A complete, beginner-friendly guide to getting the very best out of your self-hosted Navidrome music server — from installation to perfect metadata, playback, and beyond.

Based on community recommendations and best practices.

1. Overview — What Are We Building?
2. Step 1: Install Navidrome (Docker)
3. Step 2: Tag Your Music with MusicBrainz Picard
4. Step 3: Fix Genre Tags
5. Step 4: Set Up Playback Apps
6. Step 5: Remote Access with Tailscale
7. Step 6: Add Radio Stations
8. Step 7: Create Smart Playlists
9. Step 8: View Your Listening Stats (Navidrome Wrapped)
10. Audio Quality Tips
11. Alternatives Considered
12. Quick Reference Links

Here's the full stack at a glance:

┌─────────────┐     ┌──────────────┐     ┌──────────────┐
│  Your Music  │────▶│   Picard     │────▶│  Genre Fix   │
│  (files)     │     │  (tagging)   │     │  (Essentia   │
│              │     │              │     │   or Manual)  │
└─────────────┘     └──────────────┘     └──────┬───────┘
                                                 │
                                                 ▼
┌─────────────┐     ┌──────────────┐     ┌──────────────┐
│  Feishin    │◀────│  Navidrome   │◀────│  Clean Music  │
│  (desktop)  │     │  (server)    │     │  Library      │
├─────────────┤     │              │     └──────────────┘
│  Symfonium  │◀────│              │
│  (mobile)   │     └──────┬───────┘
└─────────────┘            │
                    ┌──────┴───────┐
                    │  Tailscale   │
                    │  (remote)    │
                    └──────────────┘

• Docker and Docker Compose installed on your machine
• A music library folder on your system
• (Optional) Spotify and Last.fm API keys for artist images and scrobbling

Create a docker-compose.yml file:

services:
  navidrome:
    image: ghcr.io/navidrome/navidrome:latest
    container_name: navidrome
    environment:
      - ND_MUSICFOLDER=/music
      - ND_PORT=4533
      - ND_SPOTIFY_ID=your_spotify_client_id        # For artist images
      - ND_SPOTIFY_SECRET=your_spotify_client_secret  # For artist images
      - ND_LASTFM_APIKEY=your_lastfm_api_key         # For scrobbling
      - ND_LASTFM_SECRET=your_lastfm_secret           # For scrobbling
      - ND_LOGLEVEL=info
    volumes:
      - /path/to/your/navidrome/data:/data       # Navidrome database & config
      - /path/to/your/music/library:/music:ro    # Your music (read-only)
    ports:
      - "4533:4533"
    restart: unless-stopped

1. Go to developer.spotify.com
2. Log in and create a new app
3. Copy the Client ID and Client Secret
4. Paste them into ND_SPOTIFY_ID and ND_SPOTIFY_SECRET

1. Go to last.fm/api
2. Create a new API account/application
3. Copy the API Key and Shared Secret
4. Paste them into ND_LASTFM_APIKEY and ND_LASTFM_SECRET

docker compose up -d

Visit http://your-server-ip:4533 in your browser. On first launch, you'll create an admin account.

💡 Tip: Mount your music folder as :ro (read-only) so Navidrome can't accidentally modify your files.

MusicBrainz Picard is the go-to tool for automatically tagging your music with accurate metadata (artist, album, title, track number, cover art, etc.). It looks up your music against the massive MusicBrainz database — which has excellent coverage for Japanese and Korean music too.

If your music is on a headless server, run Picard as a web-accessible Docker container:

services:
  picard-web:
    image: aandree5/picard-web:latest
    container_name: picard-web
    restart: unless-stopped
    ports:
      - "8086:5000"
    volumes:
      - /path/to/picard/config:/picard-web:rw    # Picard config/state
      - /path/to/your/music:/music:rw            # Your music (read-write!)

docker compose up -d

Access Picard in your browser at http://your-server-ip:8086.

⚠️ Important: The music volume must be :rw (read-write) here because Picard needs to write tags to your files.

Download from picard.musicbrainz.org and install normally on Windows, macOS, or Linux.

