ListenBrainz local brings recommendation features to your local Subsonic hosted music collection. Many subsonic projects lack recommendations or good playlisting tools and this project aims to create another layer on top of these projects to show what can actually be done when you have lots of good music metadata.
Note: Your collection must be tagged with MBIDs in order for this project to work. Please do not open any bug reports asking us to make ListenBrainz Local work with an untagged collection. We will never do this.
Support for LB Radio and Weekly jams exists now -- music playback and saving playlists to the subsonic hosts are now working and many other recommendation tools will be added over time.
This project is a web app that requires a subsonic hosted music collection:
So far we've only tested with Navidrome and Gonic, so we would appreciate help testing other setups.
Due to the current state of the services (navidrome, etc) not all of these services are currently supporting strong auth that does not pass plaintext passwords. The ListenBrainz team is actively contributing to these projects, but this process isn't complete yet.
In the meantime password are stored in plaintext in the database and passed in plaintext to the front-end.
Clearly we will fix this as soon as we are able.
We recommend the Docker setup for users who wish run this service; if you want to do development a python virtual environment setup is better.
Clone this repo and then in the listenbrainz-local and copy the file _env to .env and then ensure that it contains the following completed configuration lines:
SECRET_KEY="any string of your choice"
AUTHORIZED_USERS=<comma separated list of LB usernames that are allowed to use this site>
ADMIN_USERS=<comma separated list of LB usernames that are allowed to administer this site>
DOMAIN=<the fully qualified domain where lb-local will be available, without port number>
PORT=<the port numer where lb-local will be available>
MUSICBRAINZ_CLIENT_ID=<MB client id, see https://musicbrainz.org/account/applications >
MUSICBRAINZ_CLIENT_SECRET=<MB client secret, see https://musicbrainz.org/account/applications >
Note: Put values in "" if the values contain spaces.
When adding a MusicBrainz application, create the application here: https://musicbrainz.org/account/applications . For 'Callback URI' enter:
http://[:]/auth -> http://mydomain.com:500/auth
Then continue with one of the configuration sections below:
Make sure you have Docker installed and correctly setup. ( https://docs.docker.com/engine/install/ )
docker compose build && docker compose up
And LB Local will be available at the specified DOMAIN:PORT specified above!
If you wish to also create a proxy with TLS (https) support, then follow the steps above, but also add the following to your .env file:
EMAIL=<your email address>
The PORT configuration will be ignored in this setup and port 443 will be used.
Then start the service with:
docker compose build && docker compose -f docker-compose-https-proxy.yml up -d
Currently it is easy to run the app from the command line, so that is the recommended approach for development:
python3 -mvenv .virtualenv
source .virtualenv/bin/activate
pip install -r requirements.txt
./lb_local.py
The app should then be available at the URL configured in .env
