Connect Muse S

Overview

The Muse S (Gen 2, also known as Athena) is a consumer EEG headband that connects directly to the Omnio mobile app over Bluetooth Low Energy. It provides 4-channel EEG at 256 Hz, 3-channel PPG at 64 Hz, and accelerometer data — all processed on-device in real-time.

This is the only device that gives Omnio access to brainwave data, enabling EEG-based sleep staging, neurofeedback meditation sessions, and neural biomarkers like alpha/theta ratio.

What data is collected

Category	Metrics
EEG band powers	Delta (0.5–4 Hz), theta (4–8 Hz), alpha (8–13 Hz), beta (13–30 Hz), gamma (30–50 Hz) — relative power per channel. Note: gamma readings from consumer dry electrodes primarily reflect facial muscle activity (EMG) rather than cortical gamma oscillations.
Calm score	Real-time 0–100 meditation quality score based on alpha/beta ratio
Alpha/theta ratio	Lower values indicate deeper meditative states
PPG heart rate	Optical HR derived from infrared PPG (fallback when no external HR device is connected)
fNIRS oxygenation	Relative HbO and HbR (oxygenated/deoxygenated haemoglobin) from dual-wavelength PPG via Beer-Lambert law
Accelerometer	Head motion detection for artifact gating
Sleep staging	EEG-based sleep classification (deep, light, REM, awake) using spectral features
Electrode quality	Per-channel contact quality (good/fair/poor) with continuous 0–1 score

EEG channels

The Muse S has four dry electrodes:

Channel	Position	Location
TP9	Left ear	Temporal
AF7	Left forehead	Frontal
AF8	Right forehead	Frontal
TP10	Right ear	Temporal

EEG data is bandpass-filtered (0.5–50 Hz) and processed in sliding windows. Windows are rejected if motion artifact is detected (from the accelerometer) or if the spectral shape indicates muscle contamination (EMG).

Neurofeedback and meditation

During a meditation session, the Muse provides real-time feedback:

• Calm score — A 0–100 score updated every few seconds, based on alpha band power relative to beta. Higher alpha = calmer state.
• Protocol training — You can select specific neurofeedback protocols (e.g. alpha enhancement, theta training) with adaptive thresholds that adjust to your personal baseline.
• Calibration — Each session begins with a brief baseline calibration so the neurofeedback is relative to your current state, not an absolute target.

fNIRS oxygenation

The Muse’s PPG sensor (infrared + red wavelengths) enables functional Near-Infrared Spectroscopy (fNIRS):

• HbO (oxygenated haemoglobin) — Increases indicate greater neural activity.
• HbR (deoxygenated haemoglobin) — The inverse signal.

These are relative changes (Δ[HbO], Δ[HbR] in µmol/L) from session start, not absolute concentrations. They’re shown in the session analytics view.

How to connect

1. Open the Omnio mobile app.
2. Go to Settings → Heart Rate Monitor.
3. Tap Scan for Devices.
4. Turn on your Muse S and ensure it’s not connected to another app (Muse app, etc.).
5. The device appears as “Muse-XXXX” in the scan list.
6. Tap to connect. The app establishes a BLE link and initialises the device in standby mode.

First-time setup

On first connection, the app shows a setup guide:

1. Position the headband — Place it on your forehead with the earpieces hooked behind your ears.
2. Check electrode contact — The Sensor Contact card shows quality for each of the 4 channels with colour-coded indicators (green = good, yellow = fair, red = poor).
3. Adjust until all sensors are green — Gently adjust the headband position. At least 3 of 4 channels need good or fair contact for reliable data.

Connection details

• Protocol: Custom GATT service (OpenMuse protocol). Initialisation sequence: version query → status → halt → preset selection (EEG4 + PPG) → standby.
• Data streaming: EEG/PPG/ACC streams are activated on demand (when entering a session) and deactivated when idle, to conserve battery.
• Stream tiers: The app manages three tiers — idle (BLE link only), standby (keepalive pings), and active (full data streaming).
• MTU: Requested at 247 bytes. On Android, high connection priority is requested for reliable high-throughput EEG streaming.
• Auto-reconnect: Up to 5 reconnection attempts with exponential backoff (1s, 2s, 4s, 8s, 15s). The reconnect counter resets after 5 seconds of stable connection.
• Battery: Read from the Muse’s control characteristic status response.

