I had a specific need where I wanted to send sound from my DJ equipment in my living room and to my home stereo as low latency as possible over the WIFI.

I could have stretched a normal signal cable between the DJ equipment and the 4-5 meters, but that would not be any fun would it. ( also my other half don’t like cables that much ).

While figuring out how to do this, my colleague did show me a blog post he wrote about snapcast.

I had two raspberry pi’s laying around for this project, one for the DJ corner I have and one to use as a snapcast client for my stereo.

SnapCast attempt

To get the max out of the audio quality before I started testing out snapcast I ordered two of the HiFiBerry DAC+ ADC. This neat baby’s have a sample rate up to 192kHz both on input and output and designed to be used on Raspberry Pi.

Newly kernel’s support this out of the box, but if you do need to install the kernel drivers you can find them here

While following my colleague’s guide it was quite quick and simple to set up the server side. The challenge was to get the sound from the input of the HiFiBerry and over to the server as a source.

After a lot of attempt I found a way. This is what I did to get the sound to snapcast.

Dj Corner

First find the deviceid for HiFiBerry on the pi

aplay -l
**** List of PLAYBACK Hardware Devices ****
card 0: sndrpihifiberry [snd_rpi_hifiberry_dacplusadc], device 0: HiFiBerry DAC+ADC HiFi multicodec-0 [HiFiBerry DAC+ADC HiFi multicodec-0]
  Subdevices: 0/1
  Subdevice #0: subdevice #0

Then set set up a sink server for snapcast server to conenct to

pulseaudio --start
pactl load-module module-alsa-source device=hw:0,0
pactl load-module module-loopback source=alsa_input.hw_0_0 sink=auto_null latency_msec=1

Snapcast server

Set up a sink in PulseAudio to get sound from the raspberry pi at the DJ corner.

pactl load-module module-tunnel-source server=192.168.1.138 source=alsa_input.hw_0_0
parec -d tunnel-source.192.168.1.138 >/tmp/snapfifo

Add pipe in /etc/snapcast.conf

pipe:///tmp/snapfifo?name=DJCorner[&mode=read]

Snapcast Server seen the DJ Corner:

Raspberry Pi connected to the sound source:

Did it work ?

While testing it out and got it working, but it was a bit to much delay ( see video below ). Also, to be fair, snapcast is more of a open-source version of Sonos and not build to do exactly what I was trying to do. Snapcast is super cool tho. You can read more about snapcast from the github page here.

Try to remove delay

I did a couple of things to try and remove the delay.

Remove any fs protection on filesystem to try and make it more snappy on IO

sudo sysctl fs.protected_fifos=0

In the configuration of the pipe in /etc/snapcast.conf you can set more variables to adjust processing.

[&codec=<codec>][&sampleformat=<sampleformat>][&chunk_ms=<chunk ms>]

sampleformat: Default sample format of the stream source, e.g. 48000:16:2
codec: The codec to use to save bandwith, one of:
        flac [default]: lossless codec, mean codec latency ~26ms
        ogg: lossy codec
        opus: lossy low latency codec, only supports 48kHz, if your stream has a different sample rate, automatic resampling will be applied, introducing further latecy
        pcm: lossless, uncompresssed. No latency.
chunk_ms: Default source stream read chunk size [ms]. The server will continously read this number of milliseconds from the source into a buffer, before this buffer is passed to the encoder (the codec above)

Pulseaudio attempt

While sitting their and think about the major failure this project turned out to be with the use of snapcast it downed on me. The communication I have made between using PulseAudio. Why not remove snapcast and just send the feed directly to other raspberry pi that was planed to be used as a snapcast client for my stereo in the first place? I guess the research wasn’t wasted at all!

I started with formatting the raspberry pi i used on the DJ Corner for a fresh start. Since I could not remember what version of the pi I used, I was forced to remove the HAT. It was Raspberry Pi 3 Model B+’s btw.

After flashing in the new clean images I did ran the following commands to start on both devices.

sudo apt update 
sudo apt install pulseaudio pulseaudio-utils

After then added the following lines in the /etc/pulse/default.pa

load-module module-native-protocol-tcp auth-ip-acl=127.0.0.1;192.168.0.0/23 load-module module-zeroconf-publish

192.168.0.0/23 been my local network range this pi’s will operate on.

After that I restarted PulseAudio on both pi’s.

pulseaudio --kill 
pulseaudio --start

Dj Corner

On the DJ Corner side as before, I needed to find the correct input device again like before in my last attempt.

aplay -l

**** List of PLAYBACK Hardware Devices ****
card 0: sndrpihifiberry [snd_rpi_hifiberry_dacplusadc], device 0: HiFiBerry DAC+ADC HiFi multicodec-0 [HiFiBerry DAC+ADC HiFi multicodec-0]
  Subdevices: 0/1
  Subdevice #0: subdevice #0

Now edit the /etc/pulse/default.pa on the DJ Corner pi again with the following line

load-module module-loopback source=alsa_input.hw_0_0

Now if we run this command the DJ corner will try to establish connection with the other pi on 192.168.1.55

pactl load-module module-tunnel-sink server=192.168.1.55 sink_name=DJCorner

Receiver Side

Now we just need to make the receiver end of the system. To achieve this I did the following on the other Pi.

PulseAudio treats output devices as sinks. To identify the sink corresponding to your HiFiBerry, you need to list all available sinks on the receiver.

pacmd list-sinks

Look for a sink entry with a description mentioning your HiFiBerry (e.g., “HiFiBerry DAC+”) It will look something like this:

name: <alsa_output.platform-soc_sound.stereo-fallback>
device.description = "HiFiBerry DAC+ Analog Stereo"

Make a note of the sink name

To play audio from the sender over the network, you need to load a tunnel source in PulseAudio. This connects the receiver to the sender Raspberry Pi.

pactl load-module module-tunnel-source server=192.168.1.60

After the tunnel is set up, you need to ensure that the HiFiBerry output is the default sink for audio playback.

pacmd set-default-sink <hifiberry_sink_name>

Replace <hifiberry_sink_name> with the actual sink name you found earlier.