Home Theater Forum and Systems banner

Required SPL <10Hz

13K views 63 replies 13 participants last post by  Chrisbee 
#1 ·
I read at TRW web site some info about needing ~110dB at 5Hz for the sound to be noticeable. I suspect that is noticeable as an audible sound though, and not noticeable as your clothes are moving and the room is unhappy with you.

Does anyone know what sort of SPL you need at these lower frequencies in order to physically feel something, irrespective of hearing something?
 
#3 ·
Well, my current subs already hit 12Hz, which I've run up to 105dB corrected at my seat ~11' away. The thing is with a sine wave, the effect is different than with an impulse response. My clothes won't flap, for example, if I run a sine wave, although the house will respond. However, there was one point where I got my pant leg to flap a bit in Narnia.

I'm wondering what sort of dB level do I need to hit at 5Hz for example, to have the same experience?
 
#5 ·
I'm no rocket surgeon, but I didn't think it was possible to hear a 5 hz tone. Unless, that is, you played loud enough to tear your eardrums out. Then you could hear them tearing out.:explode::holycow: :R

...but then, if your eardrums were gone, you couldn't hear them, right? Okay, now I'm just rambling.:help:
 
#6 ·
Hm. Well, I have close to zero room pressurization because my room is completely open to the rest of the house, and even then it would be 5000 ft^3. You think sine wave SPL is indicative of when you would physically feel the same frequency in a movie or music? That hasn't really been my experience, which is why I'm asking.

My seats never shake, only things like lighting fixtures. I do sometimes get vibrations in the 30-40Hz range, I think, because they're leather seats.

I guess so far everyone thinks ~90dB at 5hz will be physical.
 
#7 ·
You think sine wave SPL is indicative of when you would physically feel the same frequency in a movie or music? That hasn't really been my experience, which is why I'm asking.
No, I don't. Actual material has dynamic range, so output can be higher or lower than the master volume level you are calibrated to. You can't really determine how loud you need to go with actual material because you would be faced with the chore of trying to seperate individual frequencies from a scene or clip. Unfortunately, no scene or clip I know of exists of only one frequency. Pink noise would be the next best thing.

My seats never shake, only things like lighting fixtures
Are you using a concrete floor?

Well, I have close to zero room pressurization because my room is completely open to the rest of the house, and even then it would be 5000 ft^3
Ok, so what would cause any "pant flapping"?

Are you using an IB? I'm wondering why you are concerned with 5hz output.
 
#9 ·
I currently have two SVS 16-46PC+ subs tuned at 12Hz, located about 11' away from my seating position. They are not co-located.

I don't have a concrete floor, I have a crawlspace construction, with wood on top of whatever the floor is really made out of. I've never experienced that walls flexing, floor moving, furniture moving thing people talk about when they measure subs in their closed rooms. The house does groan when I push my subs to the limit, but I've never seen visible flexing. And the furniture doesn't vibrate.

I would assume the pant flap I experienced is due just to air movement, not pressurization.

And the reason I'm asking about <10Hz is because I am trying to put together a sub that would give me noticeable output down to 5Hz. I'm having trouble figuring out the right thing to do because:

1) ported/PR has huge group delay, which I understand you believe to be okay, and driver excursion remains an issue when the tune is 5Hz instead of ~15Hz.

2) sealed rolls off and I'm not sure I can EQ to a flat response without making things worse; 5Hz at 90dB and 30Hz at 120dB.

3) IB isn't a good option for me because this is the first floor of a two-story house. The only potential IB enclosure is the fireplace, and that would be very hard to create a seal around without re-doing the fireplace.

Right now I'm kind of leaning towards a 6-driver sealed sonotube.
 
G
#10 ·
I currently have two SVS 16-46PC+ subs tuned at 12Hz, located about 11' away from my seating position. They are not co-located.

I would assume the pant flap I experienced is due just to air movement, not pressurization.

And the reason I'm asking about <10Hz is because I am trying to put together a sub that would give me noticeable output down to 5Hz. I'm having trouble figuring out the right thing to do because:

1) ported/PR has huge group delay, which I understand you believe to be okay, and driver excursion remains an issue when the tune is 5Hz instead of ~15Hz.

