Eye Tracking the Main Sequence (Sun/Goldberg Lab)

This is a Psychopy2 builder file (.psyexp extension) for use with standalone psychopy and should work in Windows or Mac OS X. To make it work a few monitor settings in Psychopy needs to be set first:

Settings for Psychopy Preferences:

In the builder view, click on the 7th icon:
(crossing monkey wrench and screwdriver)

In the General Tab
	window type = pyglet
	units = deg
	audio Library should read: [‘sounddevice’,’pyo’,’pygame’]
	audio Driver should read: [‘Primary Sound’,’ASIO’,’Audigy’]
	audioDevice should be Primary Sound Driver or Output speaker on Mac

In the Builder view, click on 8th icon

Monitor Center (More instructions to be updated later)

In the very first routine (StartupInstr), in the Begin Experiment tab there is a call to execfile script “IDMonLaunch_ioHub.py” where various monitor settings are stored, and settings for using a mouse or eye tracker are hard coded for each PC/Laptop setup. In this tab, there is a final call to open a window full screen, thus the resolution may or may not matter. The reporting of the mouse units since it is being queried from the iohub process, should be determined by the SRR_eyelink_std.yaml configuration file if eyetracker is set to TRUE. If eyetracker == False then the older launchHubServer function runs.

The rest of the builder file (MainSequenceBeta1.psyexp file) is the experiment logic.

How to turn eye position/mouse position rendering on and off:

This basic configuration can be found in the routine StartupInstr
In the Begin Experiment Tab (for displaying window boxes or eye position rendering)

Make sure to set shapes to degFlatPos And in experiment settings tab too (blue screen with white double ended arrows in it-tab 3-4)

Version History 1.1 Added RoutineSpeedUp variable in StartupInstr Custom Code to speed up testing. Should be set to 1 for actual experiments.

====================================== Programmer notes:

2/4/18: About adding custom Code components: http://www.psychopy.org/builder/components/code.html

Components in general including documentation about timing: http://www.psychopy.org/builder/components.html

Various calls you can make to manipulate a shape: http://www.psychopy.org/api/visual/shapestim.html#psychopy.visual.ShapeStim

When activating the subroutine that captures eye movements

io=iohubConnection() reports in Degrees (good!), the position of the eye rendering and units are in degrees which matches with the experiment logic, even though pixels could be used and may be more accurate, the angle in degrees is now consistent in every part of the code so control of eye position matches.

The reduced simpler io=ioHubServer() seems to report mouse position in PIXELS ONLY, which leads to a mismatch between the mouse position and the rendered eye position as expected. The mouse.getPosition fn returns in pixels which have to be divided by a number (17 or 33 depending on the platform/PC) so that the mouse position is approximate to the eye rendered in degrees.

event.mouse.getPos() bugged if use pyglet for display, pygame doesn’t work either, stick with pyglet and submit bug report for line 558 in: File “/Applications/PsychoPy2.app/Contents/Resources/lib/python2.7/psychopy/event.py”, line 558, in getPos lastPosPix = lastPosPix - self.win.size / 2

====================================== Known Bugs: on Win 10 PC, if mouse does not respond, restart script or restart psychopy (probably iohub process not closed fully)

Ryu’s Screen Screen Width 30.5 cm Screen height 17cm H 1366 V 768 ‘DESKTOP-HL8BRQT’

Alex’s Linus Amritas MBP13.3” = 28.65H 17.9cmV H 2560 V 1600 ‘Amalias-macbook-pro’ ‘Liner-MBP’ ‘Alexanders-Macbook-Pro’

Kirstie 28x15.5cm H 2304 V 1440 ‘Kirsties-Macbook’

Zikang Screen Width h 2736 v 1824 screen width 26cm ‘SurfacePro’

Arya - MBP 15 33.1 x 20.78 h 2880 v 1800 ‘Aryas-MBP.local’