summaryrefslogtreecommitdiff
path: root/20210919-raspberry-pi-white-noise-machine.md
diff options
context:
space:
mode:
authorVincent Douillet <vincent@vdouillet.fr>2021-10-11 23:03:07 +0200
committerVincent Douillet <vincent@vdouillet.fr>2021-10-11 23:03:07 +0200
commit6ee862efd612832e88b154105afe4715fd1cc78b (patch)
treef3e4d2be1615d9e37a9a120b18bee6aa3bbadf6a /20210919-raspberry-pi-white-noise-machine.md
first commit
Diffstat (limited to '20210919-raspberry-pi-white-noise-machine.md')
-rw-r--r--20210919-raspberry-pi-white-noise-machine.md56
1 files changed, 56 insertions, 0 deletions
diff --git a/20210919-raspberry-pi-white-noise-machine.md b/20210919-raspberry-pi-white-noise-machine.md
new file mode 100644
index 0000000..666157e
--- /dev/null
+++ b/20210919-raspberry-pi-white-noise-machine.md
@@ -0,0 +1,56 @@
+# Raspberry Pi white noise machine
+
+September 19, 2021
+
+A white noise machine can help find sleep or mask ambient noise or even tinnitus. It can also be used to muffle conversation from potential eavesdroppers. I wanted to experiment with a white noise machine for sleep purposes so I decided to build one with a Raspberry Pi I had lying around.
+
+For this you will need:
+
+* A Raspberry Pi equipped with a jack port and an OS installed (I used a Pi 3B with Raspbian OS)
+* Speakers connected to the Pi
+
+Actually a Pi 3 is a bit overkill for this. I have a Pi Zero lying around that could fit the bill except it needs a USB dongle to connect speakers and I don't have such a dongle.
+
+## Preparing the white noise
+
+First you need to find a white noise sample, preferably a long one. I chose to extract the sound from a 3 hour long YouTube video of a waterfall. Of course, choose what suits you. If you want to extract the sound from a YouTube video, you could use one of the many online tools but I won't recommend that to you. Instead, you can use [youtube-dl](https://github.com/ytdl-org/youtube-dl):
+
+ $ youtube-dl -x --audio-format mp3 https://www.youtube.com/watch?v=dQw4w9WgXcQ
+
+Some explanations for the options used:
+
+* `-x` asks youtube-dl to extract the audio from the video. As stated in youtube-dl's help, ffmpeg or avconv is required for this
+* `--audio-format` should be self-explanatory
+
+## Creating a white noise service
+
+Next we are going to create a service to play the white noise automatically at system startup. My Raspberry Pi is running raspbian so systemd is used to manage services. Create a `/etc/systemd/system/whitenoise.service` file with the following contents:
+
+ [Unit]
+ Description=White noise machine
+ StartLimitIntervalSec=0
+
+ [Service]
+ Type=simple
+ Restart=never
+ User=pi
+ ExecStart=mpg123 --loop /home/pi/whitenoise.mp3
+
+ [Install]
+ WantedBy=multi-user.target
+
+You should adjust your username and path to the file. I used [mpg123](https://mpg123.org/) because it's the simplest and lightest audio player I could find. Its loop option is handy to keep playing infinitely. It needs to be installed:
+
+ # apt install mpg123
+
+## Auto start, auto stop
+
+To start automatically the service on system startup, you just need to enable the service on startup:
+
+ # systemctl enable whitenoise
+
+To stop it, you can use a cron job set at the time you want it to stop:
+
+ 55 6 * * * sudo systemctl stop whitenoise
+
+This will stop the white noise at 6:55am. And that's it! Enjoy your homemade white noise machine!