2) sealed rolls off and I'm not sure I can EQ to a flat response without making things worse; 5Hz at 90dB and 30Hz at 120dB.

3) IB isn't a good option for me because this is the first floor of a two-story house. The only potential IB enclosure is the fireplace, and that would be very hard to create a seal around without re-doing the fireplace.

Right now I'm kind of leaning towards a 6-driver sealed sonotube.
Fireplaces are a no go for IB subs

You need to understand there's a HUGE impedance mismatch between the air in the room and the drivers at infrasonic frequencies. So obviously a pair of 12"s just won't cut it.......:no:

The rotary woofer is the elegant solution to this issue. The other solution is TONS of REALLY BIG woofers. Start with 18"s and think about using no less than 6 and more will be better, if you want usable output at 5Hz. Go sealed since a port is basically useless at these frequencies. Also you'll need EQ and monstrous amounts of amplifier power. Note that EQ that works that low is either very $pendy if available or must be custom made.

Earthquake proof everything if you do this since anything loose will vibrate.
 
#11 ·
For the sealed solution, I'm thinking ~700L in a sonotube, with six 12" drivers (3 on top, 3 on bottom) with 1500W of power. This is about -12dB at 5Hz compared to an IB using six 18" TC-5200 drivers, more so at 2Hz. The sealed solution also gives more SPL until the 5Hz tuning point of a ported 700L sonotube with three 12" drivers and an unrealistic port, and is only a couple of dB lower at the 5Hz point than the ported.

What sort of PEQ filters are typically applied to an IB?

Essentially, this is why I am asking about the <10Hz frequencies and what sort of SPL I need to be able to hit. If 90dB at 5Hz is going to be useless, or it is impossible to get a flat response using PEQ filters, then I'm wasting my time and money.
 
G
#16 ·
Reality check time......

My big IB as a dozen 12"s. Here's an example of it's nearfield output (without EQ). Understand this wasn't an attempt at max SPL. I was just interested in how low it went. The measurements were made with a $5000 laboratory grade mic

Freq Hz---db output
4.921-------79 db
6.200-------82 db
7.374-------87 db
8.227-------89 db
9.290-------92 db
10.42-------94.3 db
11.70-------95.5 db
12.40-------96 db
13.91-------98.5 db
14.75-------98.5 db
15.62-------100.2 db
16.55-------100.8 db
17.54-------101.6 db
18.58-------102.1 db
19.68-------102.4 db
20.85-------102.9 db
22.01-------103.1 db
23.40-------103.3 db
24.80-------103.4 db
26.28-------103.2 db
27.83-------103.2 db
29.49-------103 db
31.24-------103 db
33.10-------102.8 db
35.07-------102.8 db
37.15-------101.4 db
39.36-------100.4 db
41.70-------100.2 db
44.18-------100 db
46.81-------100 db
49.59-------100 db
 
#12 ·
Why not at least use 15" drivers? 12" drivers is going to make it that much harder. Personally, if I were you, I'd consider how much $ and work you would have to put in to get something usable at 5hz, then consider how often you will actually need to recreate 5hz, and then decide if it's still worth it, or if more headroom and lower distorton in more common bass frequencies is going to be a better application.
 
#13 ·
I modeled the same box with 15" drivers and it didn't give me anything extra. It just costs more and would require a sonotube that won't even fit through the door. Maybe less distortion, but I don't know about that.

I've been trying combinations of 12", 15", and 18" drivers within the same ~700L, in sealed, ported, and passive radiator combinations too. And six 12" drivers in a sealed sonotube seems to be the most cost-efficient approach, and doesn't seem to sacrifice much.

Granted, I won't need 5Hz often, but this project is to give me something I don't already have, i.e. <12Hz, while boosting a bit in the frequencies I already have by adding more drivers and watts. In the end, this single sub might be able to replace the two 16-46PC+ subs. I don't think I'll know until after it's built though. If I try to go for more headroom and lower distortion than what I currently have, that either means doubling up the 16-46PC+ subs which would cost about the same, or replacing them with something like what this project would end up creating.
 
