It's not 1MHz (in the usual sense), it's 1MS/s. In theory, Nyquist says it should therefore be good to 500kHz in realtime; in practice, consider yourself lucky

It's not 1MHz (in the usual sense), it's 1MS/s. In theory, Nyquist says it should therefore be good to 500kHz in realtime; in practice, consider yourself lucky if that works out. Also consider that if the 'scope has a proper antialiasing filter its rise time will be limited to something in the microsecond region - and if it doesn't have such a filter you've got another problem on your hands. I guess if you've got a repetitive signal and it can do averaging, that would help.