measure concrete room

[Sor]

Hi all, quick question. I’ve got a concrete room that is framed in on all sides. Just a rectangular “bunker” with studs along all the walls. I’m trying to plan subwoofer location so I know where to wire.

Is it worth measuring the room in its state with a box sub to determine room modes/standing waves and ideal sub location, or will it totally change when we do insulation and drywall?

As you might expect the room is an echo chamber right now. I’m thinking maybe I should wait and measure after insulation but before drywall, that will probably be much closer to what the room is like after drywall and treatments on the walls.

 

[Wayne A. Pflughaupt]

It is certainly worth measuring in its current state! Adding drywall will change the room boundary dimensions slightly, but your measurements will still be valid for placement selection.

You might run the wiring to more than one location, to give yourself more options or if you decide to add additional subs.

Regards,
Wayne

 

[EarlK]

Hi all, quick question. I’ve got a concrete room that is framed in on all sides. Just a rectangular “bunker” with studs along all the walls. I’m trying to plan subwoofer location so I know where to wire.

Is it worth measuring the room in its state with a box sub to determine room modes/standing waves and ideal sub location, or will it totally change when we do insulation and drywall?

As you might expect the room is an echo chamber right now. I’m thinking maybe I should wait and measure after insulation but before drywall, that will probably be much closer to what the room is like after drywall and treatments on the walls.

Here's a thread worth reading ( as a cautionary tale ) of subwoofer responses within a concrete ( bunker ) shaped room .

The more concrete sides ( surfaces ) out of a total ( of 6 making an enclosed box ) the worse ( the higher the "Q" ) becomes for these LF room modes ( ie; peaks & valleys ).

There are no easy answers ( or even somewhat complicated answers ) when it comes to mitigating these LF room modes ( apart from pouring new concrete into non-parallel surfaces ) .

REW has a very handy Room Mode Simulator that can give an experienced user an idea of just how intractable the problems really are ( adding subs & different sub positioning becoming secondary to "heroic room treatments" ).

:>(

 

[Sor]

Thanks, I’ll look it over. Luckily the room will not be concrete shortly, there will be air space, framing, insulation, drywall, like any other room in short order. Then another layer of rigid fiberglass. It will be interesting to test before and after to see the difference.

 

[EarlK]

Hi Sor,

Here's a thread worth reading ( located over at diyAudio ).

Acoustics of Corners

 

[AudiocRaver]

Agree with the above advice. The Q will be higher with bare block, so the dips especially will be narrow and deep (peaks, too - remember the dips will get broader with drywall added, so don''t get tempted to discount them), but the frequencies will be very close to the same with the walls more finished. So if you find locations for subs that yield good LF response in the current state, same locations should give good results with the room finished. Theoretically.

 

[Matthew J Poes]

Hi Sor,

Here's a thread worth reading ( located over at diyAudio ).

Acoustics of Corners

Hey Earl, I'll just mention I don't think the non-parallel surfaces does anything good for the bass. The bass doesn't really care what the canter of the walls are until they become great compared to the wavelength, which is never true in small rooms. Concrete walls are certainly a problem due to their stiffness, but a finished listening room with well damped walls using resilient mounting and CLD damping can and does dissipate a lot of LF energy, in fact quite a bit more than you would get from typical in room treatment. I find my own room as a good example where room effects are not as pronounced as one might expect given the room's design. I don't have floor to ceiling bass trapping.


That's a measurement with no EQ in my room and 3 subs playing, nothing was optimized.

Here was an EQed result of response and waterfall (I don't have a waterfall of the above, but don't recall seeing any strong ringing). This one shows some ringing at 70hz which is mode related, but again, in this particular measurement, there isn't really much bass trapping (and 70hz is above the point at which CLD walls offer much benefit). I've also been questioning that 70hz ringing since it doesn't show up in an older measurement of the room with different subs, and think its sub related. See the additional waterfall. Point being, concrete is bad, but is fixable in the room construction.


Here is just my mains, which have very little bass of their own due to their design (12" pro driver in a small sealed enclosure) and placed about 30" from the front wall)

 

[EarlK]

Matthew,

In review, you have a wall system in your listening room that's similar to what was mentioned within the linked-to thread over at diyAudio by E. Geddes ( correct ? )

Since you can attest to it's benefits perhaps you can convince Sor to go that direction .

FWIW, I ran REWs room mode predictor months ago for the thread I referred to in post 3, ( adjusting parameters until Room Mode frequencies and their "Q's" agreed with the thread starters own measurements ).
- I never posted the results since they were so discouraging.
- I tried to smooth the bass response by adding more subs ( as well as shifting their locations and phase ).
- I was never able to get any where near a respectable result using that methodology ( which was quite an eye-opener ).

 

[AudiocRaver]

Hmmmmm...

OK, here is my weird way of looking at it, NOT a recommendation, just a perspective...

