Setting Some Theoretical Targets
Based on the data in the previous posts, my initial take on the woofage that would deliver on paper (i.e. WinISD) is as follows.
The start point is needing 110dB at 5Hz
A sealed setup rolls off at 12dB/octave so I need 122dB at 10Hz
I get +12dB from room gain from 10 to 5Hz so I'm back to needing 110dB at 10Hz.
WinISD models at 1m but my listening position is 3m so worst case I need to add ~9dB to the model (probably less in reality? but lets be pessimistic)
Total room gain at 10Hz is ~12dB though which more than cancels out the listening distance.
Therefore I need to model for 107dB at 10Hz with 102dB as a fallback option.
Where can I put subs?
I have 2 near field locations available and 1 further away, these locations are marked in red on the pic.
REW room sim says the near field positions will not play nicely with the existing sub location. This suggests that the only way forward is to occupy that alcove completely with subs.
The alcove is ~40"x22" so I think this means an enclosure in the region of 200L is feasible.
The only danger here is that an enclosure that fills the alcove will produce a stronger, early reflection off my R speaker (whether this will impact SQ in an untreated room is another Q, it will certainly be a stronger & earlier reflection than I have now though).
Some pics of common enclosure designs to illustrate
A tube stack
A column stack
Modelling Near Field
The main doubt here is that the room sim is only accurate when I stay away from the alcoves however this is simulating a sub in an alcove so I don't know what will really happen. FWIW though
The sim says that subs in the alcoves really screws with the consistency of the response across the sofa. In the corner alone, there is a variance but it is manageable.
The most interesting Q to me right now is whether my modelling (of my room) is correct. If this is off then any attempt to design for the room will fail.
The 2nd Q is what "clean" really means. I don't have the space to chase 1% distortion at ULF and high SPL, it's just not feasible so I have to compromise. 10% (-20dB) seems like a reasonable compromise based on it being more aggressive than CEA-2010 but not as aggressive as typical HF distortion figures.
Any and all comments much appreciated!