Multi-device use

The Muse S works best as part of a multi-device setup:

• Muse S + Polar H10 or Verity Sense: The Polar device provides accurate HR/HRV from ECG or optical sensing, while the Muse provides EEG. When an external HR device is connected, the Muse’s PPG-derived heart rate is automatically disabled to avoid duplicate HR data. This is the recommended setup for sleep tracking — combining EEG sleep staging with chest-strap HRV.
• Standalone: When no external HR device is connected, the Muse derives heart rate from its PPG sensor (forehead optical, 64 Hz). This is less accurate than a chest strap but provides a reasonable fallback.

Sleep tracking

The Muse S enables EEG-based sleep staging, which is significantly more accurate than HR-only methods:

• Deep sleep (N3) — Identified by high delta power (slow waves).
• Light sleep (N1/N2) — Moderate delta, sleep spindle features.
• REM — Low-voltage mixed frequency with rapid eye movement artifacts.
• Awake — High alpha/beta, movement.

How to use sleep tracking

1. Connect your Muse S (and optionally a Polar chest strap for HR/HRV).
2. From Settings → Heart Rate Monitor, arm sleep mode.
3. The app detects sleep onset automatically from HR and EEG patterns.
4. EEG sleep epochs are recorded throughout the night.
5. In the morning, stop the session to sync the sleep summary to your dashboard.

Sleep data from the Muse is combined with HR/HRV data (from the Muse PPG or a paired Polar device) for a comprehensive sleep report. When both EEG and PPG classifiers are available, the EEG classifier takes precedence as it’s the clinical gold standard for sleep staging.

CLAS (Closed-Loop Auditory Stimulation)

When supported, the Muse can deliver precisely timed audio tones during deep sleep (N3) to enhance slow-wave activity. This experimental feature:

• Detects slow oscillation up-states in real-time from the EEG signal.
• Triggers brief audio stimuli (pink noise bursts) synchronised to the brain’s natural sleep rhythms.
• Tracks stimulation events in the session log.

CLAS requires the Muse to be the active EEG source during a sleep session.

Electrode quality

The Muse S uses dry electrodes, so signal quality depends heavily on proper positioning:

• Good (green) — Standard deviation 5–80 µV. Real EEG signal with normal biological variance.
• Fair (yellow) — Standard deviation 80–150 µV. Some motion artifact or loose contact.
• Poor (red) — Standard deviation < 2 µV (flatline/railed) or > 150 µV (very loose electrode).

The electrode quality display updates in real-time. For best results:

• Ensure the forehead sensors (AF7, AF8) make direct skin contact — push hair aside.
• The ear sensors (TP9, TP10) should sit snugly behind the earlobes.
• Stay relatively still during sessions. Head movement corrupts EEG across all channels.

Troubleshooting

• Muse not appearing in scan — Ensure the Muse S is powered on (hold the button for 2 seconds until you hear a tone). Disconnect from any other apps (Muse app, Bluetooth settings) first — the Muse only supports one active BLE connection.
• “Disconnected during initialization” — The Muse sometimes drops during the init command sequence. The app will auto-retry. If it persists, power cycle the headband.
• Poor electrode quality on all channels — The headband may be mispositioned. Refit it ensuring forehead sensors touch skin directly. Hair between the sensor and skin is the most common cause.
• EEG windows rejected (motion) — The accelerometer detects head movement and rejects contaminated data windows. Stay as still as possible during sessions. The data yield indicator shows what fraction of windows are passing quality checks.
• No PPG heart rate — PPG HR only processes during active sessions. If an external HR device (e.g. Polar H10) is connected, Muse PPG HR is disabled (this is expected).
• Battery drains fast — EEG streaming at 256 Hz is power-intensive. Expect 4–6 hours of active streaming. The standby tier conserves battery when you’re not in a session.
• Forget and re-pair — Go to Settings → Heart Rate Monitor → tap Forget, then scan again. Power cycle the Muse before re-scanning.