question markMuse Direct (iOS+Cloud) FAQ

EEG is often broken down into what are called “frequency band powers”. There are 5 commonly used frequency bands: delta, theta, alpha, beta and gamma. The power of each band is indicative of the strength of oscillations in the corresponding frequency range in the EEG data. For example, delta corresponds to the slowest oscillations in the EEG signal – that is, below 4 cycles per second – and gamma corresponds to much faster oscillations – between 30-100 cycles per second. The division into frequency bands is useful because each band is loosely associated with different mental processes.

Muse Direct provides two different versions of band powers: absolute band powers, and relative band powers. Absolute band powers give an absolute measurement of power of the EEG signal in each frequency band. Relative band powers, on the other hand, are divided by the global power of the signal. Therefore, relative band powers are always between 0 and 1, as they represent a fraction of the total signal.

In addition to offering band powers, Muse Direct also provides band scores. Band scores are computed by comparing the current power of each frequency band to its recent history. Band scores are always between 0 and 1. An alpha score of 1 means that the current alpha power is very high in comparison to the alpha power recorded over the last minute or so. An alpha score of 0, on the other hand, means that the current alpha power is very low in comparison to the recent history. For more information, please consult the documentation here.

The absolute and relative band powers, as well as the band scores, can be used to design your own algorithms based on your own or external research. Band powers are automatically computed and saved when you record data with Muse Direct, and can be streamed with OSC to other programs or computers.

OSC stands for “Open Sound Control”. It is a way of sending data from one computer to another and was originally developed for sound transmission, but it is also useful for transmitting sensor data. OSC streaming allows you to send the Muse data in real-time to another computer for analysis. For example, you can send the data to your desktop computer and use Muse Lab (for Windows and Mac)

To set up OSC streaming for Muse Direct for iOS, the first thing you will need to do is determine the IP address of the computer you want to send the data to. This can be found as the IPv4 address under Advanced Network Setting on Windows 10 or under Network > Advanced > TCP/IP on Mac. An IP address has the form NNN.NNN.NNN.NNN where NNN is a number less than 256.

Once you have the IP address, go to the Streaming screen in Muse Direct for iOS. Tap the IP Address field and enter the IP address of the computer you want to send the data to. Then tap the “Enable OSC Streaming” toggle to turn on OSC streaming.

Muse Direct for Windows will stream the data to the local machine automatically on UDP port 7000. To stream data to a different machine using Muse Direct for Windows, see the online manual.

All recordings are saved in the “Files” screen. To get there, tap on the icon in the upper left hand corner of the screen (the three horizontal lines), and select “Files”. In the Files screen, you can click on the “three dots” icon next to a file to share or send the file where you wish. All files are automatically synced to the Muse Direct cloud as long as WiFi is available.

Go to Muse Direct cloud and login with your Muse account. From there you can download your files in various file formats.

Muse Direct Cloud is a web application that lets you manage and download your Muse files. Files recorded with Muse Direct will be uploaded and stored in your account. You can then convert these to different file formats and download them for running your analysis. Muse Direct Cloud also provides basic statistics on EEG signal quality for each uploaded file.
Files that are recorded with Muse Direct iOS are automatically uploaded to the cloud when your phone is connected to WiFi. You can also directly upload files to the cloud using the “Upload Files” button on the web application.
Muse Direct Cloud supports the following export formats: MUSE, EDF+, FIF (MNE), CSV and JSON.

The European Data Format “plus” (EDF+) is a flexible file format designed to store multichannel signals such as EEG, along with annotations. The EDF+ specifications are available here. EDF+ files can be imported directly into many EEG analysis packages, such as EEGLAB, Brainstorm, FieldTrip and MNE-Python(https://www.martinos.org/mne/stable/index.html). Please note that Muse Direct Cloud’s EDF+ files currently only contain EEG and annotations data.

The FIF file format is the default file format of the MNE-Python toolbox. A FIF file can be directly imported into Python as MNE-Python’s Raw object using the mne.io.read_raw_fif function. Please note that Muse Direct Cloud’s FIF files currently only contain EEG and annotations data.
A Comma-Separated Values (CSV) file is a text file that uses commas to separate values. Muse Direct Cloud’s CSV format contains as many columns as there are data streams in your Muse file, and as many rows as there are data points in your recording. This data format contains the data streams that were recorded in your file, including accelerometer, gyroscope, muse elements outputs (band powers, blink, jaw, etc.), annotations and meta-data. Please note the data available in your file depends on what data you have selected to save in the settings screen. The CSV format can be opened in a standard text or spreadsheet editor.

See example CSV file.

The JavaScript Object Notation (JSON) format is both human-readable and easily parsable by programming languages. Muse Direct Cloud’s JSON format makes it easy to import Muse data into a programming language directly. This data format contains the data streams that were recorded in your file, including accelerometer, gyroscope and muse elements outputs (band powers, blink, jaw, etc.), annotations and meta-data. Please note the data available in your file depends on what data you have selected to save in the settings screen. The data is divided into three main keys: “timeseries”, “annotations” and “meta_data”.
See example JSON file.
Muse transmits the evenly-sampled EEG data over Bluetooth in packets of 12 samples. This means the original recording contains only one timestamp for each packet of 12 successive samples. Therefore, when converting the “.muse” file to the supported formats, Muse Direct Cloud redistributes the timestamps so that they are evenly spaced based on the sampling rate of the Muse model that was used.
Muse Direct is currently also available for Windows (In Beta)

Feedback and feature requests are welcome! Please fill out this survey to have your voice heard!