Quote:
Originally Posted by
FOH 
When tested by Ricci, and similar to most any small-sealed, the native response plunges like a rock in the low 30's ... just as does yours. However, when examining your measurements, you can see the native plunge all the sudden take a detour upon the onset of PVG around 24hz or so. Which, doing some simple calculations, 1130/24hz=47.08 You take that and divide it in half, 47.08/2=23.54hz. This is the calculated approx. onset of PVG.
So after the normal small box boosting/LT'ing is added back in, then you'll be fine. Unless you're rocking an IB, or a huge box, you're going to see this drop off.
Back to your room's PVG; when I look at the aggregate graph of all the positioning, I estimated the 24hz point visually. Around this point is the lowest freq of which can propagate freely. Any freq lower than that point, can't fully propagate, hence the pressure vessel gain ... Physically, the longest dimension of the room can no longer support full propagation of the waveform. So somewhere around 23hz-24hz, the normal propagation transitions to pressurization. The wave reproduction changes from a normal cyclic propagation, to pressurization because the wavelengths are too big for the space.
This is key;
the frequency at which this occurs is approximately the point whereby half the wavelength of a given frequency is equal to the rooms longest dimension. So looking back to previous post, seeing the longest room dimension is around 23 feet. That said, and after calculation, the transition from propagation to pressurization will occur around 23hz-24hz. This is because a 24hz waveform is (1130/24)= ~47'. Half of that is 23.5', correlating to the longest dimension of your room. This is the point of transition. Any frequency below that point begins to pressurize the room, and any frequency above that point propagates freely.
So the room gives back acoustically. One of the few examples of free lunch in audio; PVG. This reciprocity is theoretically 12db per octave. The percentage of the 12 db/octave gain one achieves, seems to entirely depend on the integrity of the boundary walls and surfaces. For example, a full tilt concrete bunker should yield a full 12db/octave boost. Typically, it seems as if somewhere between 6-10 db octave may may achievable. Also, in addition to the boundary movement/flexing, other elements can affect the onset point. And anything that consumes a certain measure of cubic feet (furniture, cabinets, etc), may slightly alter the transition merely due to occupying cubic volume.
*
Then there's another component at play, and this is very cool. A distortion lowering mechanism is a by product of the vessel gain. It would seem the acoustic reciprocity elicits a greater proportion of acoustic gain to the deep fundamental freq, moreso than that of the harmonic distortion aspects up higher. In other words, it only makes sense that any fundamental occurring below the PVG onset point, would be subject to the normal PVG, however the harmonics of the fundamental land above the transition, thus they receive much less acoustic gain. Sweet, .. have your cake and eat it too.
Context;
You're going to have plenty, I'm sure of it. It's easy to loose context around these parts. Eight eighteens, yeah, (8)18s ... hell, the vast majority of enthusiasts don't even use more than one sub,....of any size. You'll have plenty. Looking at DataBass' system charts, the SI system equated to the Velodyne DD18
@20hz (which by any technical measure is superb). They're $4999msrp, you can equal it's output with one box. Doubling to two yields 6dB, moving to 4 yields another 6dB, escalating up to 8 sums to another 6dB ... totaling +18dB over that of a single sub. That's a single, high performance 18" high excursion sub ... plus 18dB
Ample headroom and output isn't the question. The real task at hand is the
proper integration of a phenomenal subwoofer system, with those highly capable mains, into the strongly dominant influence that is your room. But you can largely win that one.
Best of luck
edited "@20hz", wrt the Velo DD18 for added clarity
Hi FOH,
Yes, this is the theory that's been used as the explanation for why sealed subs have a rising response below their knee in-room vs GP.
Unfortunately, I have not been able to reconcile it with any posted in-room measurements. In the past few years, with OM and other low cost measurement hardware/software/freeware options, many more in-room graphs have been posted, many of which I've archived against the modeled and GP measured responses.
Carp is seeing room gain begin where virtually everyone else does, regardless of the "longest dimension of the room"... in the 30s Hz. He's seeing an almost identical amount of gain as I (and notnyt, Josh, MKT, etc.) do, around 8dB/octave, beginning in the 30s Hz in rooms ranging from 2200 cubes to 6000 cubes, some open to the rest of the hose and some in an enclosed room..
I've stopped calling the phenomenon PVG and I don't believe subwoofers "pressurize" a room. Instead, they send pressure waves through the room. The gain is related to the phenomenon of progressively constructive reflections. Reflections can be constructive or destructive. As the wavelength increases, there is less chance of them being destructive and they become more and more constructive.
The evidence becomes clear when someone has been kind enough (as Carp has in this thread) to include us in his multiple placement/measuring exercise. If there was a frequency below which the room becomes everywhere pressurized, causing this PVG phenomenon, he would not be seeing 6dB differences within the so-called pressure vessel gain region simply by moving the subs, as he (and everyone else who has engaged in this exercise in his own room) clearly has shown is the case.
I have not been able to match the onset of the wavelength whose length is twice my rooms longest dimension to any measurement I've made with the subs in any location(s) in my room in 10 years. In fact, room gain begins in my room an octave above that theoretical PVG point.
Carp,
If you get the time or inclination, could you run a loopback of the SW out of your AVR? Also, I never understood the wisdom of searching for the flattest response while moving the subs around but excluding your mains. Since your mains are pretty much immovable for the best soundstage, it would be wise to include them in this placement exercise. Otherwise, you may find the perfect sub location and switch on the mains to find a terrible crossover region response.
Once you've got the best location of subs with mains, use the distance delay setting and crossover point selection setting to find the smoothest response through cross... before final EQ smoothing.
Just a suggestion that I've found works best for me, FWIW.
Awesome system, nice space, great thread. Thanks for sharing it all.
Edited by bossobass - 3/12/13 at 7:52pm
