With the way Dirac works in this device is that there two inputs 'Left' and 'Right' and the correction is applied on the inputs.
If I would do mono sub-bass that means that the subwoofers will be corrected twice, once from the left channel and once from the right channel.
I could also align the phase of each sub to its closest main and do stereo sub-bass, but that means I would use much more headroom for Dirac to achieve the same level of performance of mono sub-bass.
Am I missing something here? Is there a way to get the best of both worlds?
If the bass signals are mostly mono and you set your subwoofers to receive both left and right channels this is effectively a 6 dB boost. If you want the same output level with stereo bass as mono bass just set the output to +6 dB on each channel. In terms of headroom all that matters is what the DSP outputs, I guess you could run in to an issue if your Dirac boosts are limited because you did not have enough output from your subs but you will still be limited by the maximum output capability of the DSP, there is no free lunch.
I am personally a big fan of stereo bass as I have found that it allows for the best integration between the subs and mains. At least for sealed mains / subs I have had good experience applying a HPF equivalent to the subs rolloff to my mains to get a nice phase match.
For example I use a Linkwitz Transform such that my subs have a 30 Hz BW2 (Q = 0.7) rolloff. I then apply a 30 Hz BW2 HPF to my mains.
Here is a simple model of my mains which have roughly a 60 Hz LR2 (Q = 0.5) rolloff, both with the 30 Hz BW2 HPF and without.
I also modeled my sub as a 30 Hz BW2 rolloff with a 60 Hz LR2 LPF, combining with the main without a HPF gives you this. As you can see because this is a LR2 crossover reverse polarity gives the best response. You can also see that the phase match between the sub and main is not perfect.
Combining the sub with the main using a BW2 HPF gives you this. You can see how the phase tracks perfectly in this case and the reverse null is deeper.
Finally comparing the final system output you can see that by high passing the mains you actually get better extension due to phase alignment.
Obviously this is a somewhat simplistic look at what is going on (and completely ignores the room) but hopefully it makes the importance of phase matching easy to see. Linkwitz has a great write up on matching subs to mains which I highly recommend reading,
https://www.linkwitzlab.com/frontiers_5.htm, although he does not mention the trick of high passing the mains for phase matching he presents similar concepts.
Hope this helps.
Michael