Data Format Description

Units of Measure

For EXG channels BrainFlow returns uV wherever possible.

For timestamps BrainFlow uses UNIX timestamp, this counter starts at the Unix Epoch on January 1st, 1970 at UTC. Precision is microsecond, but for some boards timestamps are generated on PC side as soon as packages were received.

You can compare BrainFlow’s timestamp with time returned by code like this:

import time
print (time.time ())

BrainFlow Presets

Each board can accumulate data inside one of three predefined buffers. In BrainFlow they are called presets, possible values are:

  • BrainFlowPresets.DEFAULT_PRESET

  • BrainFlowPresets.AUXILIARY_PRESET

  • BrainFlowPresets.ANCILLARY_PRESET

Each of them has it’s own sampling rate, timestamp and package id. Other data types depend on exact device.

For almost all devices only BrainFlowPresets.DEFAULT_PRESET is available. But for some devices, especially if they stream different types of data with different sampling rates(e.g. Muse) we store these data types in different presets. All methods like:

# etc

Have an optional preset parameter with default value BrainFlowPresets.DEFAULT_PRESET if programming language supports default values for function arguments. Here is a code sample that you can use as a referece:

using BrainFlow


params = BrainFlowInputParams()
board_shim = BrainFlow.BoardShim(BrainFlow.MUSE_S_BOARD, params)

BrainFlow.config_board("p50", board_shim) # to enable ppg use p61, p50 enables aux(5th eeg) channel and smth else
BrainFlow.add_streamer("file://default_from_streamer.csv:w", board_shim, BrainFlow.DEFAULT_PRESET)
BrainFlow.add_streamer("file://aux_from_streamer.csv:w", board_shim, BrainFlow.AUXILIARY_PRESET)
BrainFlow.add_streamer("file://anc_from_streamer.csv:w", board_shim, BrainFlow.ANCILLARY_PRESET)
data_default = BrainFlow.get_board_data(board_shim, BrainFlow.DEFAULT_PRESET) # contains eeg data
data_aux = BrainFlow.get_board_data(board_shim, BrainFlow.AUXILIARY_PRESET) # contains accel and gyro data
data_anc = BrainFlow.get_board_data(board_shim, BrainFlow.ANCILLARY_PRESET) # contains ppg data

BrainFlow.write_file(data_default, "default.csv", "w")
BrainFlow.write_file(data_aux, "aux.csv", "w")
BrainFlow.write_file(data_anc, "anc.csv", "w")

You can get all presets availabe for your device using BoardShim.get_board_presets(board_id) method.

Generic Format Description

Methods like:

get_board_data ()
get_current_board_data (max_num_packages)

Return 2d double array [num_channels x num_data_points], rows of this array represent different channels like EEG channels, EMG channels, Accel channels, Timesteps and so on, while columns in this array represent actual packages from a board.

Exact format for this array is board specific. To keep the API uniform, we have methods like:

# these methods return an array of rows in this 2d array containing eeg\emg\ecg\accel data
get_eeg_channels (board_id)
get_emg_channels (board_id)
get_ecg_channels (board_id)
get_accel_channels (board_id)
# and so on, check docs for full list
# also we have methods to get sampling rate from board id, get number of timestamp channel and others
get_sampling_rate (board_id)
get_timestamp_channel (board_id)
# and so on

For some boards like OpenBCI Cyton, OpenBCI Ganglion, etc we cannot separate EMG, EEG, EDA and ECG and in this case we return exactly the same array for all these methods but for some devices EMG and EEG channels will be different.

If board has no such data these methods throw an exception with UNSUPPORTED_BOARD_ERROR exit code.

Using the methods above, you can write completely board agnostic code and switch boards using a single parameter! Even if you have only one board using these methods you can easily switch to dummy BrainFlow boards and it will help you during developemnent and testing.

Getting All Info About Device And Supported Channels

There is a method get_board_descr(int board_id). You can use it to get all info about specified device and BrainFlow channels for this board.

from pprint import pprint

import brainflow
from brainflow.board_shim import BoardShim, BoardIds

board_id = BoardIds.SYNTHETIC_BOARD.value
{"accel_channels": [17, 18, 19],
 "battery_channel": 29,
 "ecg_channels": [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16],
 "eda_channels": [23],
 "eeg_channels": [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16],
 "eeg_names": "Fz,C3,Cz,C4,Pz,PO7,Oz,PO8,F5,F7,F3,F1,F2,F4,F6,F8",
 "emg_channels": [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16],
 "eog_channels": [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16],
 "gyro_channels": [20, 21, 22],
 "marker_channel": 31,
 "name": "Synthetic",
 "num_rows": 32,
 "package_num_channel": 0,
 "ppg_channels": [24, 25],
 "resistance_channels": [27, 28],
 "sampling_rate": 250,
 "temperature_channels": [26],
 "timestamp_channel": 30}

Other Channels

Some boards have pretty unique data types and we do not have dedicated methods for them, for such devices we return data in get_other_channels(). Please refer to the source code to get more info about it.