syelink

Parse EyeLink eye tracker data from ASC files. For full documentation, see syelink.readthedocs.io.

Features

• Parse EyeLink ASC files into structured JSON
• Extract calibration, validation, recording, and gaze sample data
• Export gaze samples to CSV with optional raw pupil/CR data
• Visualize calibration and validation results
• Command-line interface for common tasks
• Support for binocular and monocular (left/right only) recordings

Installation

uv pip install syelink

Or install from source:

git clone https://github.com/mh-salari/syelink.git
cd syelink
uv pip install -e .

Quick Start

Convert an ASC file to JSON, text files, and CSV

uv run syelink convert data.asc

This creates:

• data.json - All session data (calibration, validation, recordings, gaze samples)
• data_samples.csv - Gaze samples with timestamps, positions, pupil sizes, and optional raw data
• Human-readable text files: recordings.txt, calibrations.txt, validations.txt, metadata.txt

To export only specific formats:

uv run syelink convert data.asc --json      # JSON only
uv run syelink convert data.asc --text      # Text files only
uv run syelink convert data.asc --samples   # CSV samples only

View session information

uv run syelink info data.json

Shows calibration/validation count, display info, and gaze sample statistics.

Export gaze samples to CSV

uv run syelink export-samples data.asc -o samples.csv

Generate validation plot

uv run syelink plot-validation data.json -i 0 -o validation.png

Generate calibration plot

uv run syelink plot-calibration data.json -i 0 -o calibration.png

CLI Commands

Options

convert

• -o, --output - Output directory (default: same directory as ASC file)
• --json - Export JSON file only
• --text - Export text files only
• --samples - Export gaze samples CSV only
• Without flags: exports all formats (JSON + text + CSV)

export-samples

• -o, --output - Output CSV file path (default: <filename>_samples.csv)

info

• Shows calibration/validation counts, display info, and gaze sample statistics

plot-validation / plot-calibration

• -i, --index - Calibration/validation index (default: 0)
• -o, --output - Output image path
• --show - Show plot interactively
• --target-image - Custom target image (validation only)

Python API

from syelink import parse_asc_file, SessionData

# Parse ASC file
session = parse_asc_file("data.asc")

# Access data
print(f"Display: {session.display_coords.width}x{session.display_coords.height}")
print(f"Calibrations: {len(session.calibrations)}")
print(f"Validations: {len(session.validations)}")
print(f"Gaze samples: {len(session.gaze_samples):,}")

# Save to JSON
session.save_json("data.json")

# Save gaze samples to CSV
session.save_samples_csv("gaze_samples.csv")

# Load from JSON
session = SessionData.load_json("data.json")

# Access validation errors
for val in session.validations:
    if val.summary_left:
        print(f"Left eye avg error: {val.summary_left.error_avg_deg:.2f}°")

# Access gaze samples
for sample in session.gaze_samples[:10]:  # First 10 samples
    print(f"Time: {sample.timestamp}, Left: ({sample.left_gaze_x}, {sample.left_gaze_y})")

Data Structure

The parsed data includes:

• Display coordinates - Screen resolution and boundaries
• Calibrations - Calibration points, polynomial coefficients, gains, results
• Validations - Target positions, gaze offsets, error metrics
• Recordings - Start/end timestamps for recording blocks
• Gaze samples - Timestamps, gaze positions (x, y), pupil sizes, and optional raw pupil/CR data
  • Supports binocular (both eyes) and monocular (left or right only) recordings
  • CSV export includes 32 columns: metadata, gaze data, and raw camera measurements

Examples

Check the examples/ directory for complete usage examples:

• basic_usage.py - Parse ASC files, save to JSON/CSV/text, and load data
• plot_example.py - Generate calibration and validation plots

Run examples with:

cd examples
uv run python basic_usage.py data/both_eyes/both_eyes.asc
uv run python plot_example.py data/both_eyes/parsed_output.json

Example data includes:

• both_eyes/ - Binocular tracking data
• left_eye/ - Left eye only tracking data
• right_eye/ - Right eye only tracking data

Citation

Cite as:

Salari, M., Nyström, M., Niehorster, D. C., & Bednarik, R. (2026).
PyeLink and SyeLink: Open-source Python tools for low-level EyeLink experiment control and data parsing.
In Proceedings of the 2026 Eye Tracking Research & Applications (ETRA 2026) Late-Breaking Work. ACM.
DOI: 10.1145/3797246.3805844

License

MIT

Acknowledgments

This project has received funding from the European Union's Horizon Europe research and innovation funding program under grant agreement No 101072410, Eyes4ICU project.