The lab streaming layer for synchronized multimodal recording

Area of Science:

• Neuroscience
• Computer Science
• Biomedical Engineering

Background:

• Synchronized data acquisition from multiple instruments is crucial for accurately recording organismal interactions with their environment.
• Hardware-based synchronization solutions are often costly and impractical for diverse systems.
• Existing methods struggle with independent instrument clocks, network delays, and potential interruptions.

Purpose of the Study:

• To present the Lab Streaming Layer (LSL) as a software-based framework for synchronizing data streams.
• To detail LSL's architecture, core features, and performance in neurophysiological and neurobehavioral applications.
• To highlight practical considerations and limitations of using LSL for reliable data recording.

Main Methods:

• LSL employs per-sample timestamps and network time synchronization for data stream alignment.
• The framework incorporates features for connection recovery, offset correction, and jitter compensation.
• Software-based synchronization is achieved across a local area network (LAN).

Main Results:

• LSL enables continuous, millisecond-precise data recording, even with network interruptions.
• The framework supports over 150 data acquisition device classes and interoperability across multiple programming languages.
• LSL has proven resilient and versatile, supporting diverse applications from basic research to commercial products.

Conclusions:

• LSL provides a robust, zero-configuration software solution for synchronizing multimodal data streams.
• It facilitates the complex task of capturing organismal dynamics and environmental changes on a common timeline.
• LSL is a major platform for multimodal human neurobehavioral recording and real-time analysis.