Recommended filler sub(s) to work with IB?

RS390HF in 5ft3 with the Dayton 250 watt amp.

Best bet is to match your main IB in output so you dont end up building new ones like i will ! I did just as you did, read my thread if you can in my signature. I really needed 15's or 18's even though they are at least 15 to 20db lower than the mains and their response is very narrow....my main subs are very capable and quite able to give me huge numbers into single digits at the LP....the Filler subs should be as capable...

..... Unless you have a setup like Geddes discribes, Welti/Devantier outline subs of equal output.

My first step will be to pick up a pro amp like my quad QSC RMX 1450 and power my 12's to see if the plate amp ( 240watt plate Dayton ) is crapping out. My though is i can squeeze a little more out of them with a good pro amp....we will see...

One reason why you'd want a high output auxiliary sub is if one owns a very high output main subwoofer system. While, yes, these auxiliary subs are there to help smooth and flatten the overall response, they should be somewhat capable or one will run into problems that Kgveteran ran into where his very capable 4x 15" system would outgun his dual 12" 250w aux subs. At high SPL levels the aux subs would run out of steam.

The aux should be nearly as capable as the main subs depending on personal use and overall level, imo.

Hey DC, Geddes starts off with a corner placement sub first, then a second sub, and a third any where you kinda want.

Im not sure where your IB manufold is located inregards to your LP, but you may very well have to do a modified multi sub method as i did, and doing so changes everything inregard to the aux subs capability. This is why i built two 12" cabs, i thought their limited bandwidth and output would make some high output 12's a perfect match.

I need to get a pro amp in my room first to see if the plate amps are just crapping out on my, which was discussed in my thread....

A MiniDSP was my best friend during this whole escapade of mine....

I'm not sure that's entirely correct. I believe he recomends the first sub being placed closest to the mains so they can run a in a slightly higher frequency range than the rest to overlap the mains in the upper modal region. He does suggest having one sub in a corner since corners "see" all room modes. The first sub can be in the corner, but it is not a requirement.

Also I do think you are probably ok with subs which are much less capable than the mains. This is certainly in keeping with the Geddes approach. Some sealed 10"s or 12" should be plenty. These smaller subs will need to be high passed though... 30-40Hz.

@kgveteran from reading your thread it really looks like you might be pushing your small plate amps into clipping. I'll be interested in seeing your thoughts when you get more capable amps that you aren't in danger of overdriving.

It depends on the multisub approach you are implementing. The Geddes method doesn't require an aux sub to have the same output capacity as the first and second position subs. The Welti method seems to require subs of equal output capacity.

IMO KGVet ran into issues because he tried to mix the two approaches.

Please give us a few more details on your present setup. Amp, drivers, room size, eq, placement of IB manifold, seating position etc.

DC,

Im chasing you around haha actually this is my homebase, in the DIY section. Have you had a chance to read the Geddes paper regarding multi-sub placement and requirements. IIRC the paper states the second sub be as capable as your main sub, or at least within 3-5db's of your IB setup. This could be a side wall or another corner sub. The third sub in his idea only needs to be within 10db's of the IB sub and would be a back wall sub, but in your case, you could at least TRY to skip to this step and do back wall placement first since, looking at the graphs, the little table tuba seems to fill in the majority of your problem anyways

If you really like what the table tuba did, why not build another one and make it's home in the theater permanent? The small sealed 15 is another great option IMO. While having seperate locations with super comparable subs is obviously ideal, most people dont have the ability or permission (WAF) to do such a thing. Me? I dont have room, no WAF factor to speak of Enter Geddes' approach, which I think at this point in time you are more aligned towards anyways. You arent going to be able to build one or two more IB setups in other corners of the room, so you need to see what you have that can get as close to the output of the IB while maintaining your budget. perhaps do something that is scalable... small sealed 15 with a decent driver, install, measure, build another if needed, rinse, repeat.

That seems about right, i only need about 10db out of them, time will tell....

Yes, my setup is a multi sub setup with no real resemblence of either multi sub setups, hence my need for more capable sides, or more capable amp or amps !

