MindVault • Muse S Athena (MS-03)

The open research client and reverse-engineering protocol specification for the Muse S Athena headband.
Stream raw 256 Hz EEG, verify real-time contact impedance, export clinical EDF+ recordings, and eliminate cloud lock-in.

---

  

---

Why This Exists

Most consumer neurotechnology software forces researchers and developers into proprietary walled gardens. Official mobile applications lock raw data behind subscription paywalls, discard high-resolution telemetry, or restrict data export to high-level "meditation scores."

Legacy community tools like BlueMuse are Windows-only and require external BLED112 USB dongles. Python scripting workarounds often suffer from Bluetooth connection drops, broken timestamp jitter, and undocumented firmware changes on the new Muse S Athena (MS-03) model.

MindVault solves this. Built with a native Local-First Architecture, MindVault connects directly to the Muse S Athena headband over your computer's built-in Bluetooth (Windows, macOS, and Linux). It delivers an instant, zero-latency 256 Hz electroencephalography visualizer, continuous impedance monitoring, and standard EDF+ / CSV exports for MNE-Python and EEGLAB. No phone app, no dongle, and no cloud upload required.

---

At a Glance

Feature	What It Means
Direct Desktop BLE	Connects to Muse S Athena via native OS Bluetooth on Windows, macOS, and Linux with zero external dongles.
High-Precision 256 Hz EEG	Full 14-bit packed telemetry on 4 clinical channels (TP9, AF7, AF8, TP10) with ~0.0885 µV resolution.
Real-Time Contact Quality	Continuous band-power impedance proxy highlights loose electrodes before you start recording.
Local-First & Private	All sessions are recorded locally to transparent CSV and EDF+ files. Raw brainwaves never touch our servers.
Fast & Lightweight	High-performance Canvas visualizer powered by uPlot renders 60+ FPS multi-channel streams with minimal CPU.
Open Science Ready	Compatible with Python, NumPy, Pandas, MNE-Python, EEGLAB, FieldTrip, and PsychoPy.

---

What It Looks Like

MindVault Desktop Interface: Live 256 Hz 4-channel raw EEG, real-time contact quality rings, and spectral decomposition.

Compatible Hardware: InteraXon Muse S Athena (Model MS-03) soft fabric headband with multi-sensor EEG (256 Hz) and optical fNIRS / PPG telemetry.

---

What's New in Muse S Athena & MindVault

Version	Summary
2.1 (Current)	Verified 14-bit EEG packet unpacking algorithm and BrainFlow battery scale fix (/256 vs /512).
2.0	Native macOS Apple Silicon & Windows 11 direct Bluetooth Low Energy stack integration.
1.8	Real-time continuous contact impedance proxy with visual electrode status rings.
1.5	Millisecond-precision CSV logging and standard European Data Format (EDF+) export.
1.0	Initial public protocol specification and standalone Python verification suite.

---

Who Will Like MindVault

• BCI & Neurotech Engineers: Build neural interfaces and brain-computer interaction loops with predictable, low-latency raw streams.
• Cognitive Neuroscience Researchers: Collect reproducible ERP (P300, N170) data without expensive clinical EEG cart setups.
• Sleep & Circadian Trackers: Record overnight frontal EEG sessions directly to EDF+ files for automated sleep staging and spindle analysis.
• Biohackers & Neurofeedback Practitioners: Monitor real-time alpha/theta ratios, frontal asymmetry, and focus states with local-first privacy.
• Students & Educators: Run psychology and neuroscience laboratory demonstrations on any student laptop.
• Open-Source Enthusiasts: Inspect every bit of data unpacking logic, packet layouts, and calibration constants.

---

Quick Start with MindVault

1. Download MindVault Desktop: Get the installer from usemindvault.com.
2. Turn on your Muse S Athena: Press the power button once until the LED indicators light up.
3. Connect in One Click: Open MindVault, select your device from the BLE list, and watch the live EEG stream initialize.
4. Check Electrode Fit: Adjust the headband until all 4 electrode indicators (TP9, AF7, AF8, TP10) turn green.
5. Record Session: Hit Record to log data to CSV or export directly to .edf for MNE-Python analysis.

