I do have Dirac. I thought that we should not allow Room Eq to correct such dips.
I use DSP and it works great for my purposes, CamillaDSP in my office and a miniDSP HTx in the family room.
When using PEQs, I take steps to prevent clipping of the low level signals going to the amplifiers.
In my family room I have multiple preset configurations. The preset for broad coverage in the room uses Dirac Live without PEQs in the bass region. The preset for when I am seated in my listening chair does not use Dirac Live, but instead exclusively uses PEQs, with numerous PEQs set for the bass region.
When tuning with the PEQs, I set the subwoofer gain in my HTx to somewhere around -10 dB. The subwoofer volume knob is in the 4/10 position. Before EQ the largest bass dip is about 6 dB below my target curve, but on average the bass is well above the target curve. Because the gain in the HTx is -10 dB and at most I am adding 6 dB with PEQ to get a flat bass response, the net gain in the digital realm is less than 0 dB. So no clipping occurs. I have a Velodyne HGS-18 subwoofer, and it is not even breaking a sweat at the loudest volume levels to which I listen. The THD is well below 1% down to 30 Hz, and not too bad below that (it is hard to tell for sure because the walls, and things on the walls, are shaking so much that they are contributing to the THD measurement).
For my office, while I am equalizing I monitor CamillaDSP for clipping. If clipping occurs, I reduce the gain in ALSA. I currently have the master gain set to 96 and the capture gain set to 92. That is adequate to prevent any clipping of the low level signals. Of course, I need to add more gain to the speakers to get the same volume level, but in my office I never play my speakers anywhere near loud enough to push them to their limit.