Now, back to the IB discussion......

This much I believe to be true, (linear capability being a given) the holy grail in transparent, fully optimized sub-woofing nirvana is optimal phase response.

Now onto this issue at hand. I may be entirely wrong, but I'd approach this from a different perspective. Yes, it appeared a secondary, rear of the room LF source helped the frequency response. However, this commonly used approach (including myself), is an acoustic band-aid,...not entirely addressing the situation. Sure, subs in the rear of the room do help in a variety of ways, but they also add other additional time domain issues that ideally must be dealt with for system optimization.

I'd suggest exploring the back-wall, and it's acoustic interaction with the front wall IB energy, and the resultant peaks and nulls that your encountering along the front to back room axis.

The null in frequency response you're encountering is acoustic interference. All the frequency response peak and null changes along the front-to-back axis, that you see upon measuring are acoustic interference.

The direct energy launched into the space off the IB drivers, encounters the reflected energy off the boundary surfaces and the waves collide. If the wave energy is in phase at the point of measurment/listening,... it results in a peak. If the acoustic energy of the two colliding waves is out of phase, a null is the result.

So, instead of creating more issues with another source, examine what can be done to dampen the rear wall's involvement. I'm not well versed in these but there's a variety of approaches that include a narrow band absorption, ie, the diaphragmatic absorber, Helmholtz type resonator, or a maybe multi-layered, multi damped membrane approach like PNW did with the Octagon rear wall.

At any rate, identification of the center frequency of interest, make an element of the rear wall surface dissipate that energy, clean up the time domain and the frequency domain follows.

Alternate approach, just sayin'...


Thoughts?

A misconception in all this is that the filler sub is just filling in a little sound. What i discovered in my project is the bandwidth was indeed narrow, with a hipass and a lopass leaving a narrow 40hz-80hz ( one octave )..... But the output of the filler sub was that of the mains, minus a few Db due to the close proximity of the fillers to my side.

I was lead to believe i only needed about 18db of output to fill in this null
It didnt work out that way at all, in the end i only shaved off a few Db to achieve a flat response, that shaved off level was only due to close proximity to the LP....

Now i'm not sure why i couldnt just add a few Db to fill in the null, but it is what it is, and this setup is not Geddes or Welti/Devantier.....its a modified version, with the only resemblence being multiples.....

Feel free to add in a theory on why the null is only -15db and i require almost full output to achieve a flat response to fill it in....at the LP

FOH-

have you tried this? Seems like it would be great if it was feasible. I think though that trying to absorb every peak and null between 40Hz to however high the last major room mode is say... 120Hz would be very, difficult. I don't know this for sure, but it seems like you would have to have tuned bass traps on the ceiling and two adjacent walls to remove most of the room modes. You would certainly need a dedicated room, or a VERY understanding wife/girlfriend.

In DC's other thread he has really worked hard and been given a plethora of suggestions on room treatments. almost to the extent of a virtual fistfight it seemed. I know there has been some work done here,more broadband absorbtion, not any helmholtz resonators or diaphramatic absorbers that I remember hearing about, but bass traps and 4inch panels have been incorporated.

It's not necessarily room mode related, what I'm referring to is essentially a 1/4 wavelength cancellation effect, SBIR. Easily calculated whereby the energy passes you from the front, encounters the next boundary and returns out of phase, thus nulling the response. Any measure of minimizing the amount of reflected energy results in less of a null. Most peaks and dips in response are from 1/4 wave interaction, not modal influence. This, however, doesn't diminish the importance of modal influence on a room's frequency response, it's just different.

Thanks

Most of this stuff is LxWxH issues. How fortunate to have a fix !!!

What I mean is the primary element that impacts a room's behavior is directly related to the 1/4 wavelength effect, not modal issues. Yes, a room's inherent modal signature is important, and certainly needs examination and EQ'ing in freq domain, and damping in the time domain. But overwhelmingly, the 1/4 wavelength effects are most important because they dominate the response strongly.