#17 ·
Well, I've been modeling everything with TC Sounds drivers. I think 15" drivers didn't give me more than the 12" drivers because of the enclosure and power restrictions. 700L and 2000W is the most I want to deal with (i.e. a single wall amp on a single wall outlet).

For example, say six 12" TC-1000 drivers in the sealed 700L volume is the reference. If I put six 18" TC-5200 drivers into that same enclosure, I can give it 5000W (not realistic) and get +8dB at 5Hz. With a more realistic 2000W peak and then only needing four drivers, +5dB at 5Hz and its super expensive. Three 15" TC-2000 drivers in 700L with 2000W peak is +1dB at 5Hz.

If I model a single 6ft^3 box with two 15" TC-2000 drivers, I can feed it 2500W peak. Three of them would be up there with the six 18" TC-5200 drivers but with 7500W peak. Are these sort of watts actually realistic? I don't think I can pull that much reliably, even with three 20A circuits running 15A outlets.

Thanks for the IB numbers. That helps me put things in perspective.
 
#19 ·
quick shot...here's 6 tc3k's in a 700 liter enclosure with an Fb around 32 Hz:




a behringer ep2500 and some creative wiring and who knows...depends on your longest room length as to how much gain you'll get...but this is a 5 dB/octave rolloff below 32 Hz...down 12 dB @ 10 Hz from the 50 Hz peak (80 Hz /4th order crossover) and down ~24 dB at 5 Hz (not factoring room gain)...qtc. =.38...infinite baffle type sound signature...assuming you'd build the enclosure yourself...$2600 for the drivers, $350 amp, and whatever the cabinet could be built for...a steep price no doubt, but infrasonic output costs...:yes:
 
#20 ·
Longest room length is maybe 30' I guess. That's about what I've been modeling too, except I've been looking at TC-1000 drivers with 2000W peak. Then it's like -34dB at 5Hz, but coming from 126dB at 50Hz that's still respectable. I've been wondering if PEQ would flatten that in a way that makes sense. Maybe it doesn't make sense, since cone excursion >20Hz isn't that much.

I've been looking at more amps, and I guess there are some that can go much higher than 2000W peak off a single 15A circuit. So those would still work. So far, unless I'm messing something up, I can get crazy SPL (>120dB) over 20Hz with almost any box I design that also tries to hit 5Hz. So I figure anything I do to try and get 5Hz is going to give me the necessary SPL, distortion, and group delay in the higher frequencies anyway.

I guess I just need to keep tinkering with some models and also look at stacking sealed enclosures; building a single 700L MDF box is just too difficult. I wouldn't even be able to move it. That's why I was looking at a 700L sonotube. But if 90dB at 5Hz is an admirable point, based on the IB numbers, then at least I know that's something worth shooting for without the pursuit being meaningless.
 
#21 · (Edited)
i think you may need more than 90 dB at 5 Hz...Ed Mullen's tests seem to indicate 100 dB+ for noticeable impact at 12 Hz and below...i'm sure 90 dB may add "something" to the experience...but "how much" something?

Reasons why this may be a futile effort:

1. Your room gain would seem to start at a low 18-19 Hz

2. electronics attenuation of VLF signals...would you get "enough" pass through < 10 Hz? very important, otherwise this is all moot.

