Real-time passive acoustic monitoring

Listening to the ocean,
around the clock

A distributed network of underwater microphones, monitored continuously by machine learning. Detecting whale and dolphin vocalizations as they happen, across two oceans.

Listen to a featured detection Loading...

Launch Dashboard

Hydrophone Streams

Total Detections

Species Detected

ML Models Active

About the project

Why passive acoustic monitoring matters

The ocean is vast, opaque, and noisy. Marine mammals communicate over distances of tens to hundreds of kilometers using sound — but human-generated noise, climate shifts, and habitat disruption are changing the acoustic landscape they depend on. Understanding who is vocalizing, where, and when is fundamental to conservation.

The challenge

Hydrophones generate continuous audio streams — thousands of hours per day across a distributed network. Manual review is impossible at scale. Vocalizations are transient, often buried in ambient noise, and span frequency ranges from 10 Hz infrasound to 100 kHz echolocation.

Our approach

Ocean Song ingests audio from 10 hydrophone streams across the Pacific and Atlantic, running 5 specialized neural networks in real time. When a vocalization is detected, the system automatically extracts, denoises, and archives the audio clip with its spectrogram — building a searchable, persistent acoustic record.

The network

10 hydrophones, 2 oceans

From the deep Monterey Canyon to the glacial waters of Disko Bay, each station listens continuously.

MBARI MARS

Monterey Bay, CA — 900m depth
Humpback, blue whale, dolphins

Shoutcast

Disko Live

Disko Bay, Greenland
Bowhead, narwhal, humpback, fin whale

Shoutcast

Orcasound — Port Townsend

Admiralty Inlet, WA
Orca, humpback

HLS

Orcasound — Sunset Bay

Sunset Bay, WA
Orca, humpback

HLS

SIMRES

Saturna Island, BC
Orca, humpback

Icecast

Orcasound — North SJC

North San Juan Channel, WA
Orca, humpback

HLS

Orcasound — MaST Center

Des Moines, WA
Orca, humpback

HLS

Orcasound — Andrews Bay

Andrews Bay, WA
Orca, humpback

HLS

How it works

From hydrophone to detection

🌊

Hydrophones

Underwater microphones streaming 24/7 from coastal and deep-ocean stations

🔄

Ingestion

Real-time audio capture at 16-48 kHz via HLS, Icecast, and Shoutcast protocols

🧠

Detection

5 neural networks analyzing every second: TF Hub, PyTorch, DSP classifiers

🎵

Recording

Automatic clip extraction, spectrogram generation, SNR measurement, denoising

Species detection

What we listen for

Humpback Whale

Megaptera novaeangliae

Model: Google TF Hub humpback_whale
Range: 80 – 4,000 Hz
Method: Mel spectrogram CNN, 3.9s windows
Threshold: 90%, 2 consecutive

Orca

Orcinus orca

Model: orcAI (ETH Zurich)
Range: 500 – 10,000 Hz
Method: 7 call-type classifier, PyTorch
Threshold: 50%

Blue Whale

Balaenoptera musculus

Model: MBARI ResNet50 A-call
Range: 10 – 100 Hz
Method: Spectrogram classifier, F1 = 0.945
Threshold: 70%

Sperm Whale

Physeter macrocephalus

Model: Spectrogram band-power
Range: 2,000 – 30,000 Hz
Method: Click train detection, ICI analysis
Threshold: 70%

Explore the live dashboard

Browse real-time hydrophone streams, listen to recent detections, and denoise whale vocalizations.

Enter Dashboard

Ocean Song / bioacoustic monitoring

0 sources

0 monitored

0 detections

System: offline

Analysis Pipeline

Google Humpback Whale TF Hub · tfhub.dev/google/humpback_whale/1 · Threshold: 90% · 2 consecutive windows

Humpback whale detection · All 7 streams · 10kHz input · 3.9s windows

orcAI ETH Zurich · github.com/ethz-tb/orcAI · Threshold: 50%

Orca vocalization detection · 5 Orcasound + SIMRES · 48kHz input · 7 call types

MBARI Blue Whale A-call MBARI · ResNet50 classifier · Threshold: 70%

Blue whale A-call detection · MBARI MARS stream · 2kHz input · F1=0.945

Sperm Whale Click Train Click detection + ICI analysis · Threshold: 70%

Sperm whale echolocation click trains & codas · MBARI MARS · 2-7kHz bandpass + inter-click interval regularity

Fin Whale · Dolphin Disabled — produced false positives, pending better models

7 streams monitored · 16kHz ingestion · 5-min ring buffers · 4 active models · Inference every 10s

Selected Event

Select an event from the sidebar

Audio Pipeline

Parameters

Snapshots

Presets

Selected Event

Select an event from the sidebar

No event selected

Source Separation

—

Agent & API Access

Base URL: https://api.oceansong.live

Swagger Docs agent.json Claude Code Skill OpenAPI JSON

Quick start:
curl https://api.oceansong.live/summary
curl https://api.oceansong.live/highlights?hours=24

No authentication required · WebSocket at wss://api.oceansong.live/ws

Questions, feedback, collabs? Telegram @roman_still →

Listening to the ocean,
around the clock

Why passive acoustic monitoring matters

The challenge

Our approach

10 hydrophones, 2 oceans

From hydrophone to detection

Hydrophones

Ingestion

Detection

Recording

What we listen for

Explore the live dashboard

Ocean Song / bioacoustic monitoring

Noisereduce

Biodenoising

Listening to the ocean,around the clock

Why passive acoustic monitoring matters

The challenge

Our approach

10 hydrophones, 2 oceans

From hydrophone to detection

Hydrophones

Ingestion

Detection

Recording

What we listen for

Explore the live dashboard

Noisereduce

Biodenoising

Listening to the ocean,
around the clock