I'm surprised at the Agilent... that's ridiculous. Using something like your test waveform (100 kHz, 10V pk-pk, DC offset = 6V), all three DSOs within reach did reasonably well. On my MSO6054A (the one I will take with me to Hell to hack the HVAC controller), the base of the waveform at ~1V was visible and accurate down to 500 mV/div, but could not be brought back onscreen at 200 mV/div. It actually looked good down to 200 mV/div on a DSO3062A (rebadged Rigol), and likewise for the old Tek 2430A.
No complaints, they performed about as I'd expected... except that I figured the $200 Rigol would have dropped the ball as soon as the waveform went offscreen. I don't expect commodity DSOs to ship with
7A13 front ends, so that TDS3032B is pretty impressive.-- john, KE5FX