The evoked IOS signal appears to start in the frame before the stimulus starts (seen in the ROI vs Time plot, all-frames display...). Not sure if this is a bug in reading of the raw data file, in averaging or analysis functions. Relative timing of frame acquisition and stimulus onsets by LabView seems correct (checked 04/25/13).