Did a single trace of 1 sec/div with about 4,000 cycles of waveform, then stopped it and zoomed in and panned about. Sometimes (always in a given session) what was known to be a 1mS wide pulse would actually show as 200us on the screen and also by the cursors. Zoom back fully out and an indicated 1 sec/div came up with a short trace, not the full screen width. Scope was calibrated ~25th August 2013 IIRC, can't get to details right now.