A room's response freq peak, will remain at that frequency regardless of position in the room. The magnitude may vary, but not the frequency. Subsequently, this is easily tamed with some freq EQ'ing for a given position.

Back to Chris' issue, I agree,....and his experimentation illustrates, adding a rear sub helps the null. However, there's complications. In his thread at the Cult, Thomas suggested relocating two of the Fi's to the rear. This, however ideal, presents unacceptable issues for Chris. Introducing a resonant design, or something like the Table Tuba, presents phase/group delay integrating issues. This is why I wanted to introduce examination of the actual root cause of the null effect.

Aggressive bass trapping, be it the typical approach of thick, porous fiberglass traps, or perhaps pressure based boundary loaded diaphragmatic dissipation of energy, will help smooth in room frequency response. Regarding wall Q, or the relative impedance of the surface, it's well understood the more rigid the boundary is, the more well defined the peaks and nulls in response. So ideally, we want a space with boundaries as to elicit the wonderful effects of PVG, but the boundaries need to be relatively lossy as to lessen the sharpness of the Q.

Acoustically, the room behaves differently above and below the transition frequency. One can view the rear wall this way too. The theoretical ideal is 100% absorption of all energy,....just like being outside. There is no SBIR, or 1/4 wavelength effects in such conditions. However, the acoustic returns off the rear wall can be valuable to create an nice listening environment. That said, you need both,....as much absorption of LF energy, yet create some reflected energy back into the room above the transition frequency. A concrete basement wall is essentially worst case scenario,....very rigid, with well defined peaks and nulls. I believe it's fully whisper clipped, drywalled etc, but I'm sure it could benefit from extensive LF attenuation (yet retain MF/HF energy) ,...which isn't very expensive.

Sorry for the lengthy post.... It's an interesting issue of which I've experimented with, so thought I'd throw in my 2 cents....



Best of luck

So, once the FR is where you want it, with the integration of filler subs, what would be the next test....

FOH, I think you bring up legitimate concerns but ultimately many of these proposals are difficult to implement in a home or even a dedicated HT. PNW's room is about as over the top, from a performance perspective as it gets. Not many people can pull that off. I would still use a multi-sub setup with a treated (meant as any and all options...not just traps) to get smoothing across multiple seats. I've never seen a small treated room that couldn't benefit. Of course I've not come across one that was able to implement all of the aforementioned solutions.

I still believe Geddes's formula is the most practical setup. The key is to get the majority of your output and subsequent smoothing effect from the two primary subs. The key there is having one in a corner and one at some other point along the front wall near the mains. With a little experimentation, this alone can achieve most of the benefit. These subs need to be capable of full output.

The third and/or fourth sub locations are meant more for reducing variance across seats than for smoothing the response. These do not need to be capable of full output. The key is to keep the levels low. I don't mean to pick on KG, but I believe he was trying to get a "filling" effect out of his rear subs and was asking for too much output from them.

You are going to get most of the filling effect from the first two front full headroom locations. Ideally, you can experiment with both locations until you get the best spatial average across all seats. This obviously doesn't apply with a single manifold IB since you are basically stuck in the mud.

KG should have separate his 4 Tumults into two boxes and placed them based on experimentation and measurements at ideal locations across the front. From there he likely would have filled all major holes and could use the rear sub to reduce variance across his seating positions.

IMO, the inability to experiment with sub locations is the biggest catch with single manifold IBs. Of course, prior to construction, you could experiment and come up with 2 locations across the front wall the yield the best results and then place your manifolds in those 2 spots.

So what should the OP do? Well, one option would be to add a 2nd front wall sub that is nearly as capable as his IB. I have a feeling that is not in the cards due to size and cost. The other option would be a rear sub that has significant capacity from ~30-80hz. This won't get results on par with the Geddes formula but should yield improvement.

I ran out of available floor space LOL.... I guess even a dedicated room has it's location problems. The Tumults are two per cabinet, and i couldnt get them where the response was even half as good as where they are....