"Non-parallel walls [in small rooms] make room modes harder to predict." True, but who cares? Predictable or non-predictable, room modes will be there and must be treated one way or another. True, the wall dimensions relative to wavelength are small at, say, 60 Hz (wavelength = 18.8 ft), but that non-parallelness would tend to "soften" the Q of the resonance in question and make it easier to deal with using traps / absorption / treatment. If there is no significant cost adder, why not?

It is worth noting that any angled wall must be REALLY SOLID to significantly affect the room dimensions and modes acoustically, which might dictate the use of more expensive construction methods, which might undo the cost effectiveness of the idea.

Also remember that room modes and reflections / cancellations are two different animals, and reflections / cancellations can happen at frequencies that look like room modes but might not be. A properly constructed angled surface might help soften / eliminate a reflection / cancellation problem. Again, it is going to happen and must be dealt with, so a cost-effective approach that makes it easier to deal with might be worthwhile.

Am I missing something here?

 

[EarlK]

Hi AudiocRaver,

What was likely missed ( since I was not clear about my intentions ) is the following.

I ( really only ) casually mentioned non-parallel surfaces and in retrospect wish I hadn't ( it was meant as a scare tactic to trigger a dialogue about implementing alternative solutions to mitigate high-Q room modes.
- From what I can gather this was all a bust since Sor hasn't been here for days and is apparently non-plussed by any of this .

The much more practical suggestion of implementing "lossy walls" became my next focus ( due to the non-practicality of reworking cement foundations ).

And now, here we are. ;>)

 

[AudiocRaver]

Yes, Earl, I can get a little carried away with a hot button, thanks for the reset.

 

[Sor]

Hi guys, I do appreciate the input. I’ve been busy with some construction and planning, intend to try to take some box sub measurements soon. I’ve never used REW before but I have my microphone and a 600w ported sub (which is not what I intend to use but should be sufficient for now, from what I gather).

Being that I’m still in the design stage I’ve got some bass traps I’m planning for soffits all around and in all corners as well.

 

[Matthew J Poes]

Hmmmmm...

OK, here is my weird way of looking at it, NOT a recommendation, just a perspective...

"Non-parallel walls [in small rooms] make room modes harder to predict." True, but who cares? Predictable or non-predictable, room modes will be there and must be treated one way or another. True, the wall dimensions relative to wavelength are small at, say, 60 Hz (wavelength = 18.8 ft), but that non-parallelness would tend to "soften" the Q of the resonance in question and make it easier to deal with using traps / absorption / treatment. If there is no significant cost adder, why not?

It is worth noting that any angled wall must be REALLY SOLID to significantly affect the room dimensions and modes acoustically, which might dictate the use of more expensive construction methods, which might undo the cost effectiveness of the idea.

Also remember that room modes and reflections / cancellations are two different animals, and reflections / cancellations can happen at frequencies that look like room modes but might not be. A properly constructed angled surface might help soften / eliminate a reflection / cancellation problem. Again, it is going to happen and must be dealt with, so a cost-effective approach that makes it easier to deal with might be worthwhile.

Am I missing something here?

I'll answer this in any case, even if Sor isn't around. It's worth the discussion I think.

From what very little research I can find on LF standing waves in rooms with non-parallel walls, I can't find good evidence that it mitigates them or lowers their Q, as you indicated. As such, I guess I'm still a bit dubious of this idea (for standing waves). I suspect it won't have a benefit, but would potentially shift them around. It would mitigate flutter echo, which isn't such a bad thing. While I know some folks feel this is of little value if you are to treat the walls anyway, but I do think the biggest lesson I learned as I began to treat rooms was that absorbers don't absorb like you think they will, so the room itself needs to be part of the effort.

As for the walls being true barriers, that is another issue. The true wall or acoustic barrier and the room's physical barrier are not typically the same. It is my understanding (and I'll admit to not understanding this all that well) that folks aren't fully sure about all this yet. Toole wrote in one of his articles a bit about adding some inches to the rooms to account for the midpoint of the physical wall depth to account for the walls impedance better and better predict modes. I've always wondered if I misunderstood him because a few inches wouldn't make much difference, and I have tried this trick to no avail. My new theater is interesting, it has a unique set of circumstances that gave me some experimental opportunities to try some calculations (but still befuddles me). The concrete outer wall is actually many feet from the loss resilient but massive inner shell in various locations. This means my room's hardest barrier would be so far from the inner shell that it should have a huge impact on the location of peaks and dips and would make the effect obvious (more than a few inches of extra wall depth). In fact, I find this to be true, but they still don't match the true modes. I find that I can only explain them using a kind of hybrid approach with a lot of dimension fudging. Some modes seem to be due to the true exterior wall barrier (and thus suggesting that the LF's are passing right through 2 layers of drywall and damping compound). Other modes seem to match the rooms inner dimensions, but also happen to align with SBIR. Yet others match nothing I can calculate and I can't find their source, I rack it up to errors in such a modeling approach. I'm taking an online course on BEM modeling offered by a well regarded Universities physics department online but not for credit. Hopefully I'll learn enough to begin using the freely available BEM software and maybe get good at modeling rooms accurately.

