I don't know the frequency resolution of this REW software except that the menu says something about "512". I think one just might draw some conclusions about this abstract point by examining the curves, eh.
Not the time or place or need to defend my little hifi.
Holes are not esp. audible. Again, the fallacy of "hearing by eye". Are you diss'ing a "huge 25 Hz wide "hole"" - showing up at one particular location for one single mic? Or are we talking about a typical speaker tolerance band of, say plus-and-minus 6 dB?
In which case, the "huge "hole"" really starts to look like it is outside the tolerance band by not a whole lot and in a region where precise hearing just ain't happening? If you are going to "hear by eye," then you should "look" appropriately. But more important, just how would you propose to fix that fraction-of-an-octave hole (electrically, acoustically, or magically) and without making something else worse?
Did the charts make their little point? Do you have something substantive to add?
Footnote for builders: it may sound sad, but I had almost no control of acoustic parameters in setting up my subs. The rather small room had only one corner suitable for the Klipschorn and the height, width, depth, and location of the giant OB sub were fixed by the furnishings... as was using my 50 yr old Stephens driver.
Last edited:
It took me about 20 years after I got out of school before I realized that woofer/enclosure systems could be easily and straightforwardly analyzed using Newton's second law of motion as applied to forced oscillation. It is a standard well known equation with a standard solution. This is true for any woofer system. Response depends on tuning three parameters, moving mass, spring constant, and damping. In principle, an acoustic suspension system or any other system can be tuned for any arbitrary response. Acoustic Suspension is the easiest to design for. Although there is some mechanical restoring force to minimize the manometer effect, most of the restoring force is due to air pressure. It should have been called a pneumatic suspension system. (It's the same equation used for many other phenomena such as designing automobile suspensions.) The problem with ported systems is that the spring constant K is a function of frequency. At resonance, K is very low and offers almost no resistance to moving air through it. The same is true to lesser degrees for multiples of the resonance frequency. Between resonance frequencies, at the halfway points, the port impedence becomes very high. This is why there is often a roller coaster response curve. In an AS system, the restoring force is independent of frequency, subject only to the ideal gas laws. The amount of air trapped inside determines K. B, the viscosity or velocity related frictional damping factor is controlled by the stuffing, that is the packing of fiberglass or other material. The speaker must work to push and pull air between the fibers. The best design goal is for a system damping factor of .707 with the lowest F3. Because of the constant 12 db per octave falloff below resonance, the response is equalizable. The 24 db per octave falloff of ported systems means that the resonance frequency is for all practical purposes the low end cutoff of the system.
I thought this was all rather clever of me until I was reading Sams Audio Engineers handbook and found that the Thiel Small analysis had beat me to the punch by years and had essentially rearranged it to make it easy to design enclosures for specific drivers using a simple cookbook formula. Still I like to look at it from Newton's point of view.
There are a couple of factors that determine the frequency resolution. It's very important to know what frequency resolution you're looking at. For example, switching from 1/6 octave to 1/24 doesn't mean much if there are only 3 data points per octave.
Depends how wide and deep they are, e.g. a hole from 0Hz-20kHz isn't especially audible
There are a couple of factors that determine the frequency resolution. It's very important to know what frequency resolution you're looking at. For example, switching from 1/6 octave to 1/24 doesn't mean much if there are only 3 data points per octave.
Depends how wide and deep they are, e.g. a hole from 0Hz-20kHz isn't especially audible
Those points are trivially true and therefore a waste of bandwidth.
Does my chart really look like "3 data points per octave"? Gimme a break.
Just where are you going?
So you do agree that the hole around 50 Hz in your graphs needs to be fixed?
I cannot know what the frequency resolution of your graph is (but it's important to know), so I had to ask you. That's where I'm going.
Last edited:
Studying my response plots, if I visually split them at what looks like the Schroeder frequency, it is about 400-600Hz. I understand that this is not a brick wall change and the calculators suggets fs is lower.
I have been running my subs up to 250Hz and it isn't easy to localise them, and it isn't a problem in practice. Given the potential smoothness, should I try running them up to 400Hz?
I have been running my subs up to 250Hz and it isn't easy to localise them, and it isn't a problem in practice. Given the potential smoothness, should I try running them up to 400Hz?
If you had a mixed-bass sub directly behind the L and the R speakers, you could have them play really high and it wouldn't do much harm for the stereo image. But as you moved them away from the main speakers, the freq for non-detectability would have to go down. Nothing too brilliant in me saying that.
There is a freq which I'd say is like 120-140 Hz*, at which there is no place where there is any detectability on acoustic music provided no highs or noise is hearable from the subs. But some boastful person is going to pop up and say he/she can localize to 5 Hz, or better if they don't have a head cold or when playing 2 Hz sq waves.
