By all means ask; I'm out here on the forum to get input and informaton, so I appreciate you taking the time to do so.
I agree that if my NHT offer falls through and I'm left to current drivers, then something has to give. I had already modeled a few prospective subs to 4th order ported designs that would give me a box q of 0.7 in 3.5 cu ft with an f3 of 20Hz. That meets my frequency response criteria, and would require only adding a port to the existing boxes. I've also looked into partitioning my existing enclosures and running a modern sub with eq correction for LF. The third non-NHT alternative is to run my old JL W1 with its f3 of 35Hz and call it good enough, accepting a less than stellar upper 20's response.
It's easy to explain why I'm reluctant to eq. I don't have one, and don't want to add it if I don't have to. Doesn't mean I won't, it's just not my primary option at the moment.
Ported: This one may be in the category of misconceptions you alluded to. I believe a critically damped sealed enclosure sounds better than any of the ported alignments. I'm aware that some disagree (the Exodus app notes, the "subwoofer myths" link on Annunaki's signature), but my beliefs are based on the larger body of what I've read, in combination with my own empirical observations. That doesn't make it so, but I definitely have a bias.
The subwoofer myths article pointed out that the properly tuned port limits cone movement at that tuning frequency much more than a sealed enclosure does, and I buy that. And I have a subsonic filter, so I'm not worried about the loss of damping below that frequency. But what about the next octave above the tuning frequency, where the speaker and port go in phase and cone movement rises ? Could it be that I'm hearing the looser, undamped band near 30Hz on a vented design that's otherwise well controlled above and below that notch? What's your take?
Thank you for your detailed and honest reply.
Now this is where things get difficult. I will take the ported issue first. Ported enclosures seem to satisfy a very large number of people, but clearly they don't entirely satisfy you. Well they don't satisfy me either.
I think it comes down to a matter of total system Qt. It is a very widely held belief indeed that the quality of the system resonance leads to an optimal total system Q of 0.7. That's too high, I think and not natural.
Now we get into the genre of music debate. There is music of natural instruments and music that has no existence without a loudspeaker and everything in between that has varying degrees of dependence on loudspeakers. I would concede that it may be that resonant reproduction has been round so long that it is in many peoples musical memories and part of their expectation. That embellishment by the speaker is part of the intended presentation.
But what about the reproduction of natural instruments and compositions that long proceed the era of reproduced sound?
I have worked with musicians who know the sound of their instruments, recorded them and attend concerts and the opera house regularly. I can be certain a Qt of 0.7 does not give rise to natural reproduction and that Qt 0.5 is much nearer the mark. In other words reproduction systems must be essentially non resonant and that a Qb4 box can never be!
Further the port assists the woofer over a very narrow frequency range and in general has a different quality to the sound emitted from the cone. I'm conscious of that port kick in of ported Qb4 boxes, even such highly regarded speakers as the B & W 200D.
So what are the alternatives.
The sealed enclosures can be made essentially non resonant. However the air volume in the box needs to be small enough to properly react against the cone and control movement, especially as the suspension needs to be highly compliant and the air in the box provide the principle restoring force. The drive motor needs large excursion as only the cone can radiate sound energy. A loudspeaker cone is very inefficient at coupling to the room. The roll off will of necessity in such a system start significantly above the free air resonance of the driver at 12 db per octave. However since the driver does not decouple from the box because of the restoring force of the air in the box, Eq can be applied as long as the motor system allows for sufficient linear travel.
So yes, equalization is an essential part of a well designed sealed alignment. This also demands very large amplifier power for high spl in the last octave. This makes for what I call a brute force approach, but probably the only feasible solution where space is at a premium.
Next get rid of the box! Open backed dipole systems are favored by some. However the cone is never coupled to a box. Because of cancellation loses and reinforcement complicated equalization has to be provided. Large cone excursions are again called for and the only restoring force is the driver suspension. As you would expect distortion is significant, but there is no box resonance. I don't want to suggest that these systems can not be accomplished, but by getting rid of the box you have a whole lot of other issues to confront.
Infinite baffle. The problem here is you need a huge space in your house not needed for much else than to provide a non resonant cavity behind speakers that is large enough to not have a Vb in audio range. It creates the effect of the rear of the cone radiating into free space cut off from the front of the cone. Multiple large drivers are required and again the restoring force is provided entirely by the driver suspension. If you have the right situation and enough of the right drivers, this solution can work very well, but for most it is impractical.
Horns. A horn is the most efficient acoustic transformer of all. Pressure is very high behind the cone, and cone excursion highly controlled. In addition efficient low Qts with high flux density magnet systems can be used. Individuals like Dr Bruce Edgar now have good mathematical models for horn design to minimize coloration that has plagued horn systems. Bass extension is determined by the width of the horn mouth, so for reproduction of the last octave horns are large.
Now to pipes. Pipes were first used to load speakers by Paul Voight seventy or so years ago. The type of pipe used for loading speakers is the closed (one end) pipe, otherwise known as a Gedeckt pipe. The length of the pipe is one half wave length. In addition to the fundamental there will be odd order harmonics generated, no even harmonics. Odd order harmonics are not pleasing to the ear, so at first site this is not a promising system.
However if the driver is placed at a the point of the first node of pressure of the third harmonic, then harmonic generation is to a large degree suppressed.
This point is about 1/4 to 1/3 the pipe length from the close end. Now a pipe has an anti node of pressure at the closed end and a node of displacement at the closed end, a node of pressure and an anti node of displacement at the open end. So there is high pressure at the speaker location, which makes for excellent control of driver excursion.
Now if the pipe is tapered we can broaden the frequency over which the driver is assisted to about 1.5 octaves.
Now an undamped pipe is resonant. It has two peaks of impedance just like a Qb4 box, but port output is high. Fp is obviously determined by pipe length and has to be related to driver F3. Now there is a lot of similarity with a Qb4 box, in that the pipe volume is related to driver VAS. However a pipe can be critically damped without totally suppressing pipe output, although it will be reduced compared to an undamped pipe. The trick is to damp the pipe by stuffing it to the point where you just get one peak of impedance like a sealed enclosure. At that point Qt is around 0.5 roll off 12 db per octave within driver excursion limits.
So where is the snag. It is size. A transmission line pipe designed to reproduce the last octave is a large structure taking up a lot of real estate, so we return to practicality issues. You can see this from my pictures in my signature. Those speakers are dual lines tuned about one half octave apart, resulting in very smooth driver support from 20 to 100 Hz.
For me the result is entirely worth it to have typms, bass strings, bassoons, piano and organ pipes sound just as they should.
Next large membrane speakers. For very large membrane systems this essentially means electrostatic loudspeakers. To reproduce the last octave requires a very large panel indeed and again we are back to practicality issues.
So you can see the wisdom in the old adage: - "Does a speaker have to be large? No but it really helps."
I hope this length response will assist you in your decision of how you want to take your poison.