1. Add files — Drag & drop your music folder into Picard
2. Cluster — Click "Cluster" to group files by album
3. Lookup — Click "Lookup" to match against MusicBrainz
4. Review — Check the matches are correct (green = good match)
5. Save — Click "Save" to write the tags to your files

Picard can also rename and reorganize your files based on metadata. Writing these scripts manually is complex, so use the Picard Filenaming Script Generator:

📦 WB's Picard Filenaming Script Generator

This interactive CLI tool lets you:

• Choose from 6 ready-made presets (e.g., Artist/[Year] Album/01 - Track.flac)
• Or use the custom builder wizard to configure every detail
• Preview example output before saving
• Export directly into Picard — no scripting knowledge needed

# Install and run
git clone https://github.com/WB2024/WBs-Picard-Filenaming-Script-Generator.git
cd WBs-Picard-Filenaming-Script-Generator
python3 picard_script_generator.py

💡 For Japanese/Korean music: MusicBrainz has strong coverage of J-Pop, K-Pop, anime OSTs, and more. Picard will pull the correct metadata including original-language titles when available.

Picard is great for most metadata, but its genre tags are often generic or inconsistent (e.g., tagging everything as "Rock" or "Pop"). There are two approaches to fix this:

📦 Essentia Music Tagger

This tool analyzes the actual audio waveform using machine learning to predict accurate genres and moods. No internet required after setup — everything runs locally.

What it does:

• Classifies across 400 genre categories (Discogs taxonomy)
• Detects moods (energetic, dark, happy, aggressive, etc.)
• Writes GENRE and MOOD tags directly to your files
• Supports FLAC & MP3

Quick start:

# Clone and set up
git clone https://github.com/WB2024/Essentia-to-Metadata.git
cd Essentia-to-Metadata

# Create virtual environment
python3 -m venv venv
source venv/bin/activate

# Install dependencies
pip install essentia-tensorflow mutagen numpy

# Download ML models (~87MB)
bash download_models.sh

# Run (always do a dry run first!)
python tag_music.py

Or via Docker:

# CPU mode
docker compose build essentia-tagger
MUSIC_DIR=/path/to/music docker compose run --rm essentia-tagger

# Dry run first!
MUSIC_DIR=/path/to/music docker compose run --rm essentia-tagger /music --auto --dry-run

Example output:

Input:  2Pac - Temptations.flac
Output: GENRE = Hip-Hop - Gangsta; Hip-Hop - East Coast Hip Hop
        MOOD  = Energetic; Dark

🧪 Always run in dry-run mode first to preview what changes will be made!

📦 Artist Genre Metadata Enforcer

If you prefer your own genre definitions rather than what an algorithm or database decides, this tool lets you define a genre once per artist and apply it to their entire discography.

Why use this?

• You decide that all 2Pac is Hip-Hop, period
• No more "Hip Hop" vs "Hip-Hop" vs "Rap" inconsistencies
• Define once, enforce everywhere — even across thousands of tracks

Quick start:

# Install dependency
sudo apt install python3-mutagen   # Debian/Ubuntu
# or: pip install mutagen

# Download and install
sudo curl -o /usr/local/bin/music-genre-enforcer \
  https://raw.githubusercontent.com/WB2024/Artist-Genre-Metadata-Enforcer/main/music-genre-enforcer.py
sudo chmod +x /usr/local/bin/music-genre-enforcer

# Run
music-genre-enforcer

Interactive workflow:

Genre for [Aphex Twin] (or l/p/?/s): p
Existing genres:
  [1] Electronic - Ambient
  [2] Hip-Hop
  [3] Post-Punk
  [4] Rock - Art
Pick number (blank cancels): 1
  -> Set [Aphex Twin] = Electronic - Ambient

Expected folder structure:

/your/music/library/
├── A/
│   ├── Aphex Twin/
│   │   ├── [1992] Selected Ambient Works 85-92/
│   │   │   ├── 01 - Xtal.flac
├── B/
│   ├── Bowie, David/
├── #/
│   └── 2Pac/
└── Various Artists/

💡 Pro tip: You can use both — run Essentia first for a baseline, then use the Genre Enforcer to override specific artists with your preferred labels.

📦 Feishin — A modern, Spotify-like desktop music player.

Why Feishin?

