown
                                             - distribution   The icon of your distribution is shown
                                             - application    The audio application's icon is shown
    --debug                                enables detailed debug messages.
    --fake-http-content-length             If set, the content-length of HTTP requests will be set to 100 GB.
    --disable-switchback                   If set, streams won't switched back to the default sink if a device disconnects.
    --disable-ssdp-listener                If set, the application won't bind to the port 1900 and therefore the automatic discovery of new devices won't work.
    --disable-device-stop                  If set, the application won't send any stop commands to renderers at all
    -v --version                           Show the version.
    -h --help                              Show the help.

Samples:

  • pulseaudio-dlna will start pulseaudio-dlna on port 8080 and stream your PulseAudio streams encoded with mp3.
  • pulseaudio-dlna --encoder ogg will start pulseaudio-dlna on port 8080 and stream your PulseAudio streams encoded with Ogg Vorbis.
  • pulseaudio-dlna --port 10291 --encoder flac will start pulseaudio-dlna on port 10291 and stream your PulseAudio streams encoded with FLAC.
  • pulseaudio-dlna --filter-device 'Nexus 5,TV' will just use devices named Nexus 5 or TV even when more devices got discovered.
  • pulseaudio-dlna --renderer-urls http://192.168.1.7:7676/smp_10_ won't discover upnp devices by itself. Instead it will search for upnp renderers at the specified locations. You can specify multiple locations via urls seperated by comma (,). Most users won't ever need this option, but since UDP multicast packages won't work (most times) over VPN connections this is very useful if you ever plan to stream to a UPNP device over VPN.

Device configuration rules

Most times the automatic discovery of supported device codecs and their prioritization works pretty good. But in the case of a device which does work out of the box or if you don't like the used codec you can adjust the settings with a device configuration.

If you want to create a specific configuration for your devices you can do that via the --create-device-config flag. It will search for devices on your network and write a config for them. It will look for / write them at:

  • ~/.local/share/pulseaudio-dlna/devices.json (prioritized)
  • /etc/pulseaudio-dlna/devices.json

The purpose of this is that the application should do the most work for the user. You just have to edit the file instead of writing it completely on your own.

Let's make an example: I started the application via pulseaudio-dlna --create-device-config and that is what was discovered:

    "uuid:e4572d54-c2c7-d491-1eb3-9cf17cf5fe01": {
        "rules": [], 
        "flavour": "DLNA", 
        "name": "Device name", 
        "codecs": [
            {
                "rules": [], 
                "bit_rate": null, 
                "identifier": "mp3", 
                "mime_type": "audio/mpeg"
            }, 
            {
                "rules": [], 
                "identifier": "flac", 
                "mime_type": "audio/flac"
            }, 
            {
                "channels": 2, 
                "rules": [], 
                "identifier": "l16", 
                "sample_rate": 48000, 
                "mime_type": "audio/L16;rate=48000;channels=2"
            }, 
            {
                "channels": 2, 
                "rules": [], 
                "identifier": "l16", 
                "sample_rate": 44100, 
                "mime_type": "audio/L16;rate=44100;channels=2"
            }, 
            {
                "channels": 1, 
                "rules": [], 
                "identifier": "l16", 
                "sample_rate": 44100, 
                "mime_type": "audio/L16;rate=44100;channels=1"
            }
        ]
    }

It was detected that the device supports the following codecs:

  • audio/mp3
  • audio/flac
  • audio/L16;rate=48000;channels=2
  • audio/L16;rate=44100;channels=2
  • audio/L16;rate=44100;channels=1

If you don't change the configuration at all, it means that the next time you start pulseaudio-dlna it will automatically use those codecs for that device. The order of the list also defines the priority. It will take the first codec and use it if the appropriate encoder binary is installed on your system. If the binary is missing it will take the next one. So here the mp3 codec would be used, if the lame binary is installed.

You can also change the name of the device, adjust the mime type or set the bit rate. A null value means default, for bit rates this is set to 192 Kbit/s.

In that case I want to rename my device to "Living Room". Besides that I don't want the L16 codecs, so i simply remove them and i want my mp3 to be encoded in 256 Kbit/s.

    "uuid:e4572d54-c2c7-d491-1eb3-9cf17cf5fe01": {
        "rules": [], 
        "flavour": "DLNA", 
        "name": "Living Room", 
        "codecs": [
            {
                "rules": [], 
                "bit_rate": 256, 
                "identifier": "mp3", 
                "mime_type": "audio/mpeg"
            }, 
            {
                "rules": [], 
                "identifier": "flac", 
                "mime_type": "audio/flac"
            }
        ]
    }

But as it turns out this device has a problem with playing the mp3 stream when you don't specify the --fake-http-content-length flag. Let's say flac works without the flag. So, you can add a rule for that to that device.