Standalone Python Quickstart (Developer Spec)

You can also run our standalone, zero-dependency reference decoder directly from this repository:

# Clone this repository
git clone https://github.com/mind-vault-lab/muse-athena-protocol.git
cd muse-athena-protocol

# Run verification unit tests on synthetic Athena packets
python3 athena_decode_spec.py --test

---

Understanding Core Components & Protocol

MindVault is engineered on ground-truth reverse engineering of the Muse S Athena GATT architecture:

[PACKET HEADER: 14 Bytes]
├── [0..3]   pkt_time    (uint32 LE: 256 kHz hardware clock)
├── [4..5]   pkt_index   (uint16 LE: rolling counter)
└── [6..13]  reserved    (firmware metadata)

[PAYLOAD: 1 or more SUBPACKETS]
├── [0]      tag         (uint8: payload identifier)
├── [1..4]   sub_header  (subpacket timestamp & flags)
└── [5..N]   data        (packed bitstream)

1. The 14-Bit EEG Correction (vs 20-Bit Myth)

A widespread misconception claims that Athena transmits 20-bit EEG. Cross-validation across OpenMuse and BrainFlow implementations proves that EEG is 14-bit packed with a scale factor of:

$$\text{EEG Scale} = \frac{1450.0}{16383.0} \approx 0.088506,\mu\text{V/LSB}$$

The 20-bit width belongs strictly to the optical fNIRS/PPG sensors (OPTICS_SCALE = 1.0 / 32768.0).

2. Numerical Calibration Table

Modality	Bit Width	Packing Layout	Scale Factor	Output Unit
EEG (4 Ch)	14-bit unsigned	LSB-first in 28-byte block	1450.0 / 16383.0	Microvolts ($\mu\text{V}$)
Optics (PPG/fNIRS)	20-bit unsigned	LSB-first in 30/40-byte block	1.0 / 32768.0	Raw Intensity
Accelerometer	16-bit signed	2 bytes LE (int16)	6.10352e-5	Gravitational units ($g$)
Gyroscope	16-bit signed	2 bytes LE (int16)	-7.4768e-3	Degrees/second ($^\circ/\text{s}$)
Battery	16-bit unsigned	2 bytes LE (uint16)	1.0 / 256.0	Percentage ($0\text{--}100%$)

Note on Battery Calibration: Tag 0x88 / 0x98 uses a divisor of /256 to convert raw uint16 to percentage. BrainFlow's default channel erroneously applies /512, causing fully charged headbands to report ~50% battery.

3. Connection Handshake Sequence

sequenceDiagram
    autonumber
    participant Host as Host (MindVault Client)
    participant Athena as Muse S Athena (MS-03)
    
    Host->>Athena: Connect BLE (MTU >= 247)
    Host->>Athena: Subscribe to 0x0001 (Control), 0x0013 (EEG), 0x0014 (Telemetry)
    Host->>Athena: Send "v6\n" (Protocol Version)
    Athena-->>Host: Response: Device Version
    Host->>Athena: Send "s\n" (Device Status Query)
    Athena-->>Host: Response: Status String
    Host->>Athena: Send "h\n" (Halt active stream)
    Host->>Athena: Send Preset "p1041\n" (EEG + 16ch Optics)
    Host->>Athena: Send "dc001\n" (Start Transmission)
    Host->>Athena: Send "L1\n" (Enable Low Latency)
    Athena-->>Host: Continuous 256 Hz EEG & 64 Hz Optics Notifications

Loading

---

Electrode Placement & Sensor Montage

MindVault Real-Time Electrode Placement: Live contact quality percentage and signal amplitude for TP9, AF7, AF8, and TP10.

• EEG Montage (10-20 Standard):
  • TP9: Left temporoparietal (mastoid reference area behind ear)
  • AF7: Left anterior frontal (forehead)
  • AF8: Right anterior frontal (forehead)
  • TP10: Right temporoparietal (behind right ear)
  • Reference: FPz (center forehead active ground)
• Frontal Optical Clusters:
  • LO (Left Outer), LI (Left Inner), RI (Right Inner), RO (Right Outer)
  • Multi-wavelength optical paths: 730 nm (Near-IR), 850 nm (IR), 660 nm (Red), and Ambient.