• Beautiful UI that looks and feels like Spotify
• Connects directly to Navidrome (and Jellyfin)
• Dual playback engines: MPV (high-quality native) or Web player
• Smart playlists, lyrics, scrobbling, themes
• Available on Windows, macOS, and Linux

Download: github.com/jeffvli/feishin/releases

Setup:

1. Download and install Feishin for your OS
2. Open Feishin → Add Server → Enter your Navidrome URL and credentials
3. Start browsing and playing!

For the best audio quality on desktop:

1. Use MPV as the playback engine in Feishin's settings
2. Connect a USB DAC (Digital-to-Analog Converter) to your computer
3. Set audio output to WASAPI (Windows) in Feishin/MPV settings
4. Enable audio-exclusive mode — this prevents Windows from resampling your audio

This bypasses the Windows audio mixer entirely, sending bit-perfect audio to your DAC.

📦 Symfonium — The best Android music player for self-hosted servers.

Why Symfonium?

• Connects to Navidrome (plus Plex, Jellyfin, Emby, cloud storage, and more)
• Offline caching — download music for when you're without internet
• Gapless playback, replay gain, and high-res audio support
• Cast to Chromecast, DLNA, Sonos, and more
• Android Auto & Wear OS support
• One-time purchase, no subscriptions, no ads

Setup:

1. Install from Google Play Store
2. Add your Navidrome server (URL + credentials)
3. Enjoy your music on the go!

Want to listen to your music when you're away from home? Don't expose your server to the internet with port forwarding — use Tailscale instead.

📦 Tailscale — Creates a secure, private network (VPN) between your devices with zero configuration.

1. Install Tailscale on your server (where Navidrome runs)
2. Install Tailscale on your phone/laptop (where you listen)
3. Both devices get a private Tailscale IP address
4. Access Navidrome using the Tailscale IP — encrypted, secure, no port forwarding

# On your server (Debian/Ubuntu)
curl -fsSL https://tailscale.com/install.sh | sh
sudo tailscale up

# Note the Tailscale IP it gives you (e.g., 100.x.y.z)

Install the Tailscale app on your phone/laptop from tailscale.com/download.

Then in Feishin or Symfonium, set your server URL to:

http://100.x.y.z:4533

🔒 Why Tailscale? No ports to open, no dynamic DNS to configure, encrypted WireGuard tunnel, and a generous free tier for personal use.

📦 Navidrome Radio Station Manager

A CLI tool to search, browse, and add internet radio stations to your Navidrome server from the Radio-Browser community database.

# Install dependency
pip3 install requests

# Download and install
curl -o /usr/local/bin/navidrome-radio \
  https://raw.githubusercontent.com/WB2024/Add-Navidrome-Radios/main/navidrome-radio.py
sudo chmod +x /usr/local/bin/navidrome-radio

# Run
navidrome-radio

On first run, enter your Navidrome database path (e.g., /path/to/navidrome/data/navidrome.db).

Then search and add stations:

MAIN MENU
1. Search and add radio stations
2. View existing stations in database
3. Inspect database (debug)
4. Change database path
5. Exit

Search by name, genre, country, or browse top-voted stations.
Select stations → type "add" → they appear in Navidrome immediately!

💡 Docker users: Your database path is on the host, not inside the container. Check your docker-compose.yml volumes to find it.

📦 Navidrome Smart Playlist Creator

Navidrome supports Smart Playlists — dynamic playlists defined as rules stored in .nsp files that automatically populate based on criteria you define (e.g., "all songs I've loved from the 80s" or "high-quality tracks I haven't played recently"). This fully guided CLI tool builds those rule files through numbered menus — no JSON editing, no memorising field names or operator syntax.

• 30+ ready-made presets — deploy instant playlists covering essentials, discovery, moods, decades, quality, and complex nested logic
• One-click deploy all — save every preset at once with a single menu choice
• 100+ fields — every field Navidrome supports, from core metadata to MusicBrainz IDs and ReplayGain values
• Nested rule groups — create sub-groups with their own AND/OR logic (e.g. "loved OR highly rated" inside an AND query)
• Multi-field sorting — sort by multiple fields with individual ascending/descending direction
• Fully guided numbered menus — no typing field names, operators, or syntax
• Beautiful UI — enhanced display via rich (optional, degrades gracefully)

git clone https://github.com/WB2024/Navidrome-SmartPlaylist-Generator-nsp.git
cd Navidrome-SmartPlaylist-Generator-nsp
pip install rich       # Optional — recommended for best UI
python navidrome_smart_playlist_creator.py

On first run, set your playlist directory — the folder where Navidrome scans for .nsp files. The path is saved and remembered for future sessions.