Also remember that room modes and reflections / cancellations are two different animals, and reflections / cancellations can happen at frequencies that look like room modes but might not be. A properly constructed angled surface might help soften / eliminate a reflection / cancellation problem. Again, it is going to happen and must be dealt with, so a cost-effective approach that makes it easier to deal with might be worthwhile.

Am I missing something here?

I guess I would want to use different terms to clarify what you mean here. A room mode is a standing wave, meaning the wave is reflecting off boundaries and it is hitting other reflected waves, causing interference. This interference will cause peaks and dips, but because they are due to wall reflections (so the wall is the source) the existence of a peak or a dip (or not) depends on your position. A speaker boundary effect still has the speaker as the source and is interference caused by reflection that interferes with the speakers direct radiation. Those peaks and dips don't move with your change in position, only with the change in the position of the speaker itself. Is this was you are referencing?

So then I just would again question why a cantered wall would soften the interference if the angle of the wall is small relative to the wavelength? Are there maybe some research studies that looked at this to suggest this would be so and why? I only could find some really old papers which appeared to have their results disputed more recently. I found some more recent work but I can't get the whole paper, and from what I can tell, their claims seem a little hard to believe (so I would want to see the whole paper). All I could find to support this notion is that the energy in the axial modes are shifted to tangential modes and as such their strength would be lower. That has some logic behind it but I would imagine that it would strengthen the tangential modes quite a bit, so that means it mostly shifts the modes.

Since a proper experiment is not really possible (as you noted, the walls need to be really stiff, so testing a space that is square, then not square, and fully working out the change in modes as a result of the cantering vs change in room dimension would be very difficult if not impossible. Instead models using BEM approaches would allow a far more free set of parameters to be tested. I think this is probably where it needs to be examined, but as of yet I can't do this and I can't find much in the way of published work looking at this.

Since Dr. Geddes did his dissertation on LF's in small rooms, I'll ask him if any of his modeling specifically looked at the impact of cantered walls. That might give some insight.

For what it's worth, I did find a paper that looked at curved walls. Apparently the curved walls were very bad for bass modes.​

 

[Matthew J Poes]

Hi guys, I do appreciate the input. I’ve been busy with some construction and planning, intend to try to take some box sub measurements soon. I’ve never used REW before but I have my microphone and a 600w ported sub (which is not what I intend to use but should be sufficient for now, from what I gather).

Being that I’m still in the design stage I’ve got some bass traps I’m planning for soffits all around and in all corners as well.

Hey Sor, As you probably saw in both my comments and Earl's, the room's actual wall construction impacts the bass response as well. Bass traps of the type you are talking about are only effective down to so low. They operate on velocity absorption which is fairly inefficient at low frequencies. The very stiff cement walls are great at trapping sound and when sound is trapped all it can do is bounce around. This is why bass in a cement bunker is so wickedly uneven. You should consider using a lossy wall construction technique. If your only goal is LF absorption and not sound proofing, then you can use thinner 1/2" drywall with something like Green glue and you can use clips and hat channel or resilient channel if you can be certain to use very careful construction (resilient channel can be short circuited by screwing through to the stud). You will want to insulate the walls, this helps. You are effectively turning the wall into a giant bass trap. This concept hasn't been studied in an experimental sense as the study would be next to impossible, but it has been simulated and small prototype samples have been measured, suggesting that it works quite well. The room's construction technique can be one of the most important acoustic treatments you can make.

 

[Sor]

Ok. I had intended to insulate but had not considered using channel, as I expected this was primarily for isolation and avoiding sound transfer. If it can aid in this manner as well for absorption I’m in.

 

[Sor]

Here's my first shot at using REW. For the most part I just see a whole lot of ringing, as one would expect. I tested four subwoofer positions, I'd be curious to hear opinions about which positions (or mixes of positions) might be preferred based on this.

I used a 600W ported sub and a UMIK-1 to take these measurements. Just turned the laptop volume all the way up and left the measurement at -12db.

 

[EarlK]

The fourth trace ( position ) looks the smoothest ( 60 to 200hz ) to me .

That said, none of them have what I would call a sub-woofer curve.

Instead, they all look like the lower frequencies from a full-range box.

 

[Sor]

Maybe it’s because I disabled the crossover on the sub. Most people are probably testing subwoofers in a finished system trying to tune the room and have low pass filters turned on. It made sense to me to get a full sweep since I’m testing the room, not tuning an installation.

I'll probably put subs in multiple places. Here are the locations, FYI:

#1 - center of room, on floor immediately in front of riser

#2 - Right front corner

#3 - Front center

#4 - Right rear corner