Freezing, using movie files


My program chronically freezes after a few trials. A typical run proceeds as follows:

Event 1) Sentence appears, response = hit space bar (96 sentences in stimulus list)
Event 2) Movie file (.avi, ~ 10 MB), followed by a text field, response = enter a number (8 videos in stimulus list)
Event 3) If the number is incorrect, “INCORRECT” appears on the screen (no response)
Event 4) Sentence appears, response = hit N or Y (96 sentences in this stimulus list)

The program contains about 20 blocks, each of which are linked to one trial (1-20), which is linked to the four relevant events (one for each trial; the sentences had to be presented in random order). Each participant sees only one block, but there are 40 text events (192 sentences altogether) and 1 movie file event (with the 8 videos).

The system: Windows XP, Pentium 4, and from what I can remember, decent memory and hard drive space.

At first I thought the movie files were too large. However, I replaced the video files with much smaller versions (~ 300 KB, not experiment quality though), and the program continued to freeze up (but less frequently). Might I have too many events?

Also, after clicking RUN and entering the participant name (etc), it takes 1-5 minutes for the SAVE dialog box to appear.


Footnote: I should mention that, perhaps 1 out of 5 times, the program runs fine and does not freeze. I’m fairly certain the problem is not related to keyboard input, for instance. Also, the point at which it freezes varies.

SuperLab on Windows has difficulty with loading/presenting multiple movies at once. The precise limitation depends on the video card. I’m not certain that this is what’s going on in your case, but it sounds like you’re only presenting one movie at a time. You may run into this limitation if you tell SuperLab to load all stimulus files before the start of the experiment. This is in the Experiment Options dialog. Try telling SuperLab to load the stimulus files on a per-trial basis (which is the default).

If you’re already loading stimulus files on a per-trial basis, then we’re probably running into a different issue.