readme.md

Tangle Eye tracking test

Setup

make test will install lncdtasks with pip in a venv and exercise ./eyetest.py.

Run like

# once to setup and confirm working
git clone https://github.com/NPACore/tangle-eye
cd tangle-eye
make test

. .venv/activate/bin # per session to use locally-installed python libraries
./eyetest.py         # to get the task dialog box

Use tracker = "None" as in the screenshot to demo without an eye tracker

Eye Data

An edf2asc converted eyetracking file is in example/wf_20241126.asc

Retrive with git lfs pull. Plotting with read.R

Flip images benchmark

Is it better to pre-render and flip one giant texture or generate display on the fly? For real-time performance, it doesn't matter. See imgflip_benchmark.py. Flip times are still limited by refresh rate (2 win.flips() at 60Hz = 0.033s).

flipped 10 images in 0.033s per iteration one large image flipped 0.033s per iteration

Resources

For more eye tracking specific interface see

• pylink_help.py for lncdtask code
• more info for use at the MRRC/scanner
• free-regirstation required SDK download on the EyeLink forums
• SR Research EyeLink in Psychopy YouTube video