---

Advanced Use Cases & Research Scenarios

Scenario 1: Event-Related Potential (ERP) P300 Oddball

Run auditory or visual oddball paradigms with millisecond timestamp synchronization. The 256 Hz temporal resolution enables robust detection of P300 and N170 waveforms with as few as 40 epochs.

Scenario 2: Overnight Sleep Staging & Slow-Wave Analysis

Track overnight delta (0.5–4 Hz) power and sleep spindles (12–15 Hz) on frontal channels. Export continuous 8-hour sessions to standard EDF+ for automated hypnogram scoring with YASA or Sleep-EDF toolkits.

Scenario 3: Real-Time Neurofeedback & Contemplative States

Compute instantaneous alpha asymmetry (AF7 vs AF8) and frontal theta power to guide meditation neurofeedback protocols, stress management, and mental workload estimation.

Scenario 4: Prefrontal Hemodynamic fNIRS Research

Utilize the 16-channel frontal optical array to study prefrontal cortex activation via the modified Beer-Lambert law (mBLL) during cognitive tasks.

---

System Requirements

Metric	Minimum	Recommended
Operating System	Windows 10, macOS 12 (Monterey), Linux (Ubuntu 20.04+)	Windows 11, macOS 14+ (Apple Silicon), Ubuntu 22.04+
Bluetooth	Bluetooth 4.2 LE	Bluetooth 5.0+ LE
RAM	4 GB	8 GB+
Storage	200 MB for app + 50 MB per 1 hr session	1 GB SSD
Python Interop	Python 3.9+	Python 3.10–3.12 with MNE-Python

---

Comparison: MindVault vs. Alternatives

Feature	MindVault	Mind Monitor	BlueMuse	OpenMuse	Official Muse App
Target OS	Windows, macOS, Linux	iOS / Android	Windows 10/11	Linux / macOS (Python)	iOS / Android
Primary Focus	Research & Live Desktop EEG	Mobile OSC / CSV	LSL Streamer	CLI / Scripting	Consumer Meditation
Hardware Setup	Direct Built-in Bluetooth	Built-in Bluetooth	BLED112 Dongle often needed	Python BLE stack	Phone Bluetooth
Athena (MS-03)	Native Verified Support	Supported	Experimental	Supported	Consumer modes only
Export Formats	CSV + Standard EDF+	CSV + OSC	LSL Stream	LSL + Raw Dump	Proprietary Cloud
Contact Quality	Real-Time Impedance Ring	Numeric Horseshoe	Console Log	OpenGL visualizer	Audio feedback
Price / License	Open Spec + Free App Access	Paid Mobile App	Open-Source	Open-Source	Free + Subscription

---

Academic Citation

If you use MindVault recordings, protocol specifications, or open-source tooling in peer-reviewed research, please cite our project:

@software{mindvault2026athena,
  author = {The MindVault Project Contributors},
  title = {MindVault: Open Reverse-Engineering Specification and Desktop Client for Muse S Athena (MS-03)},
  url = {https://github.com/mind-vault-lab/muse-athena-protocol},
  year = {2026}
}

---

Tags & Discovery Keywords

MindVault Ģ Muse S Athena Ģ MS-03 BLE protocol Ģ open-source EEG client Ģ 256 Hz EEG streaming Ģ Muse without phone Ģ Muse on macOS Ģ Muse on Linux Ģ Muse on Windows Ģ Muse EDF export Ģ MNE-Python Muse Ģ brain-computer interface Ģ raw brainwave recording Ģ dry electrode EEG Ģ Muse fNIRS Ģ Muse PPG heart rate Ģ Muse contact impedance Ģ BlueMuse alternative Ģ Mind Monitor alternative Ģ OpenMuse Athena Ģ sleep staging headband Ģ ERP P300 EEG Ģ neurofeedback software Ģ local-first biosignals Ģ InteraXon reverse engineering Ģ EEG research tools

---

_{Maintained with precision by the MindVault Team • Local-First Biosensing Platform}