[...] No one ever seems to mention that Harmon’s experiments and theories on multiple subs were derived in shoebox rooms. It’s good and fine if you happen to have your set-up in a room like that, but it’s not of much use for the rest of us. [...]
This is not completely true, as Harman's research gives a good idea of what is possible in most rooms. They have covered limitations of multiple subwoofer integration, as well.
For instance, I have a ~80 m² cross-gable roof shaped room - far off from Harman's "shoebox rooms". The room is not (yet) treated with bass-traps and unfortunately, I only have 2 fixed subwoofer positions available: right and left of and in-between the main loudspeakers. Those are really suboptimal conditions!
However, using 2 subwoofers instead of one gives me much more flexibility on the application of EQ [besides this, a higher combined SPL at lower THD]. More precisely,
Multi-Sub Optimizer (MSO) calculated two filters, one for each subwoofer and additionally a (+)delay block for one, and a (-)gain block for the other subwoofer. Something that clearly cannot be achieved with a single subwoofer. The result is pretty astonishing: With 2 subwoofers which, by the disadvantage of fixed positions in quite close proximity to each other, do excite very similar room modes; MSO was able to smoothen the frequency response significantly
over 5 listening positions = an entire couch! The frequency range of my interest was from 15 Hz to 120 Hz for the room corrections, following a slightly modified Harman target response curve. Other calculations showed, that I could smoothen the frequency response down to 10 Hz but this would rise distortion levels; and I prefer less THD in addition to concentrating on the audible range [
which would start at ~16 Hz for "younger" people with an undamaged sense of hearing]. Visceral impact is more than enough for my taste; so even in movies, I do not miss the additional -5 Hz; to reach 10 Hz "flat".
Without any doubt, the results - even with a single subwoofer - will be of another level, once proper acoustic room treatment has been put in place!
====
On the following MSO graphs,
no smoothing has been applied. The last two graphs, REW distortion measurement, have a fixed
1/24 octave smoothing applied:
..
[ EDIT
]
I added a distortion measurement, which was taken after the calculated MSO tunings had been applied to both the 'dual-opposed 15-inch drivers, sealed' subwoofers. The highest distortion [
< 0.7%] shows, where the big 40 Hz room mode had to be corrected for. [
Dirac Live room correction will be applied on top of that; for the whole multi-channel setup.]
- miniDSP UMIK-1 (factory calibrated: 10 Hz - 20 kHz)
- ~10 feet [~3 meters] listening distance, at the MLP
- REW; single sweep @ 90 dB SPL / theoretical equivalent of 99.54 dB SPL @ 3.28 feet [= 1 meter]
Enjoy!