Birdsong ID · on-device

1

Analyze

What birds are in your recording?

Audio file

Pick a file — its recording date is read from the file’s metadata or filename.

Location — click the map, use your current location, or enter coordinates (filters species to your location, on-device)

Latitude Longitude

🔒 Your location drives the species filter entirely on-device and is never sent anywhere to filter. If you’re online, an approximate point (rounded to about a kilometer) is sent to OpenStreetMap only to show the nearby place name — nothing is tied to your identity.

Advanced options

Saved results JSON (optional)

Already classified this recording? Add its *_detections.json (with the audio) to review it again without re-analyzing.

Min confidence Recording date (blank = auto-detect from file) High-pass filter (Hz) — 0 = off (matches desktop) Loudness normalization — off matches desktop BirdNET

Loading model…

Your recordings

Tap a recording to review it again without re-analyzing. Audio isn’t saved — reopen the file to play clips. Export bundles them to a file you can import (merge) on another device.

2

Triage — does this look right?

Note (optional)

You can come back to triage any time.

Saves this recording to your library and unlocks the detections table, JSON export, and mastering.

3

Master & export

Auditiontune while you listen

Compare

Original Mastered

Loudness

None Peak RMS

Leaves the recording at its original level.

Target level−3 dB

High-pass filter Remove low rumbleCuts wind, handling & traffic below the birdsong.

Cutoff120 Hz