If you have any gaps on each layer, caulk them.
You also probably want to add extra clips near the corners for support.
rabident, good point about the extra clips.
In my case, the exterior walls are already framed and are rigidly attached to the floor joists above. One of my walls also supports an 18" LVL as well. So, with that in mind, and clips and channel all the way around gives each wall the same impedance, I'm planning to use clips and channel on all the walls.
Working under the assumption that this thread has served its purpose, maybe I can hi-jack it to work from this point...
Who cares if the walls have equal impedance?
I've read that they should - I just fail to find compelling reasoning to support that claim. Any insight?
In fact, my intuition suggests that the acoustics of a room might be better given differing impedance.
Topic for a new thread you think? Maybe someone with more knowledge could chime in? I'd certainly like to know, myself.
I was thinking about that. Trying to come up with a way to present the situation without presenting a bias. Also trying to decide which sub-forum...
If you don't see a new thread from me in a couple days, it's up to you to start it.
Personally I think this level of design consideration if for an enthusiast with commensurate level of expectations for the rooms performance.
I just like using clips and channel on the walls and ceiling because it does a wonderful job of soundproofing. Keep in mind that once a room is built It is doubtful that you would redo a wall because you thought you could skip the channel.
It's a big deal with room designers
The Soundproofing Company
Right, so let's deal with this here if we can...
It's a big deal with room designers - this is what it seemed like to me. Whereas, the do-it-yourselfer doesn't seem to get hung up on it. One would easily say "that's because the do-it-youslfer doesn't know what he's doing. If he had the knowledge to make better decisions, he would ensure that all the surfaces had equal impedance."
That may be true, but it's flawed logic. I need more data...
As I suggested to J_P_A, my gut says this is baloney. Here's my understanding and rationale:
Wall impedance is a complex, frequency-dependent phenomenon. It is an expression related to a wall's tendency to reflect and absorb sound. Some sound will be absorbed within the wall structure. Some will be transmitted out the other side. In general, we (as do-it-youselfers) concentrate on the transmitted sound, trying to minimize it, but ignore the difference between reflected and absorbed sound.
Since all of these behaviors are frequency-dependent, changing construction techniques leads to transmission, absorption, and reflection at differing amounts at differing frequencies.
A theoretical space is constructed with the same materials and techniques on all 6 surfaces. That surface has a very high STC rating at high frequencies, but shows a minimum on the STC graph at a low frequency - maybe 50Hz, for argument's sake. Similarly, the damping characteristics of the surface reach a minimum at a different low frequency, let's say 40Hz. I don't have good exposure to reflection data to know what a wall assembly may do at certain frequencies (those data are not often talked about in DIY circles, in my experience), but I'm forced to assume that there is a similar pattern for reflection, such that the wall becomes more "transparent" to sound as the frequency drops. So what does that get us?
That gets us a very predictable set of characteristics - if we have the data.
So what does that do for us? How do we use that data?
If we have sophisticated modeling software, we can plug in those data into the model and get high resolution predictions for modal response and bass reflections. Anything else?
What if we don't have access to that modeling? We get a strong pattern of response from all the surfaces that either works with or against the modal characteristics of the space. I have no way to predict what the impact of that interaction will be. So what value is the matched impedance to me?
In contrast, my gut says that if the walls are as variable as possible (while still meeting my needs for STC) I get less strongly correlated response. What good is that? Well, it seems to me that it means that my chances of fighting both wall impedance and room-size problems at the same frequencies has dropped to almost zero. And what have I lost?
"What have I lost?"
wall impedance and room-size problems at the same frequencies has dropped to almost zero.
You could also win the lottery; but, then again, what's a room-size problem? Did you mean room response?
When I said room-size problems I meant modal response.
Can you clarify what you mean by decorrelated response? This is probably the aspect of calibration that I have the least understanding of. It sounds to me like you're saying that there will be unpredictable phase shifts (phase shifts that vary with frequency) in the reflected sound, but like I said - I don't follow.