The wizard guides you through five steps to build a playlist:

1. Rule logic — ALL must match (AND) or ANY can match (OR)
2. Rules — pick a field → operator → value. Supports nested rule groups for complex AND/OR logic to any depth, with a live summary panel showing your rules as you build
3. Sort order — one or multiple sort fields, each with ascending/descending direction
4. Limit — optionally cap the number of tracks
5. Details — name and optional description

After completing the wizard, the generated JSON is previewed and you confirm before saving.

Choose Deploy preset playlists from the main menu to instantly save 30+ ready-made .nsp files:

Place .nsp files anywhere inside your music library folder — Navidrome finds them automatically on the next library scan. Smart Playlists refresh automatically when accessed.

💡 Recommended: Create a SmartPlaylists/ subfolder inside your music library root and save all .nsp files there to keep them organised.

⚠️ Docker users: If you use a separate PlaylistsPath volume, place .nsp files inside your music library folder instead — that's the most reliable approach across all Navidrome versions.

📦 Navidrome Wrapped

Navidrome Wrapped is a beautiful, Spotify Wrapped–inspired client-side web app that generates comprehensive all-time listening statistics from your Navidrome server. See your top artists, albums, songs, genre and decade breakdowns, diversity scoring, rating analysis, audio quality insights, and shareable summary cards — all in a stunning interactive slideshow.

• Pure client-side — no backend server needed; runs entirely in your browser
• Privacy-first — your credentials are never stored or sent anywhere except directly to your Navidrome instance; all processing is local
• Interactive slideshow with 8 sections: Welcome, Listening Time, Top Artists, Top Albums, Top Songs, Diversity, Ratings, and Summary
• Interactive charts for genre and decade distribution
• Shareable cards — generates 9:16 share images of your stats
• Demo mode — try it with sample data before connecting your server
• Caches results in localStorage for 90 days so repeat visits are instant
• Live hosted version available at mdeik.github.io/Navidrome-Wrapped

Just visit mdeik.github.io/Navidrome-Wrapped and enter:

• Your Navidrome server URL (e.g., http://your-server-ip:4533 or https://your-domain.com)
• Your username and password
• Click "Generate All-Time Wrapped"

It's just static files — no build step needed:

git clone https://github.com/mdeik/Navidrome-Wrapped.git
cd Navidrome-Wrapped
# Open index.html in your browser, or serve it:
python3 -m http.server 8080
# Then visit http://localhost:8080

💡 Tailscale users: If accessing your Navidrome remotely, use your Tailscale IP as the server URL (e.g., http://100.x.y.z:4533).

🎮 Try demo mode first — Navidrome Wrapped includes a demo mode with sample data so you can explore the interface before connecting your own server.

These tips will help you get the best listening experience:

Feishin → MPV Engine → WASAPI (Exclusive) → USB DAC → Amp → Speakers

This chain ensures:

• ✅ Bit-perfect audio output
• ✅ No Windows audio mixer resampling
• ✅ Direct hardware access
• ✅ Maximum quality from your files

These alternatives are worth knowing about:

Here's the recommended sequence for setting everything up from scratch:

1. 🐳  Deploy Navidrome via Docker
2. 🏷️  Tag your music with Picard (get metadata right first)
3. 📁  Generate a filenaming script to organize your library
4. 🎸  Fix genres with Essentia and/or Genre Enforcer
5. 🎧  Install Feishin (desktop) and/or Symfonium (mobile)
6. 🌐  Set up Tailscale for remote access
7. 📻  Add radio stations
8. 🎼  Create smart playlists
9. 📊  View your listening stats with Navidrome Wrapped
10. 🔊  Optimize your audio hardware chain
11. 🎶  Enjoy your music!

Happy listening! 🎶

If you found this guide helpful, consider buying me a coffee!