OK, thanks Angela, that's very useful info. If you don't mind, let's try this: Take that copy in which ONLY the problem input file exists (along with its stim and style files/folders). Minimize the input file to say 10 trials or so--just to the point where it's really small/simple but sufficient to produce no data. Eliminate any unnecessary stimuli and make sure this self-contained simplified version still shows this problem. Then send it along to me and I'll run it here, try to replicate the problem, and also see if I can see anything that is unusual about the files or input file structure.