Rigol MSO5000 series - rise time versus bandwidth for non-hacked 'scope

[6SN7WGTB]

Sorry if this is a stupid question.

My Rigol MSO5104 (100MHz and not yet hacked) has a -3dB bandwidth of 146MHz.

I seem able to measure rise times down to ca. 800-1000ps for sharp pulses.

So, the software is limiting the BW to 100MHz nominally, but am I correct that the 350MHz hardware front end can indeed measure rise times within its hardware capability?

I'm judging the 'hardware capability' to be roughly determined by 350E6/0.4 = 1.2ns, or clearly slightly better if it does exceed 350Hz.