Skip to main content

BrainStreamingLayer real-time framework for online neuroscience research through LSL-compatible devices.

Project description

Brain Streaming Layer

BrainStreamingLayer provides a real-time brain signal streaming framework. BSL is a wrapper around the python interface to the Lab Streaming Layer (LSL). BSL goal is to simplify the design of a study using the Lab Streaming Layer which provides sub-millisecond time synchronization accuracy.

Any signal acquisition system supported by native LSL or OpenVibe is also supported by BSL. Since the data communication is based on TCP, signals can be transmitted wirelessly. For more information about LSL, please visit the LSL github.

BSL is based on NeuroDecode. The original version developped by Kyuhwa Lee was recognised at Microsoft Brain Signal Decoding competition with the First Prize Award (2016) after achieving high decoding accuracy. BSL is based on the refactor version by Arnaud Desvachez for the Fondation Campus Biotech Geneva (FCBG). The low-level functionnalities have been reworked and improved, while the decoding functionnalities have been dropped.


BSL supports python >= 3.6 and requires:

  • numpy
  • scipy
  • pylsl
  • mne
  • pyqt5
  • pyqtgraph

Optional dependencies for trigger via a parallel port (LPT):

BSL can be installed in normal mode with python install or in developement mode with python develop. Optional dependencies can be installed using the keywords:

  • parallel, parallelport, lpt
  • doc, doc-build, documentation
  • test, testing


Documentation website.

BSL is centered around 4 main modules: stream_receiver, stream_recorder, stream_player and stream_viewer.


The stream receiver connects to one or more LSL streams and acquires data from those. Supported streams are:

  • EEG
  • Markers


from bsl import StreamReceiver

# Connects to all available streams
sr = StreamReceiver(bufsize=1, winsize=1, stream_name=None)
# Update each stream buffer with new data
# Retrieve buffer/window for the stream named 'StreamPlayer'
data, timestamps = sr.get_window(stream_name='StreamPlayer')

The data and its timestamps are returned as numpy array:

  • data.shape = (samples, channels)
  • timestamps.shape = (samples, )

The data can be returned as an MNE raw instance if return_raw is set to True.


The stream recorder connects to one or more LSL streams and periodically acquires data from those until stopped, and then saves the acquired data to disk in pickle .pcl and in FIF .fif format.


import time
from bsl import StreamRecorder

# Connects to all available streams
recorder = StreamRecorder(record_dir=None, fname=None, stream_name=None,

When the argument record_dir is set to None, the current folder obtained with pathlib.Path.cwd() is used. When the argument fname is set to None, the created files' stem use the start datetime.

CLI: The stream recorder can be called by command-line in a terminal by using either bsl stream_recorder or bsl_stream_recorder followed by the optional arguments -d, -f, -s respectively for record_dir, fname, and stream_name, and the optional flags --fif_subdir and --verbose.

bsl_stream_recorder -d "D:/Data"
bsl_stream_recorder -d "D:/Data" -f test
bsl_stream_recorder -d "D:/Data" -f test -s openvibeSignals


The stream player loads a previously recorded .fif file and creates a LSL server streaming data from this file. The stream player can be used to test code with a fake LSL data stream.


import time
from bsl import StreamPlayer

sp = StreamPlayer(stream_name='StreamPlayer', fif_file=r'path to .fif')

CLI: The stream player can be called by command-line in a terminal by using either bsl stream_player or bsl_stream_player followed by positional arguments stream_name and fif_file and the optional arguments -r, -c, -t respectively for repeat, chunk_size and trigger_def, and the optional flag --high_resolution.

bsl_stream_player StreamPlayer data-raw.fif
bsl_stream_player StreamPlayer data-raw.fif -c 16
bsl_stream_player StreamPlayer data-raw.fif -c 16 -t triggerdef.ini


The stream viewer creates a 2-window GUI composed of a control GUI and a plotter GUI to display the data acquired from an LSL server in real-time.

CLI: The stream viewer can be called by command-line in a terminal by using either bsl stream_viewer or bsl_stream_viewer followed by the optional argument -s for the stream_name. If no stream name is provided, a prompt will ask the user to select the desired non-marker stream to display.

bsl_stream_viewer -s StreamPlayer


Triggers includes functions to mark time event by sending a trigger which will be saved on the TRIGGER channel of the on-going recording. Triggers can be achieved either through hardware or through software.

Currently, the supported hardware triggers use an LPT port.


import time
from bsl import StreamRecorder
from bsl.triggers import SoftwareTrigger
from bsl.triggers import ParallelPortTrigger

# Software trigger
recorder = StreamRecorder()
trigger = Softwaretrigger(recorder)
for k in range(1, 5):

# Hardware trigger through Arduino LPT converter
recorder = StreamRecorder()
trigger = ParallelPortTrigger(address='arduino')
for k in range(1, 5):

Note that closing the trigger before stopping the recording may not be required for all kind of triggers.

Copyright and license

The codes are released under GNU Lesser General Public License.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

bsl-0.4.0.tar.gz (1.8 MB view hashes)

Uploaded source

Built Distribution

bsl-0.4.0-py3-none-any.whl (1.8 MB view hashes)

Uploaded py3

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring Facebook / Instagram Facebook / Instagram PSF Sponsor Fastly Fastly CDN Google Google Object Storage and Download Analytics Huawei Huawei PSF Sponsor Microsoft Microsoft PSF Sponsor NVIDIA NVIDIA PSF Sponsor Pingdom Pingdom Monitoring Salesforce Salesforce PSF Sponsor Sentry Sentry Error logging StatusPage StatusPage Status page