3. Room size, i would'nt leave it to 12" drivers...3 good 15"ers would undo (6) 12's (sealed)

4. Costs...99.9th percentile bass requires you spend a proportionate amount on the driver system.

5. Size/weight...(3) 2x15" units would come at about 200 lbs. a piece...6 cu. ft. for each unit/24" cube...The TC3000 would be the perfect driver for your application...small footprint, 3.5" peak to peak excursion envelope, and low qts. However, the drivers are $425 shipped (maybe a discount on four?)...if you had a room in the 1500 to 1800 cu. ft. area, i would say you could take a stab at it with (2) 2x15's:scratch: Kgveteran on AVS has a setup like this with 4 Tumult 15's and a Bassis sub equalizer, needless to say it is very potent.:holycow: ...Bossobass also on AVS, has the most extreme sealed setup that i know of...a Rebase system he calls it, also using a Bassis equalizer (3-$400)...i think 4 (2x15)...if you could get hold of these two guys, they could give you some insight into what you could expect...but they use 15" drivers so...:dontknow:

It would take some fancy eq to engineer something close to a house curve, a lot of midbass output would have to be sacrificed of course...the simulation i did for 6 tc3k's would be down only 24 dB which is better than your 34...a 10 dB difference down that low would definitely be noticeable...and depending on how far down you can get the Fb, you may realize better performance than the models indicate...

I'm only running a lowly single 2x15 so i can't offer much consolation (i do have the Vd equivalent of 5 12" drivers though)...that said, i've been testing with some infrasonic tones (down to 3 Hz) and i have to turn the receiver up to between -10 and 0 dB to score some palpable impact...i'm down 7 dB @ 8 Hz (from 10 Hz/ -10 dB overall)...the 8 Hz tone, i think, was genuine, but the 4 Hz is probably all harmonic, i could'nt differentiate between the two...whirring helicopter blades is the best way to describe what i heard...the tones were supposedly recorded at 0 dBfs so i probably have pass through issues myself...the drivers do follow the tones, excursion-wise...i would'nt say that i was blown away by the output/extension however...even with a first order rolloff, shelving eq and a small room, 9 Hz is pretty much the bottom of the barrel for me...8 Hz shows up but it won't add too much to the experience...below that?:huh: :dontknow:...i'll soon be adding a third tc3000 to the mix, so we'll see what improvements can be had with an extra 5-6 liters of Vd...good luck:)
 
#22 ·
Thanks very much for the info. Do you recall where Ed Mullen's test results are? I don't remember seeing anything posted by him regarding this. I only find graphs of the audible threshold, which isn't the same thing. The "how much" something is exactly what I'm trying to determine, to see if this is a waste of time and money.

And as you mentioned, roll-off in the low frequencies occurs in the electronics. I've been looking at the QSC RMX amps, for example, and they are -3dB at 5Hz. I don't know if the balanced out on the Emotiva DMC-1 is also -3dB, it might be, but I sort of think it's not as much as -3dB based on the other measurements I've seen.

I'd forgotten about the Bassis unit, because previously I was looking at a passive radiator design. I should look at that again.

Do you recall the SPL you needed at 8Hz to feel it?
 
#23 ·
Hm. Linkwitz Transforms don't seem to be helping. Oh well.

Oh, could you tell me where how you see three 15" TC-3000's having more displacement than six 12" TC-1000's? When I multiply out the Vas, it seems like the six 12" TC-1000's have more.
 
#26 · (Edited)
Vd is the measurement that is critical here...this is the swept displacement of the driver, peak to peak (linear)...when the driver was first released, the stated xmax was 34mm linear...winisdpro autocalculated:

34mm /Sd=829.6 cm^2 and Vd=2.82, or 5.64 liters swept/ 109.5 dB @ 20 Hz

Later the xmax was dropped to:

30mm /Sd=829.6 cm^2 and Vd=2.49, or 4.98 liters swept/ 108.5 dB @ 20 Hz

Discussions with kyle, indicated that the excursion envelope extends out to around 46mm, assuming you have the power to drive it to that point...this would be over 7.5 liters displacement, of course distortion would be high, but it's more protective than usable...watching a single driver during "F' Irene" would "illustrate" this capacity quite vividly...point of the fact is that it has much more excursion headroom than the TC-1000...more linear xmax and sensitivity (+3 dB)to boot. Tale of the tape?
15 liters (3x15") vs. 13 liters (6x12")...and the cost would be a tad less for 3 15's (3x395) vs. 6 12's(6x198)...you may save more on shipping also...

