Skip to content

Multiple object tracking with circular trajectories. With Eyelink interfacing, and option to draw the discs using textures to achieve high refresh rates.

License

Notifications You must be signed in to change notification settings

alexholcombe/MOTcircularClean

Repository files navigation

Multiple object tracking experiment

This program is super-complicated, legacy of many different papers using tracking. I don't recommend you use it without my guidance. My other repositories, such as attentional-blink, are more user-friendly.

In the long term, want to set up exchange of objects among rings. Necessary to conduct an identity-tracking experiment. Can also push harder on speed limit that way.

Analysis workflow

  • In the dataPreprocess folder, its Rproj loadAnonymiseSaveData.R is used to match up Psychopy datafiles with EDF files, and possibly combine multiple sessions together.
  • the analysis directory with its Rproj is used to analyze the data.

Eyetracking

Eyetracking data analysis

New way: Use eyelinkReader R package, see notes

Old way: For Fixation Report analysis from EDF files (instructions about making the fixation report here), see this code

Drift correction: will need to be done at the analysis stage because we are not able to re-calibrate the eyetracker due to it not being able to draw to the same graphicsEnvironment even though it's supposed to be able to. So we'd have to close Psychopy grahpics environment and open an eyetracker one for each calibration.

Eyetracking problems Nov 2023 - Jan 2024

On Alex's Mac, it never connects to the eye tracker, like an ethernet problem. Eyelink provides some advice on how to configure the IP address to remedy that, which we reprint here, which I haven't tried.

While Eyelink's example program, picture.py in their EyeLink Developers Kit, works on some machines for Eyelink and Psychopy sharing a graphics environment (so both Eyelink can draw its calibration targets and Psychopy can draw things), we never got that to work with MOTcircular.py, even though it was working years ago - since then, Eyelink has updated their Developers Kit.

So, we instead have the Eyelink quit drawing (pylink. ) after initial calibration, even though it does work on Windows, you just have to tab between windows.

On Joshua's Win laptop, we had some sounds crashing errors. * the sounds crash if I use the old genv way of allowing the Eyelink to draw the calibration targets, but not if I use the pylink.openGraphics()
* Seemed to be fixed after Josh changed the sound encoding of all sound files to be at the same sampling rate.

On Joshua's Win laptop, if oldSchoolWayOfEyelinkDrawingToScreen=False, then Eyelink successfully starts drawing the calibration targets in a separate graphics window you need to tab betwen.

Subsidiary experiments

  • Old versus young

In old repository, separate from this new Clean repo

About

Multiple object tracking with circular trajectories. With Eyelink interfacing, and option to draw the discs using textures to achieve high refresh rates.

Resources

License

Stars

Watchers

Forks

Packages

No packages published