takeaim:
Although some people prefer 2' to 3' deep porous (fiberglass/rockwool/cotton/polyester) absorbtion (e.g. Paul Woodlock, Mark Edmonds, etc) to absorb bass, I believe that it's more common to use resonant absorbers to absorb bass. You've probably read about Dennis Erskine using them in columns and soffits and risers. Partially because it takes up less floor space when combined with 1" porous around the room. The disadvantage of resonant absorbers is that they shake the whole wall at that frequency, possibly making that frequency transmit through the wall reducing soundproofing.
"To get broadband passive absorption across the frequencies of most interest to human acoustic design, usually requires a combination of resonant and porous absorption." (Trevor Cox in Room Acoustics 2004-5)
There are several different types of resonant (non porous) absorbers: membrane, slat helmholtz, perferated helmholtz, single port helmholtz. (beware there are lots of wrong Slat Helmholtz formulas on the www -- the correct one can be found here
http://forum.studiotips.com/viewtopic.php?t=94 )
Anyway, on to membrane formula:
A example membrane trap would be an airtight box with a 4'x8'x3/4" sheet of plywood on a 2x4 frame about 10" deep, with more plywood on the sides and back of the frame, with fiberglass/rockwool inside, not touching the front membrane. You can put horizontal bracing from front to back (the 10" dimension) near the sides but not under the membrane (the front 4'x8' panel).
http://www.bobgolds.com/BassTrap_JeffCooper_pg127.jpg
Membrane trap formula: d = 28900 / (M * f^2)
d = depth of airspace in inches
M = surface density of panel, lb / ft^2
f = peak absorbing frequency
M for 1/2" plywood is about 1.375 lb / ft^2
M for 3/4" plywood is about
2.06 lb / ft^2
Let's try to design one for 38hz.
d = 28900 / (
2.06 * (38^2)
d = 9.7"
Umm handy. We can make the sides out of 2x10's with that dimension, instead of 2x4's and plywood sides.
Let's try to design one for 43hz.
d = 28900 / (2.06 * (43^2)
d = 7.6"
That's a 2x8, with 3/4" plywood membrane.
Let's try to design one for 50hz.
d = 28900 / (2.06 * (50^2)
d = 5.6"
That's a 2x6, with 3/4" plywood membrane.
Let's try to design one for 63hz.
d = 28900 / (2.06 * (63^2)
d = 3.5"
That's a 2x4, with 3/4" plywood membrane.
Let's try to design one for 75hz.
d = 28900 / (1.375 * (75^2)
d = 3.5"
That's a 2x4, with 1/2" plywood membrane.
Let's try to design one for 97hz.
d = 28900 / (2.06 * (97^2)
d = 1.5"
That's a 2x2, or a 2x4 on it's narrow side, with 3/4" plywood membrane.
Let's try to design one for 118hz.
d = 28900 / (1.375 * (118^2)
d = 1.5"
That's a 2x2, or a 2x4 on it's narrow side, with 1/2" plywood membrane.
The key point is the density of the panel, so weigh the plywood when you get it home, and cut the side frames to give the correct depth.
Building these things to a target frequency apparently is difficult, as the real world for a variety of reason's doesn't match the formula. Also there's the problem of resonance/ring back into the room, i.e. that it'll not only absorb at that frequency (a good thing) but might even ring at that frequency (a bad thing). The fiberglass inside should help damp the ringing, but it widens the Q of the absorber and reduces its ability to absorb.
For example, as you make them smaller area than a 4'x8' sheet, structural stiffness becomes more of a factor. Obviously a 4"x4"x9.7" isn't going to resonate at 38hz. But as for just where a smaller surface area starts deviating significantly from the formula, I don't know. I believe a 4x8 sheet is safe to use the formula, especially in the 50hz to 100hz range. And you probably need a significant surface area anyway to deal get enough sabins of absorbtion out of it anyway.
The lower the frequency, the less likely you'll be able to build one. I know of at least one acoustician (Eric Desart) who tried to make a 31hz absorber panel, and wrote "Be careful with traps going that low in frequency [e.g. 20hz, 30hz]. Not only the resonance frequency is important but also the damping. Such devices very easily can cause secondary reverb curves, radiating sound back into the room at a minus 20 to minus 30 dB level. ... The resonance frequency itself is only ONE SINGLE parameter. I once tried to correct a 31 Hz problem with a panel resonator (horizontal). I could tune the panel resonator to the EXACT frequency. But I couldn't get the internal damping high enough. And when I got it high enough it lost its efficiency. So even with the correct resonance frequency I just got other problems instead. So I stopped and demounted the whole thing (which had cost already time, energy and money)."
If you can find some instructions on how to build an accelerometer tester, you can find out what the resonance frequency is of something you've built.
Or you can purchase resonant absorbers some from companies like RPG who sell traps tuned to a target frequency and Q.
http://www.bobgolds.com/WideQ_NarrowQ.gif