Ur not pickin at all, it is, what it is .......
I'm hoping to pick up a used EP2500 for the AUX subs in the next two weeks or so, but ur right on, the AUX subs arent able to kep up as they sit with 240watt plate amps... If that doesnt give me the output i need to get back to a flat response, i think a nice pair of 18's would do....

I hear ya. There is always some compromise that must be made. That makes things fun. In your case, you do need more output from your Aux subs. I'm quite certain you are running into amp limits and not speaker limits.

Your setup might actually be a better model for the OP since he is similar stuck with a single location, high output sub presently. He will need to use a sub that has output capacity on par with his IB. The good thing is that it really only needs to do this from 30hz and up and this means he won't need a huge box or huge volume of displacement.

I guess at this point, i would stay away from plate amps and go straight to pro amps, unless you have to... Once the cost of a plate amp rises above $250.00 you are now ready for a used Behringer EP2500 on ebay with tons of power in a neat package to store anywhere you or i want.....

Just for yucks, the dual Infinity thread looks to be a winner. Do a nice opposing duals with a pro amp and away you go. It will either be a work horse if you need it or loaf along like my mains do unless we are watching UnderWorld !!!!!!!

I must subcribe to this thread.... This is good stuff, cant wait to see what worx

I am looking to revisit my friends system i installed about 6 or 7 years ago. Check out the slots cut in his hardwood floors, i did a dual manufold, before i owned a calibrated mic. He has quad sound splinter 15's.....

I guarantee he has major problems at the LP and some nice little side subs will do the trick

What I'm suggesting is merely examining the cause of the null.

The course of action I put forth had it's genesis in the fact that the OP didn't want to relocate his IB drivers to the rear, and wanted an economical approach of just a few hundred dollars. This is why I shifted the attention to acoustics. Yes, PNW's Octagon is over the top, however I used it as an example of a planar array LF section, which is essentially what Chris has implemented. That said, PNW's acoustic treatment approach to his rear wall is dirt cheap. Either using roofing felt, or equivalent limp mass membrane, or some plywood (or equiv) diaphragmatic absorber attenuating the problem return frequency. By sizing the plywood appropriately, and with an adjacent 703/705 fiberglass to change the resonating plywood energy into heat. Simple, not too expensive, however it would take some simple experimentation.

It is said that plywood at these thicknesses target these respective frequencies; 1/4"=110 Hz, 3/8"=87 Hz, so increasing to 1/2", or 3/4", or perhaps adding mass would fine tune the native resonant frequency of the panel. One could easily implement this pressure based approach at the wall, then in front of this, a method of velocity based absorption bass trap working in concert but more broadband.



-----


I agree. He's currently using a multi-sub setup, the current IB is not addressing front to back modes, but he is addressing lateral modes by spreading the LF sources side to side. I do agree, a multi-sub set up of distributed sources works wonders in smoothing frequency response aberrations. I just wanted to look at the cause.

The more I look at 1/4 wave cancellation, it does dominate the peaks and valleys of our response.





Many good points. Again, I agree on the multi-sub approach, Geddes or otherwise. I think keeping things simple, wrt capability and strain, and retaining identical output per sub. Clearly this would be preferable all things being equal. I'll be implementing many of these principles soon, blending an IB with highly capable filler subs. But high in my priority list is phase linearity over as wide a freq range and over as wide a listening area as possible.

Yes, any IB install should ideally measure with whatever method possible prior to committing to a location. Stiff boundaries sharpen the Q, lossy boundaries lessen these effects. My biggest issues in my room aren't front to back, but laterally. This is easily addressed. Actually, whichever axis in which the problem arises, spreading the LF sources across the median point typically addresses the issue with somewhat of a selective mode cancellation.



The highly capable rear sub you suggested would help, just as it did in the experiments. But I've got no experience with blending alignment types and maintaining decent phase response etc., so I don't know how it ultimately would blend and subsequently sound.