Keep in mind that the TC-3000 is probably the best for this application, given size and budgeting, but it's not the best 15" driver offered...the LMS 5400 series should have a 4" peak to peak excursion envelope, retaining all of it's motor force over 38mm one-way...of course you'd have to pay twice the price of a tc3000 for this performance...

It's hard to say how much i felt at 8 Hz...

I'll disconnect the mains and run the 8 Hz tones tomorrow and let you know more...a high load on the ep2500 though:help: ...it causes severe flickering with my lamps while playing it back...:blink:

While the threshold for hearing is supposedly ~110 dB @ 5 Hz, how much spl does it take to feel this frequency? that's the question...i'll see if i can find Ed Mullen's findings on VLF. I think Mark Seaton has done some work in this area also.

The Linkwitz transform is more efficient/accurate than shelf filtering but there "is" a limit to how much you can "squeeze" from a driver. I will say this though...below 9 Hz, the response falls off at a more rapid rate...the FBQ 2496 and Audigy2ZS soundcard have a stated FR of < 10 - 44/46 KHz...who knows what the < 10 means but it would appear to be accurate given the 7 dB loss between 8 and 10 Hz...the ep2500 is also on its way out at 5 Hz (-3 dB)...the cones are still moving down past 4 Hz, but they are'nt producing anything tactile...:thumbsdown:

I'll check this all out tomorrow though. I'll run trueRTA also and be looking for the 2nd and 3rd harmonics...i won't be able to "see" the fundamental though so it'll be hard to measure actual distortion for the 8 Hz signal.:rolleyesno:...I may also bypass the Yamaha and run the signal straight from the soundcard to the ep2500 and see if that makes a difference.:scratch:

That said, you may not be able to overcome both the room and the electronics roll off without considerable expense in new electronics and SOTA drivers...again that 99.9th percentile thingy...that's why i was saying that maybe 8 Hz would be a more feasible goal...

Btw...i want that cat!!!...i'll trade you for a snooty himalayan named "Sasha" with tons of fur...hehe:)
 
#24 ·
Well the Vd of the tc-3000 is 2.4L(800cm^2 x 3.0 cm) and the Vd of the tc-1000 is 1.1L (470cm^2 x 2.4cm), so a single tc-3000 will out-displace dual tc-1000s. I just used Sd measurements from the Sound Splinter Rl-ps since the actual tc-3000 and tc-1000 measurements weren't available on the TC Sounds website.
 
#25 ·
Yes, electronics rolloff is another good point. Processor's are typically rated as -3db at 5hz, as are pro amps (though some are rated at 10db below max output, so at max output, it's even further down). An EQ unit is also gonna have low end rolloff. I'd count on being at least 8db down at 5hz as a baseline. Ilkka took some measurements, and low end rolloff in electronics is indeed a real issue.
 
#27 ·
here's something you might find interesting also...



TRW-17 transducer $12,900.00

Motor Controller $350.00

Amplifier and crossover $700.00

Design and installation, typical $8,000-$12,000*


Total $21,950-$25,950

the 100th percentile solution...:eek: ...and it ain't pretty:rolleyesno:
 
#28 ·
I've been applying a first order Bessel 5Hz highpass in my models so far, to account for the SMS-1. I guess it would make sense to make it a 2nd or 3rd order highpass, to account for the processor and amp as well.

Thanks for explaining the difference between Vd and Vas. I didn't realize they were different. Perhaps I entered the information incorrectly into WinISD also. I'll have to double-check. Thanks very much for taking the time to look into what SPL you can feel at 8Hz. I appreciate it.

The problem with the Linkwitz Transform is that boosting such a low frequency, i.e. if I try to get Fb to be 5Hz and Q to be .707, results in WinISD saying I can only use something like 2-10W (depending on the model) before I reach max cone excursion. I'm not really sure I understand that. Is that like 2W into the Linkwitz Transform, and somehow 1000W comes out the other end (in terms of line-level signal)?

I think everything ends up either running a highpass filter at 5Hz in some form or another, like -3dB in frequency response or the 5Hz filter in the SMS-1. :(

The way I figure it, if I can construct something that will replace the SVS 16-46PC+ subs, then a budget of ~$2000, or maybe a few hundred more, is reasonable.