Anyway, Gary S Kendall of Northwestern University published a paper in 1995, and the paper came up in a Google search. His paper begins with a background discussion where he defines decorrelation as "a process whereby an audio source signal is transformed into multiple output signals with waveforms that appear different from each other, but which sound the same as the source." Taking this along with Big's understanding that "wall structures and their particular impedance have an effect on in room acoustics and in particular the resulting frequency response of the room" brings me to a partial understanding of what Dennis is getting at. I conclude (at least for now) that the concern Dennis has about calibrating a system in a room with unequal impedance boundaries is that (again, as Big suggested) the reflected sound will not have the same sonic character as the original signal. This is related, it seems, to what I've often read from some of the more loquacious posters on the subject here as "eq"ing the reflection. The result is that no single set of EQ parameters can satisfy the needs of both the direct and reflected sounds. Further, if one were to find a set of EQ adjustments appropriate for one seat, they would be wrong for another seat as the balance of direct and reflected sound will be different at each location.
Can I get some feedback about my "new knowledge?"
Also useful in building my understanding: http://www.etfacoustic.com/correlation.html
You brought it up.
Just think about the totally hosed up in room response you'd have (at any frequency) where the left wall is 100% reflective of a given frequency and the opposite wall is 100% transmissive. Imagine the challenges if, at a given frequency range the Q of the response from that wall is radically different than the Q of the response (peak or dip) of the opposite wall. If you'd like to stick to the "baloney" theory, go with it. That's your choice; but, to suggest all audio room designers have no sense of logic or are otherwise idiots, is not a position which will garner favor from those professionals in the community from whom you may be seeking advice ... free or otherwise.
In retrospect, it should have been been obvious to me that if it (any concept, for that matter) were important to pros, a frame of mind that suggests it's not important is clearly wrong. I should have started by learning more about the concepts before I began stating any conclusions (no matter how tentative and preliminary) about their usefulness. In some ways, my skepticism does me a disservice - I try to avoid just taking generalities from the forums as best practices. In this case, equal impedance is the best practice but the underpinning reasons aren't part of the easy-to-give advice, so my knee-jerk reaction is "prove it."
The good news to me is that I feel like I've learned a lot about this at this point. Certainly not enough to use it effectively. Maybe not even enough to decide what time, room space, and money pursuing equal impedance boundaries is worth.
Is 16" stud spacing a best practice for low frequency transmission loss maximization?
This link discusses SRI and TL, and while I don't have the exact details that would be required to connect SRI or TL to impedence, it seems they would exhibit the same regions of response characteristics: specifically, a resonance region. (Sorry about the image; it's transparent and better viewed against the black background of its source webpage.)
Dr Marsh's online lesson continues with some summarized points that I've copied here:
Resonance and coincidence effects cannot be eliminated. If the designer aims to create the maximum SRI, an attempt should be made to get resonant frequencies as low as possible (preferably well below the audible range) and the critical frequency as high as possible (preferably well above the audible range). Whilst it is not possible to apply a generic solution to all panels, the following relationships do hold:
- Reducing the stiffness of a panel lowers it's resonant frequency and raises it's critical frequency, basically increasing the region for which the mass law applies.
- Increasing panel mass also lowers resonant frequencies and raises the critical frequency.
- Decreasing panel thickness raises the critical frequency but generally reduces panel mass.
- Increasing the amount of damping applied to the panel will not alter the frequencies of resonance and coincidence but will act to reduce their effect.
- Good insulation is therefore a combination of low stiffness, high mass and high damping (given cost constraints).
If reducing the stiffness of a wall panel lowers it's resonant frequency, shouldn't that be accomplished by spacing studs farther than normal, allowing the gypsum some room to oscillate? There must be some reason this is never recommended; anyone know? Also, any ideas how mobile or weak a wall panel should be to decrease the frequency of its resonance region below the audible range? Is there some practical reason that this is not pursued?
The flexibility of the system is generally at odds with the requisite mass. Very massive things tend to be stiff.
