We have tried to recreate the problem in order to determine whether we can do so with any predictability. It seems that it only occurs after the first run through of the task. Restarting the computer after every participant keeps the error at bay. While this appears to be a short term fix, it obviously is not ideal. The task is set up as an experiment file consisting of 6 .que files. Each questionnaire file contains 20 short videos (4-5 seconds). The error usually doesn't occur on the first .que file but appears sometime between the second and fourth .que files within the experiment. It doesn't appear to be a particular video that causes it either but it does happen across multiple computers.

Any thoughts?

Thanks,

Kevin