You want my cat?
 
#35 ·
i saw a cat in one of your HT photos...i think that only 24 dB of boost is available with the Bassis unit...the amplifier will determine how much boost you can get and at what level without clipping...that far down in the power band most amplifiers will exhibit serious power loss...

I ran the test tones today...at 4-5-6-8-10-12 Hz...i'll sum it up as pressure and harmonic:

Behringer calibrated ecm8000:

12 Hz pressure 105 dB @ 3.5 meters (5% 3rd HD)
10 Hz pressure 100 dB @ 2 meters (10% 3rd)

Radio Shack digital @ 2meters (uncorrected readings):

4 Hz harmonic 87 dB no pressure
5 Hz harmonic 90 dB no pressure
6 Hz harmonic 89 dB no pressure
7 Hz no test tone
8 Hz pressure 93 dB pants flap..."definitely" present with "impact"
9 Hz no test tone
10 Hz pressure 96 dB
12 Hz pressure 97 dB

I'll run this again several times to confirm...but it looks like i'm -4 dB @ 8 Hz.:scratch:

I was'nt able to bypass the receiver and run more tests but perhaps next week i could do this also...

The 8 Hz signal comes through strong enough to add to the experience...like BHD, which has a strong 7-8 Hz signal...i'm pretty sure that this is what i'm feeling on "Irene":yes:

In summary, it seems that i'm getting at least 100 dB, assuming modest correction factors and relatively low distortion, down to 8 Hz.:T

I'll do more testing next week.:)

Thanks to Mark for chiming in...i was hoping for some help here...:clap:
 
#29 ·
I've often wondered whether a really long pipe out into the garden and back again (or folded under the floor) would make a serious infrasonic sub. Probably too many harmonics unless it was very large in cross section. Somewhere between 64 and 120 feet long sounds like a good starting figure to compete with the TRW. :D
 
#30 ·
That's the approach used by pipe organs and that huge underground subwoofer design (concrete and bricks with several large drivers). But a pipe would only have one tune; unless you want that to be your port length so it's coupled with an in-room driver. You'd get a really low tune but still have the group delay problems. For SPL, I think you'd not want something so long, but rather multiple wide tubes of maybe a dozen or two feet in a ported configuration???
 
#33 ·
Hi Steve,

I caught some of this in various places and know there is a roll off, but could you point me to any measurements? Am I correct in recalling CJ from AV Talk taking some?
 
#32 ·
Hi Guys,

Interesting discussion that seems to be going in circles somewhat. Since my name was mentioned, I'll note that in my own experience, assuming there is significant headroom at higher frequencies, exceeding 100dB @ the listening position in the 10Hz range with minimal compression has been an approximate point at which I've felt such extension to be useful on real program matieral. Don't get me wrong, you don't want to dial in the response to be +10dB relative to 80Hz with only 100dB capability at 10Hz, but I would try and keep this range within ~6-8dB of the 80-100Hz range. In my personal experience, I prefer the sound and experience of this over a system that has a steeper roll off below 20Hz. Subjectively such extension has always sounded deeper reaching and larger in scale as compared to inserting a high pass in the chain.

The "art" lies in balancing your playback expecations and the capabilities of your system. That deeper reaching and larger scale sound isn't so pleasant if you are always hearing your sub amp clip or a driver bottom. This of course is dependent on many factors involving your listening space, environment, personal preferences, and of course the capabilities of the rest of our system.

To break down the above, once you get over the threshold of around 100dB @10Hz (PLEASE, PLEASE don't get crazy with this as any sort of brickwall/absolute, it's a ballpark figure) you should experiment with EQ or other means of including as much of this VLF capability as your system can comfortably take. If there wasn't much limit to the VLF headroom in an ideal system, I would probably want the response to be about +6-8dB @ 10Hz relative to 80-100Hz, again heavily dependent on the rest of the system. Even so, I still feel that even -10dB in the same context can be worthwhile.

Just my take.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top