I have what is probably a convoluted way of handling my calibration setup with two systems... or really just two processors, one for two-channel and one for multi-channel. My room is indeed dedicated, so it makes a difference in how I setup the calibration. Everything in the room is pretty much symmetrical. My two front subs are equidistant to the MLP, as are the two rear subs. I use the acoustic reference timing in REW to time align those two sets of subs from front to back, which in essence makes them one sub, and I daisy chain them front to back. Then I use the PEQ's on the subs to fix a couple of minor issues. I measure the combined response of all four subs, since they are mono and all play the same bass at the same time, and should be adjusted likewise. The combined response looks good, so Dirac takes care of the super minor anomalies. I could use the PEQ's in the SHD as well... just whatever is needed. I do have a couple of high-pass filters on my subs at maybe 12-15Hz (can't remember exactly), but it takes care of a little hump I have down really low that I can't get to with the SVS PEQ filters. The HTP-1 also has a host of features and adjustments, including PEQs, shelf filters, etc... very much like the SHD. I also have no limitations to placement of the subs, so they are placed optimally for the best response, therefore not a lot of filters are needed... thankfully.
I use PEQs in the subs because we do a lot of speaker evaluations in my room, and we don't want to always use Dirac, although we might use the subs on some speakers. Having worked for SVS and knowing their engineers, etc, using PEQ before Dirac or Audyssey in no way affects the performance. A lot of guys do this... never heard of anyone complaining about it having any adverse affects... and I've done it for years on top of years.
I have two inputs on my subs. One set is from the HTP-1 and the other set is from the SHD. One or the other is OFF so that only one set of inputs are fed live by whichever processor I'm using. Then as previously mentioned, front subs are daisy chained to the back subs with timing adjusted.
I run my HTP-1 calibration first, which does separate the subs from the mains. This also gives me the correct timing between my mains and subs, which I can plug into the SHD. For example... the mains are delayed by 12.4ms to time align with the subs... so I set the delay on the mains in the SHD to 12.4ms. Then I run Dirac.
The issue with Dirac normalizing the levels in the HTP-1 is that I already have my subs at the proper level to match my mains... which is typically about 10dB higher than the mains (gradual slope / house curve). Dirac removes it when it normalizes the response measurement, and I have to set my target curve so that it boosts the bass back up that 10dB. I can start with the subs level down, but I don't want to do that... as it is perfect for the SHD, and I'd have to boost it there as well. In essence it's not really a boost because all Dirac did was lower the output of the subs. Yet, I still like the way the SHD handles it with the actually level since it is 2.0.