You can think of this another way. If you have Ionovac colossal tweeters, ALL the instruments with even a tiny bit of upper harmonics seem to come straight out of the little blue balls of gas. Rather impressive stereo image, eh. And as realistic as could be. In that case, hardly matters where you put any of the other speakers, kind of.
*I have a lot of experience living with odd corners for my Klipschorn mixed-bass with a 140 Hz, 24dB/8ave cross-over in many rooms over 40 yrs.
Last edited:
I guess I'd agree. As I alluded to, I can only just make out a sub if I turn around and look at it but even so, I can live with it.
I may try this first with the sub I have close to the mains.
Off topic, but have you heard some? I once built a plasma tweeter. It was directly driven by a pentode that was driven by it's screen. I was using ScanSpeak revelator tweeters at the time. The benefits of the plasma were freedom from resonance and a smooth power response, but what I heard during the limited nearfield listening I did was very similar to the revelator. They both seemed veiled by something, whether it was my preamp, the source/source material (which IIRC, I suspected at the time), or room acoustics. Now that I see acoustics as being such a concern, I see the plasma tweeter as less essential than I used to. They are also difficult to wrap a horn around.
I may try this first with the sub I have close to the mains.
Off topic, but have you heard some? I once built a plasma tweeter. It was directly driven by a pentode that was driven by it's screen. I was using ScanSpeak revelator tweeters at the time. The benefits of the plasma were freedom from resonance and a smooth power response, but what I heard during the limited nearfield listening I did was very similar to the revelator. They both seemed veiled by something, whether it was my preamp, the source/source material (which IIRC, I suspected at the time), or room acoustics. Now that I see acoustics as being such a concern, I see the plasma tweeter as less essential than I used to. They are also difficult to wrap a horn around.
Illusion-risk. Pure illusion-risk. That kind of testing has to be blind. One can easily "hear" all the bass coming out of a flowerpot if we look that way*. Try it.
(Haven't heard plasma tweeters in decades. But the memory never fades.)
*Bit different with real instrument recordings as test sources since they'll always have upper partials which are easier to localize. But those tones are coming from the upper-range speakers. Which, again, is why location of subs, even with highish crossovers, isn't deleterious to stereo.
Last edited:
I count 3-4% of my room boundary opens up into other rooms (i.e. doorways). Is this affecting fs or causing any other issues of importance? I think this ties in with another question: does modal density, or more a lack thereof at a point in the spectrum, affect the ability to analyse its behaviour as steady state?
Large inter-room coupling can be a big effect. This is because substantial delays in signals can begin to appear outside of the integration time of the ear. For example, a 100 Hz tone would take several ms to even detect let alone analyze in a transient nature. In a small room, the sound has virtually reached steady state in this time, so no "transient" effects could be detected and a steady state analysis is not only appropriate, but the only option. But if this small room is coupled to another room that has say a resonance at 100 Hz, then a strong reflection (or better stated an energy feedback) could be seen at times as long as 10-20 ms and this then can no longer be considered steady state. In auditoiums it is known that coupled rooms need to be avoided because of this effect. 3-4% is probably nothing, but many open floor plans in homes basically couple the whole home into the listening room. I don't believe that is a good situation.
Agreed. We moved recently. The new room where I've got everything set up is approximately the same size as the old one, however this one has a doorway into the kitchen, where the other one was just open.
This new room seems a bit more boomy than the last, and I've actually had to reorient my subs. Normally, I have 2 subs on my front L/R (big "bipolars") 2 subs on my rear channels (trianglular "corner loaders") and a sub on my center channel behind the TV. Normally for 2 channel music, only the big subs were playing, but in this house, I've had to put the corner loaders on the front channels instead for listening to music to cut down on the resonance which I think is coming from the new orientation of this room. It seems to be about 45 cycles or so, and it's terribly annoying.
Can't wall off the kitchen in a rental though, bummer.
What do you mean how come?? Its all about displacement, that means surface area * Xmax. To have enough output at 12Hz for it to matter you need displacement. The RE XXX 18" at > 50 mm one way is the badest boy on the planet for ULF. The popular LMS5400 18" woofer has pretty impressive XMax put 4 of those in sealed dual boxes (7 cuft) and you have a perfect sub system down to 10Hz).
100% agree that sealed designs are the only way to get ultimate bass and that does include IB designs which allow use to maximize displacement and power requirements. I do own many sealed designs (A couple of LMS5400s in 4 cuft boxes) and I also own an IB design that displaces 27 liters. You can figure out how many 12" woofers are needed to get that
Its easy to unders 12" woofers are not ULF designs period a high end custom HT room needs better subwoofers then those.
Last edited:
What do you mean how come?? Its all about displacement, that means surface area * Xmax. To have enough output at 12Hz for it to matter you need displacement. The RE XXX 18" at > 50 mm one way is the badest boy on the planet for ULF. The popular LMS5400 18" woofer has pretty impressive XMax put 4 of those in sealed dual boxes (7 cuft) and you have a perfect sub system down to 10Hz).
