12The problem was present during two observing runs in the winter 2004 before it was identified and corrected by CFHT. The source of this was a faulty video board which was subsequently replaced.