    "uuid:e4572d54-c2c7-d491-1eb3-9cf17cf5fe01": {
        "rules": [], 
        "flavour": "DLNA", 
        "name": "Living Room", 
        "codecs": [
            {
                "rules": [
                    {
                        "name": "FAKE_HTTP_CONTENT_LENGTH"
                    }
                ], 
                "bit_rate": 256, 
                "identifier": "mp3", 
                "mime_type": "audio/mpeg"
            }, 
            {
                "rules": [], 
                "identifier": "flac", 
                "mime_type": "audio/flac"
            }
        ]
    }

That's it. pulseaudio-dlna will automatically use that config if you don't use the --encoder or --bit-rate options.

Known Issues

  • Distorted sound

    If you experience distorted sound, try to pause / unpause the playback or changing / adjusting the volume. Some encoders handle volume changes better than others. The lame encoder handles this by far better than most of the other ones.

  • There is a delay about a few seconds

    Since there is HTTP streaming used for the audio data to transport, there is always a buffer involved. This device buffer ensures that even if you suffer from a slow network (e.g. weak wifi) small interruptions won't affect your playback. On the other hand devices will first start to play when this buffer is filled. Most devices do this based on the received amount of data. Therefore inefficient codecs such as wav fill that buffer much faster than efficient codecs do. The result is a noticeable shorter delay in contrast to e.g. mp3 or others. Note, that in this case your network should be pretty stable, otherwise your device will quickly run out of data and stop playing. This is normally not a problem with cable connections. E.g. I have a delay about 1-2 seconds with wav and a delay of about 5 seconds with mp3 with the same cable connected device. You can decrease the delay when using wav or using high bit rates, but you won't get rid of it completely. My advice: If you have a reliable network, use wav. It is lossless and you will get a short delay. If you have not, use another encoder which does not require that much bandwidth to make sure your device will keep playing. Of course you will be affected from a higher delay.

Troubleshooting

  • My device does not get discovered by pulseaudio-dlna

    The computer pulseaudio-dlna is running on and your device needs to be in the same network. In uncomplicated home LANs this is normally the case. You can test if other applications are able to find your device, e.g. BubbleUPnP (Android application). If they do it is likely that you are using a firewall / iptables. Try disabling it. If you verified that your firewall is blocking, you should use the --msearch-port <port> option and open port 8080/tcp, port 1900/udp and port <port>/udp.

  • The device is successfully instructed to play, but the device never connects to pulseaudio-dlna

    Check if your are using a firewall / iptables. If that works, open port 8080/tcp and port 1900/udp.

  • The device is successfully instructed to play, but the device immediately disconnects after some seconds

    Some devices do not stick to the HTTP 1.0/1.1 standard. Since most devices do, pulseaudio-dlna must be instructed by CLI flags to act in a non-standard way.

    • --fake-http-content-length

      Adds a faked HTTP Content-Length to HTTP 1.0/1.1 responses. The length is set to 100 GB and ensures that the device would keep playing for months. This is e.g. necessary for the Hame Soundrouter and depending on the used encoder for Sonos devices.

Tested devices

A listed entry means that it was successfully tested, even if there is no specific codec information available.

Device mp3 wav ogg flac aac opus l16
D-Link DCH-M225/E 🚫 🚫 🚫
Cocy UPNP media renderer 🚫 🚫 🚫 🚫 🚫
gmrender-resurrect
BubbleUPnP (Android App) 🚫
Samsung Smart TV LED60 (UE60F6300)
Samsung Smart TV LED40 (UE40ES6100)
Samsung Smart TV LED48 (UE48JU6560) 2 🚫 🚫
Xbmc / Kodi 🚫 2 2
Philips Streamium NP2500 Network Player
Yamaha RX-475 (AV Receiver)
Yamaha RX-V573 (AV Receiver) 6
Majik DSM
Pi MusicBox
Google Chromecast 🚫 🚫 🚫
Google Chromecast Audio 🚫 🚫 🚫
Sonos PLAY:1 3 3 🚫 🚫
Sonos PLAY:3 3 3 🚫 🚫
Hame Soundrouter 1 🚫 🚫 1 🚫 🚫 🚫
Raumfeld Speaker M
Pioneer VSX-824 (AV Receiver)
ROCKI
Sony STR-DN1050 (AV Receiver)
Pure Jongo S3
Volumio
Logitech Media Server
Panasonic TX-50CX680W
Yamaha CRX-N560D 4
RaidSonic IB-MP401Air
Medion P85055
Onkyo TX-NR509 🚫
Denon AVR-3808
DAMAI Airmusic
Panasonic TX-50CX680W
Xbox 360 5 🚫 🚫 🚫 🚫

1) Works when specifing the --fake-http-content-length flag

2) Is capable of playing the codec, but does not specifiy the correct mime type

3) Works since 0.4.5 (--fake-http-content-length is added automatic)

4) The device needs to be in SERVER mode to accept instructions

5) Was reported to buffer really long. Approximately 45 seconds

6) Was reported to have issues being discovered. Make sure you run the latest firmware

Supported encoders

Encoder Description Identifier
lame MPEG Audio Layer III mp3
oggenc Ogg Vorbis ogg
flac Free Lossless Audio Codec flac
sox Waveform Audio File Format wav
opusenc Opus Interactive Audio Codec opus
faac Advanced Audio Coding aac
sox Linear PCM l16

You can select a specific codec using the --encoder flag followed by its identifier.