VinylScrobbler

Listens to audio from your turntable or any physical source, identifies tracks with Shazam, scrobbles them to Last.fm, and saves local listening statistics you can browse from any device on your network.

Runs fully headless on a Raspberry Pi — no display, no graphical environment required.

---

What it does

1. Captures audio from your source (turntable USB output, splitter, or microphone).
2. Identifies the track with Shazam (artist, album, genre, cover art).
3. Scrobbles automatically to Last.fm.
4. Saves local statistics to a SQLite database (artists, albums, tracks, genres, listening time).
5. Serves a web dashboard accessible from any device on your local network.

---

Stats dashboard

!(https://i.imgur.com/hBoUxfR.png) Open from any device connected to your network:

http://<your-raspberry-ip>:8000 [Or the port that you designated in the setup]

Includes:

• Top artists
• Top albums
• Top tracks
• Genre breakdown
• Total listening time
• Full playback history

Filterable by: Today / Week / Month / Year / All time.

---

Setup

Requirements

• Raspberry Pi running Raspberry Pi OS (Bookworm recommended)
• Python 3.11+
• Audio source connected (turntable USB output, splitter + USB converter, or microphone)
• (OPCIONAL)Last.fm account with API key and secret — https://www.last.fm/api/
• (OPCIONAL)Spotify account with API key and secret - https://developer.spotify.com/dashboard

Installation

# 1. Clone or extract the project
cd VinylScrobbler

# 2. Run setup (without sudo)
chmod +x setup.sh
./setup.sh

The script will:

• Create a virtual environment and install dependencies
• Generate backend/config.py and backend.sh When the setup is over, you just execute it with

./backend.sh

Then you enter the dashboard, where it will

• Let you select the audio input device
• Ask for your Last.fm and Spotify credentials

Autostart on boot

chmod +x autostart_setup.sh
./autostart_setup.sh

# Run once with sudo to keep the service alive without an active login session:
sudo loginctl enable-linger $(whoami)

This creates a systemd user service (VinylScrobbler.service) that starts with the system, no graphical session needed.

Run manually

./backend.sh

---

Project structure

VinylScrobbler/
├── backend/
│   ├── main.py            — FastAPI: detection loop, WebSocket, stats API
│   ├── audio.py           — Audio capture
│   ├── shazam_service.py  — Track identification (includes genre extraction)
│   ├── lastfm_service.py  — Last.fm scrobbling
│   ├── stats_service.py   — SQLite: save and query statistics
│   ├── state.py           — Global app state
│   ├── logger.py          — Logging
│   └── config.py          — Generated by setup.sh
├── stats_web/
│   └── index.html         — Web dashboard (served at /)
├── setup.sh
├── autostart_setup.sh
└── requirements.txt

---

API endpoints

Endpoint	Description
GET /	Stats dashboard
GET /api/now	Currently playing track (JSON)
GET /api/stats/summary?period=week	Summary: top artists, albums, tracks, genres, listening time
GET /api/stats/history?page=1	Paginated playback history
WS /ws	WebSocket for real-time status updates

period accepts: today, week, month, year, all

---

Logs

tail -f ~/logs/vinylscrobbler.log

---

Configuration

Edit backend/config.py directly. Common parameters:

Parameter	Default	Description
CHUNK_SECONDS	10	Seconds of audio sent to Shazam per attempt
RETRY_DELAY	15	Seconds between identification attempts
SILENCE_TIMEOUT	30	Seconds of silence before switching to idle
VOLUME_THRESHOLD	0.01	Minimum RMS level to consider audio present

Restart the service after changing config:

systemctl --user restart VinylScrobbler

---

Audio source options

There are three common ways to feed audio into VinylScrobbler:

1. Direct USB output from the source (e.g. i use a Audio-Technica AT-LP120XUSB, so it has a native usb output).
2. A splitter between the source and your amplifier, with one output going to a USB audio converter.
3. A microphone positioned near the speakers — less precise, may trigger on ambient noise.