Calibration Issues

Support
1

Hi Chris,

I believe Simon may have emailed you about this some time ago, but I’m having issues with getting a tight calibration with the calibration protocol. Consistently, the top left, middle, and bottom right calibration points are off by about 2-3 degrees in a consistent direction for each point for every calibration the monkey completes. This has made it so that I can’t lower the fixation window below 5 degrees - otherwise the monkey gets consistent errors and becomes frustrated. All of the other points are spot-on, which leads me to believe the monkey is in fact looking at the points and it is not an issue with the eye tracker.

This issue isn’t fixed by implementing a clear calibration or stopping and restarting the protocol.

I very much need to be able to get a fixation window of 2-3 degrees because I am working in V1. I’m not sure what could be causing this. Any thoughts?

Thank you,
Yvonne

2

Hi Yvonne,

Can you send me an event file from one of your calibration runs? It would be helpful to see the actual numbers that the calibrator is producing.

Also, have you tried using the linear eye calibrator (as opposed to the standard eye calibrator, which uses second-order fit functions)? That might be more appropriate for calibrating digital EyeLink data.

Thanks,
Chris

3

Hi Chris,

Switching to the linear eye calibrator seemed to fix the issue! I wasn’t
aware it might be better for digital EyeLink data. Thank you for suggesting
it.

Best,
Yvonne

4

Hi Yvonne,

Switching to the linear eye calibrator seemed to fix the issue! I wasn’t
aware it might be better for digital EyeLink data.

It was just a guess, based on my assumption that the transformation from EyeLink “pupil” coordinates to MWorks eye coordinates is linear. It’s not clear to me why a second-order fit wouldn’t also work.

In any case, I’m glad that switching solved your problem!

Chris