Snapcast (Synchronous audio player) plays audio streams time synchronized on multiple devices over network using a server and a client component. The server picks up an audio stream from a fifo pipe, thus it can be combined with any audio source that is able to write to a pipe. The Snapserver is able to handle multiple streams at once. The assignment of clients to streams and their volume is controlled via a JSON-API.
USE flags for media-sound/snapcast Synchronous multi-room audio player
|Build and install Snapcast client component
|Enable the use of dev-libs/expat for XML parsing
|Add support for FLAC: Free Lossless Audio Codec
|Enable Opus audio codec support
|Build and install Snapcast server component
|Build with TREMOR version of vorbis
|Add support for the OggVorbis audio codec
|Support for DNS Service Discovery (DNS-SD)
emerge --ask media-sound/snapcast
The Snapcast server should run as its own user:group e.g. snapserver:snapserver (default created during emerge).
A sample config that makes snapserver pick up an audio stream from the named pipe /tmp/snapfifo looks like this:
SNAPSERVER_OPTS="-d -s pipe:///tmp/snapfifo?name=default"
To start snapserver now run:
rc-service snapserver start
To start snapserver at boot time run:
rc-update add snapserver default
For all snapserver options please see man snapserver.
The basic configuration for the client looks like this:
This will run snapclient as daemon using the default sound card. It will try to find servers on the network using Avahi, if avahi-daemon is installed and running.
To start snapclient now run:
rc-service snapclient start
To start snapclient at boot time run:
rc-update add snapclient default
Desktop environment/selecting sound card
To make use of PulseAudio/use snapclient in your desktop environment (to be able to use snapclient alongside with other audio sources, control its volume, etc.) you would start it as your desktop user instead of letting the init system control it.
To get a list of sound cards, and select the right one, run:
Discard all samples (playback) or generate zero samples (capture)
PulseAudio Sound Server
To use PulseAudio for example, use
1 with the
-s command line option:
snapclient -s 1
Snapclient should now appear in your list of audio playing applications of PulseAudio.
To make some testing noise, random bytes can be shoved into the snapservers pipe:
cat /dev/urandom > /tmp/snapfifo
Snapcast can basically be used with anything that is able to write PCM audio to a fifo.
To make MPlayer play something over Snapcast use:
mplayer <input> -novideo -channels 2 -srate 48000 -af format=s16le -ao pcm:file=/tmp/snapfifo
To hear music with MPD over Snapcast create a new audio_output in the mpd.conf using the fifo module:
The sample rate setting is the default one used by Snapcast. Different sample rates can be used but must be set in the snapserver config file first.
Multiple input streams
Snapcast is able to handle multiple input streams per server instance. The server also assigns the clients dynamically to streams.
Each stream is added with another
-s option and its own fifo to the snapserver config.
SNAPSERVER_OPTS="-d -s pipe:///tmp/snap_kitchen?name=Kitchen -s pipe:///tmp/snap_livingroom?name=Livingroom"
Assignment of streams to clients as well as volume, name, and latency of each client is controlled over a JSON-API provided by snapserver. The API documentation can be found on GitHub.
There are some (GUI-) implementations using that API, like
The developer of Snapcast also made an Android app, Snapdroid.
It can connect clients to streams, set volumes, etc. Moreover it implements a Snapcast client on Android, making it possible to play a Snapcast stream on an Android device.