How tphakala/birdnet-go works
BirdNET-Go is a self-hosted, real-time wildlife sound analyzer. Audio sources feed capture and analysis, where installed classifiers produce detections that are stored, exposed in the web interface, and evaluated for alerts. The interface also supports configuration, analytics, imports, and system operations. README statements establish the product workflow and major capabilities; source excerpts show selected audio, model, alerting, storage, and server responsibilities. Several UI and integration relationships are documented but their implementation was not sampled.
Main entry points: audio_pipeline_service.go Analysis pipeline · App.svelte Web application · engine.go Audio engine · model_manager.go Model manager
Components 21
Audio and detection
- Audio engine
internal/audiocore/engine/engine.go - Stream pipeline
internal/audiocore/stream/pipeline.go - Analysis pipeline
internal/analysis/audio_pipeline_service.go
Data and models
- Model manager
internal/classifier/model_manager.go - Detection storage
internal/datastore/v2/manager.go - Species lookup
internal/speciesindex/service.go
Web experience
- HTTP server
internal/api/server.go - Web application
frontend/src/App.svelte - Dashboard
frontend/src/lib/desktop/features/dashboard/pages/DashboardPage.svelte - Analytics pages
frontend/src/lib/desktop/features/analytics/pages/Species.svelte - Live audio view
frontend/src/lib/desktop/features/live-stream/pages/LiveStreamPage.svelte - Settings interface
frontend/src/lib/stores/settings.ts
Operations and integrations
- Application lifecycle
internal/app/app.go - Configuration
internal/conf/storage.go - Alert rules
internal/alerting/engine.go - MQTT integration
internal/mqtt/client.go - BirdWeather
internal/birdweather/birdweather_client.go - Support diagnostics
internal/support/database_collector.go
Other components
- Operator
- Audio sources
- Alert destinations
Connections 21
- Operator → Web application: uses
- Audio sources → Audio engine: provide audio
- Audio engine → Stream pipeline: streams frames
internal/audiocore/engine/engine.go - Audio engine → Analysis pipeline: feeds analysis
- Analysis pipeline → Model manager: uses classifiers
internal/analysis/audio_pipeline_service.go - Model manager → Analysis pipeline: classifies audio
- Analysis pipeline → Detection storage: stores detections
- Analysis pipeline → Species lookup: looks up species
- Detection storage → HTTP server: serves detection data
- HTTP server → Web application: serves app and API
internal/api/server.go - Web application → Dashboard: renders
frontend/src/App.svelte - Web application → Analytics pages: routes to
frontend/src/App.svelte - Web application → Live audio view: routes to
frontend/src/App.svelte - Operator → Settings interface: configures
- Settings interface → Configuration: updates settings
- Analysis pipeline → Alert rules: emits detection events
internal/analysis/audio_pipeline_service.go - Alert rules → Alert destinations: dispatches alerts
- Analysis pipeline → MQTT integration: publishes events
- Analysis pipeline → BirdWeather: sends detections
- Operator → Support diagnostics: requests diagnostics
- Application lifecycle → Configuration: starts with settings