We accomplish our goals by first decoupling the two leaves. This allows the added mass and absorption to further reduce the MAM resonance point. That wasn’t pointed out and is a big deal.
We then add as much of an air cavity as is practical, and most importantly, add as much mass as practical. Mass is the bigger driver of those two variables.
Also keep in mind that we can start to significantly attenuate frequencies 1.5 X the MAM resonance frequency. That is, if the MAM resonance frequency is 50Hz, we’ll be able to get respectable TL at 75Hz and above. To isolate the 20Hz audible low end you’d need a MAM resonance point of 13Hz or so. This would require a decoupled double leaf assembly with something along the lines of a few feet of cavity depth and 10 or more sheets of drywall on either side.
The Soundproofing Company
That diagram is grossly over simplified and, in some circles, be considered incorrect. We typically view the mass controlled region to be from 50Hz and below (16Hz for audible concerns, lower for vibration control). The diagram doesn't address the Primary Structural resonance, Antiresonance, and secondary resonance regions.
I'm relieved to know that the coincidence region is not of particular concern, as I think I may overheat my brain trying to deal with that all at once. Refraction always makes me stop and think clearly; I have to go back to the basic analogies from a conceptual physics book I have.
I appreciate you shining a light down the rabbit hole. Certainly an online forum (much like an online course, such as Dr. Marsh's) would be a poor place to try to learn about this sort of stuff - not to mention that most of us (speaking for myself with certainty) don't have adequate math skills to deal with what I'm sure becomes multi-variable calculus and worse very rapidly. Your explanation does enough to confirm for me that I have no practical way to model and predict the behavior of the walls I'm about to build (should have built today, in fact), so unless my librarian friend can help me find a manageable resource, I'll be decoupling, adding mass, and trying to forget about the influences I can't control - that is until I become (or hire) some poor fellow to drag his gear into my basement and try to calibrate my system. Though, I suspect that the features one would need access to in a processor will not be part of the equipment I'm going to install, at least at this point.
The as built flex characteristics cannot be reasonably predicted. The diaphragmatic absorption characteristics of any given wall section may not (and likely will not) perform as predicted either in center frequency nor amount of energy absorption. Equally problematic is the vast majority of modal room prediction models only predict frequency, not magnitude...you may be attacking a problem which will not exist. Here is an example: if you install a 4x8 sheet of drywall vertically, the diaphragmatic characteristics will be radically different than when installing the same sheet horizontally!
Several years ago there was a well respected theater designer (I know the man and have enjoyed his company) who advocated this type of approach to reduce, or eliminate, modal response problems within a room. Once constructed, this approach "seemed" to work. In the end, what was really happening was the walls had so little mass the majority of the energy was passing through the wall. Remove the walls, no modal response problems. But, what else was happening? For one, there was a triple leaf in most of the builds.
The others were more obscure. For example, now that a wall section is moving we have turned the wall into a speaker. The section compresses absorbing energy and then springs back into place. That creates an entirely different set of issues. We cannot accurately predict the amount of energy absorbed, which is problematic; however, we also cannot predict the frequency, phase, or amplitude of the energy created when the diaphram snaps back into place.
This method, with good intent, was to resolve a projected problem; however, the method itself induced other, if not more disturbing, issues. A key consideration also overlooked is once built, how do you fix the problem? Basically, you tear out the wall and start over. So here is where DIY (or self proclaimed 'designers' or never built a room before lab rats) get into deep dodo. They read about the concept (as you have), latch on to that concept, and implement the concept without having the practical experience or depth of knowledge to understand the full suite of impacts of that concept or method. When you latch onto some tidbit of information, understand it is exactly that ... a tidbit.
People like Ted bring that experience and knowledge to the table. It's fair to ask "why". OTOH, understand a long list of "well, but, I read that..." gets old fast, especially when the same dialog comes up from several different individuals, in different threads, and different forums. (No real solution to that.)
May I ask where is a good place to learn "basics" ? I don't aspire to learn more than fundamental concepts - but I'd like to learn more than I know now. Advice ?