The task that I am running is acquiring both eye data (from EyeLink 1000+ via TCP/IP) and joystick data (via analog input channels on BNC-2090A for X and Y). Looking through the output .h5 file, I noticed that the length of the arrays of eye data and joystick data are different from each other frequently (every 2-3 trials) by 1-2 frames.
When comparing the length of the arrays to the length of the trial (using the timestamp for eventmarker 'End Trial'), the length of the trial is always shorter than the length of the eye data/joystick data by 2-5ms.
I have a few questions that I'd like to ask, to ensure that I'm analyzing properly shifted/adjusted eye and joystick data:
1) Eye data and joystick data could be truncated to match the length of the trial. Do you have an idea about whether those additional eye/joystick samples should be cut from the beginning or end of the array?
2) Do you know why the length of the eye data array is longer than the joystick data on some trials, and shorter on other trials?
The task is attached, as well as an example session output file with 100 trials (converted to a tab-delimited csv in output_data.txt given the size/filetype restriction).
Thanks for your help.