I just thought we could start examining the time domain. Since I've began exploring as much acoustics related topics as I can, both in intensive reading and mucho hands on in my room and my LP, everything is so strongly rooted in the importance of the time domain. Although I've really immersed myself into the acoustic properties of my room, all I've learned is essentially how much I don't know. What I do know is all the smart guys consistently point to the time domain, and it's relative importance. So I'm pointing too. Additionally, the more I learn, the more I realize the importance of 1/4 wave interaction.

Myself included, we always are quick to suggest adding a rear sub,..and it does typically work. Instead of just looking at a rear sub, or just examining the rear wall's contribution, a prudent approach would include both elements.


Thanks

It nice to just bounce idea's around. I dont think there's alot of guys doing the multi sub thing. The multi sub method requires measurement gear, and thats a little intimidating at first, like it was for me.

I really think there are more methods to this other than Welti/Devantier and Geddes. If we do venture off their methods, we go there a little blind, as i did. So, WTS, i agree, it has been very polite and informative, hoping more guys jump on the band wagon and we can start to compile more data and build what we need based on good solid numbers.......


FOH, lend a little light on this.... and how to modify phase to best suit the response, i just let Audyssey do that part of it.... i'm clueless

Phase describes the relative time difference between two signals.

It's a multi-fold answer. In your case, you want to line up, in time, the leading edge of the bass transient, from each subwoofer location so that they sum accurately at your listening position.

Phase is expressed in degrees, which measure the completed portion of a circular period (wavelength). To help understand, picture 90 degrees of phase delay is a quarter of a period (wavelength) at any frequency. Every piece of gear in the chain, from the cables to the speakers, can affect the phase of the signal. Speakers/Subs can easily affect phase differently at every frequency because they possess a changing inductive load.

Upon getting to the point of visualizing distance in wavelengths, then you can see how subs will interact with each other and your room. The acoustic sum of the subs at different locations will either add or subtract depending on the relative phase alignment between the subs output. When two signals of the same frequency are combined, the summed response will be either greater than, or less than the original signals, depending on the phase.

Not only do the subs need to play nicely in phase, the subs summed total, must be optimized with the mains at the frequency at which the two are crossed over. This is very, very important for realism in playback. The reason is the alignment in time of all the harmonic elements of a lower fundamental. An E string on a bass guitar has a fundamental about 41 hz, but the complex nature of it's harmonic structure needs to be preserved across a system consisting of multiple drivers covering differing frequency ranges, and most importantly in this case multiple locations of drivers covering the same frequency bandwidth.


---


A nice exchange of ideas is always a pleasure.

Yeah, the rear speakers do add a level of complexity to a rear boundary wall panel absorber. But it could be placed between, below, or anywhere the surrounds aren't.

The image below, is taken from Ethan winers superb acoustic treatment article.




Ethan's paper I linked to is so, so good. Even if one is reasonably well versed in such aspects, it's so well written and so comprehensive I believe a great deal can be gleaned from it. contained therein is a brief section on the panel absorbers in question.

Here's another such article from Ethan.

Also, Ethan monitors an acoustics forum
here, so you may search there or even post a very specific question regarding a pressure panel trap, on your back wall.

This will also aid in calculating your specific scenario.

Here's a forum discussion with Ethan describing elements, and interaction from various individuals.

I've earmarked many of these because I too am interested in implementing a pressure panel bass trap in my new config. I implemented one several years ago, but I had no idea what I was doing and have since removed the panel. Walls of varying impedances dissipate LF at different freqs and amounts. I always felt as many ways one can attack bass trapping the better. A triple attack of the boundary wall, the panel absorber, both combined with velocity based absorption seems ideal.

In your case, it seems like you could experiment quick and dirty by suspending some plywood, closely damped by fiberglass or equivalent insulation, and see the result. Then, varying the amount of suspension, point of contact etc, varying airspace, dampen the ply with additional thickness, less thickness etc, then attempt to correlate which direction to proceed in the experiment.

You could leave the mic at the second row position, while testing,....then attempt to lessen the null by absorbing that freq at the rear wall. Maybe it'd work, maybe I'm way off here. Oftentimes in acoustics, elements that would seem very simple and intuitive are anything but.

I hope someone else can enter in here as well.