I think in an ideal world you would have 4 subs well placed, but 3 is pretty tough to get right IMO. I disagree that you need more than one sub for good bass response. Certainly exceptional response would require more, but the average person would be fine with a single great sub.
I disagree completely.
A single sub, used the way subs are conventionally used (lowpassed in the ~80Hz range, with the mains highpassed at same) is going to have
more variance in the modal region than "full range" mains, because there are fewer pressure sources exciting room modes. (A single sub with big mains run full-range will be a little better, but still not up to the level of a well-designed multisub system.)
Also, sub quality is not nearly as important as you make it out to be. Yes, one needs enough volume displacement and amplifier power to hit desired SPL targets. And one should use subwoofers with low inductance. But otherwise, sub quality is overrated.
(True, some of us - I won't name names here
- can be snobs about audio parts, and for our reference systems would not consider anything less than Aura's phenomenally-linear underhung NRT drivers.)
Lastly, it is neither that difficult nor that expensive to get right, assuming one picks parts wisely, considering that most of the subs will be lower down in level from the "main" sub - see Geddes, Toole - the auxiliary subs can be smaller and less powerful. (And thus, easier to hide in a room.) I am assuming one is competent in the use of basic audio measurement tools. Let me give you a demonstration of a modest multisub system in a temporary rental flat. Placement is per Geddes: one in a corner, one far away, the third far away from both), and measurement procedure (adding in subs progressively, based on distance from mains) per
Geddes as documented by
Mehlau.
Sub 1 (XLS12 Application Note powered by Dayton SA240), no EQ, 48dB/oct highpass below tuning (~17Hz):
Sub 2 (KEF HTB2, ~25% down the right wall), no EQ, no highpass:
Sub 3 (KEF HTB2, along 70% down the left wall, practically nearfield), no EQ and no highpass:
End result of summing them, just playing with levels and delays on the miniDSP. (i.e.
no EQ)
After Audyssey calibration (DynamicEQ turned off) and three bands of fairly mild EQ cuts, here's the summed, spatially averaged response:
Few have the kind of cash or spouse to invest in a multi-sub system.
Another huge misconception.
Let's take an equivalent version of the system above, made from all off-the-shelf subwoofers. For the main sub, the SVS PB12-NSD will likely offer better maximum output, because it uses a newer development of the same woofer, in a larger box with more power. It costs $750. The smaller subs I used are a little expensive, because they're "designer" pieces. They seem to go for about a grand each new right now. However, one can substitute any sub that can be hidden in one's room and has low enough inductance to extend up to at least ~150Hz. Not to mention the used market.
And point of fact, right now in that same room I have three new subwoofers: a DIY unit with an Aurasound NS12-794-4A in a 65L closed box, an Aurasound NS10-794-4A in a 15L closed box, and a Peerless XLS10 in a 10L closed box.
None of them are the least bit visible. The main sub is hidden in a chest, the NS10 sub is hidden behind a basket panel/kitten-scratch-panel inside a cubical shelving unit, and the XLS10 sub is hidden inside the coffee table. (Which is, of course, to the side and not between listener and front mains.
On-axis can tell you plenty about how bad a speakers is.
But they can't tell you if they are any good. That is to say, a one-point measurement can be used to
exclude, but not
include.
Of course off-axis and waterfall plots help a ton, but a measurement does tell us some things.
Waterfalls are just pretty pictures. They don't add anything not obvious from FR and polar maps...
You do seem to be strongly opposed to Ribbons. We all have preferences. Just realize it's likely your strong preference. I do wonder if a blind test might change the outcome of the tests. I'm sure you've given it a go.
I am no more "opposed" to ribbons than I am to dome tweeters mounted on 180deg waveguides with 6"+ midwoofers...
I haven't done a blind test with ribbons. I have done a blind test recently between two small coincident drivers (KEF HTS3001SE, Tannoy Arena), a mini-monitor with wide but constant midrange directivity (Zaph ZBM4), a cheap high-performance speaker with an 8" woofer and a 1" dome in a waveguide (Berry B2031P), and a conventional 7" "high end" 2-way (Usher Tiny Dancer). (All of them were on subwoofer-stands, crossed at I think 200Hz.) The midrange flaws of the Tiny Dancer were instantly obvious compared to the four better speakers. (Though I found that I much preferred the little KEF eggs to the little Tannoy eggs, which were the only speakers I brought to the comparison. That led me to buy a set of the little KEF eggs shortly thereafter...)
You probably understand more about this stuff than I, but doesn't the implementation effect diffraction and what not?
Physical implementation, yes. (Roundovers, minimal surface protrusions, etc.)
Crossover, no.
I bet there's no way he hears "diffraction" off the mounting screws of the RAAL ribbon in a blind test.[/QUOTE]
Probably not. But diffraction from the chamber and the diffraction slot? Very likely. And the deleterious effect on power response of the crude 180deg waveguide? Unless the midrange is very small (and note that "Mr. Raal" seems to favor omni radiation), absolutely.
