On seeing temporal gaps between gratings: a criterion problem for measurement of visible persistence.

Previous estimates of visible persistence based on subjects' ability to see a "blank" between brief flashes of a grating are remarkably long when compared with studies of temporal integration. We show firstly that subjects can detect gaps as short as 5 msec at low spatial frequencies, rising to about 20 msec at high spatial frequencies, and secondly that artefactual luminance transients such as might be found in tachistoscopes do not greatly increase the detection thresholds. Much longer estimates were obtained through the use of a more conservative criterion, which was subjectively ill-defined and very variable across subjects. Studies which use this method may not measure visible persistence very effectively.