Do the samples work for you ok? I took a look at the files you sent to support, and I noticed the time ranges for the probes were a little odd. There should be increasing values because they are cumulative--not relative to the last trial (which is what it looks in your input file). The start time of a probe should never be less than the end time of the previous probe. Otherwise, there is no way to know which probe the subject is responding to. Does that make sense? I'm actually surprised that this approach doesn't send MediaLab into a meltdown. It would make sense to have MediaLab inform you of this automatically.