Adding screen distance to raw data log

Marsh

Hi developers of PyGaze,

Thank you for the great work!

I am using a Tobii Spectrum 600 Hz with Tobii PRO SDK (TobiiPro.SDK.Python.Windows_1.7.0.3) to collect task-contingent eyetracking data through PyGaze implementation in Open Sesame software. In my raw data log, data is lacking about the distance between the eyes and the screen (screen distance), while I need this information to analyse oculomotor behavior (e.g. with GazePath).

How can I change the libtobii.py file in PyGaze in order to add the screen distance for both eye seperately to the raw data output? Or would it be better to calculate distance post-hoc after data collection?

Also see the attached script that I am using.

Thanks again and kind regards,

Marsh

eduard

Hi Marsh,

screen distance is something that should be fixed during the recording, no? Or do you have some sort of mobile eye tracking setup? If it is a fixed setup, the screen distance shouldn't change, so then, you could manually send a log message to the eye tracker with the screen distance. If it is easy to add screen distance posthoc, that is of course also a good option (given that you just have to add a single value). Logging with pygaze works with the log_var command: http://www.pygaze.org/documentation/eyetracker/#EyeTracker.log_var

Hope this helps,

eduard

Marsh

Dear Eduard,

Thank you for you reply!

The eye-tracker is fixed to the screen, but we cannot use a head mount (we are studying babies and infants). The distance is determined during calibration and logged. However, the kids can wiggle around with their heads, slightly changing the distance to the screen. Therefore, I consider that it may be wise to calculate the distance as a continuous variable to go along with the gaze data during data processing, or do you disagree?

All the best,

Marsh

eduard

Hi Marsh,

No, sure if you can take the distance into account, but do you have such a measure? I am not familiar with that specific eye tracker, but if the kids move enough to make the change in distance really matter for dva, I would expect that the eye tracking itself becomes problematic, signal loss, etc.

Also, I don't think there is an eye-tracker specific way to measure screen distance, as this is usually a constant. You would need to make this measurement yourself (with some sender/receiver setup), and merge that information with the eye data (probably post-hoc). Not sure that's worth the trouble....

Eduard