100% agree that sealed designs are the only way to get ultimate bass and that does include IB designs which allow use to maximize displacement and power requirements. I do own many sealed designs (A couple of LMS5400s in 4 cuft boxes) and I also own an IB design that displaces 27 liters. You can figure out how many 12" woofers are needed to get that
Its easy to unders 12" woofers are not ULF designs period a high end custom HT room needs better subwoofers then those.
That's why they invented the Rotary Subwoofer
Youtube it, it's awesome.
Edit: That 18 is pretty damn impressive as well, I think I'll put a couple of those on my mid-life crisis wishlist that includes refrigerator sized subs.
Last edited:
doug20 - you misunderstand my meaning (because I wasn't clear enough). I see the value of a 12-Hz 12-inch driver not to produce 12 Hz tones badly (as you correctly point out).
Rather, it would be great to be able to buy such drivers to create say, 3 cu foot boxes that went to maybe 30 Hz (or woofers like the 1.4 cu foot AR-1 that goes to 37 Hz).
Why doesn't anybody make them?
(I don't have to the skill with modeling programs, but you'd think adding weight to cones would to make-your-own 12-Hz 12-inch driver would make sense too.)
Rather, it would be great to be able to buy such drivers to create say, 3 cu foot boxes that went to maybe 30 Hz (or woofers like the 1.4 cu foot AR-1 that goes to 37 Hz).
Why doesn't anybody make them?
(I don't have to the skill with modeling programs, but you'd think adding weight to cones would to make-your-own 12-Hz 12-inch driver would make sense too.)
I do not have that answer, maybe asking the question outside of this thread would create some discussion and give us an answer.
Oh, I'd love a DIY version of the rotary. I'd probably build the cabinet behind a wall in a closet that's in the living room and port the door (after screwing it shut)
Anybody can make a 12Hz woofer! See below.
Hoffman's Iron law: Every bass speaker must negotiate 3 big tradeoffs: Enclosure size, Bass extension, and Efficiency.
. . . You only get to 'choose' two.
To make a 12Hz/12'', simply grab any massive-motor 12" woofer with a very low Qts and a low Fs. A "good" candidate would have Qts in the .20 range, and Fs in the low 20's.
Now pour a heavy coat of epoxy on the cone. Do the backside too. You might want to cut some holes thru the cone so the epoxy can bond to iself thru the cone. Heck, add some rebar while you're at it. Cure. Measure Fs and repeat as needed to reach 12 Hz.
If you're really lucky, the Qts won't be too high to still be useable in a box. Compared to its former self, it will be very inefficient, but it will work . . . until the voicecoil frys, or the coil former or suspension breaks, or your amplifier smokes.
If the Qts is now too high to model well in any box, simply bolt your new outboard bass shaker to the sofa and enjoy.
Many today say that there's no problem with low efficiency: "watts is cheap!" . . . but inefficiency requires an extremly engineered and overbuilt driver. While watts may be cheap, a driver that performs with comparable fidelity at extreme temperatures/excursion to a lesser driver under more reasonable conditions is not cheap!
Sure, one can engineer a woofer and amplifier to survive the efficiency problem; The Carver Sunfire successfully plays Loud and Low in a tiny box. The amplifier could also power a toaster, the voice coil can ignite hardwood, and the pricetag suggests (to me at least) that the tradeoffs are pushed well past the point of diminishing returns. With the same investment, one could build a bump-out addition and a less-compromised sub that performs as well or better. All tradeoffs have a point of diminishing returns. If it weren't so, you'd see more 12 Hz woofers in the marketplace.
Tubamark -
Nice that you took my inquiry seriously enough to make meaningful jokes of the gross-exaggeration variety... albeit based on my own earlier suggestion of weighting the cone.
Right, you can't just mindlessly mangle an existing driver. But you sure can modestly re-engineer one with benefit. There exists a whole range of drivers of every sort - just none that go "all the way" to the Vilchur pole, as I bemoan.
Maybe a life-long character defect of mine, but I have no iron-clad respect for existing designs of drivers or much else. Few serious DIY experimenters do. But horror about changing an off-the-shelf driver seems the basis of your post.
Nice that you took my inquiry seriously enough to make meaningful jokes of the gross-exaggeration variety... albeit based on my own earlier suggestion of weighting the cone.
Right, you can't just mindlessly mangle an existing driver. But you sure can modestly re-engineer one with benefit. There exists a whole range of drivers of every sort - just none that go "all the way" to the Vilchur pole, as I bemoan.
Maybe a life-long character defect of mine, but I have no iron-clad respect for existing designs of drivers or much else. Few serious DIY experimenters do. But horror about changing an off-the-shelf driver seems the basis of your post.