im surprised to hear that the clips and hat channels are not very effective, everything i have read tells me otherwise,
As Greg already mentioned, they DO work. They just don't work as well as a fully-decoupled inner-leaf wall does. There's a limit to how much isolation you can get from clips with hat channel (or resilient channel), and for what you need, that limit is not good enough. For someone who needs modest to decent isolation (let's say around 40 to 45 dB, give or take a few), either RC or clips + hat is fine: it can do that. But once you need isolation beyond about 50 dB, you've hit the limit of what can be achieved. That's when you need to move to fully-decoupled framing. It's sort of the same as the issue with staggered studs: You can get more isolation with a staggered-stud wall than you can with a single stud wall, for sure, but there are limits there as well, becuase there are still flanking paths at the top and bottom of the wall, where the studs are on the same plates, and fully connected, mechanically. If you need more isolation that can be achieved with a normal staggered stud wall, then there's no point just adding mass to the wall, as it's a case of diminishing returns: You'd need to add a huge amount of mass to get anywhere, but putting much less mass on a separate from will work better. That's the issue here with clips: You CAN get get isolation with clips, but not extreme isolation. An with clips, you CANNOT add more than three layers of drywall, since there are structural limitations as well: clips simply cannot support the extra mass.
what i'm understanding is that the small gap that it creates would actually turn the wall into a drum head (so to speak)?
Welllll,,,, sort of, but not really! It's a bit complex, but basically you have to stop thinking about the individual parts of the wall, and understand that all the parts working together act as one single resonant system. In other words, the entire wall will resonate at one specific fundamental resonant frequency. If you happen to play that frequency inside the room, then the wall will not only allow it through to the other side, but will actually AMPLIFY it, so it is LOUDER on the other side. And it's not just that one frequency: the effect extends out to a range of frequencies centered at that resonant point, and that range goes all the way up to 1.414 times the resonant frequency. For complex mathematical reasons, the resonant peak amplifies everything up to 1.414 x F0. At that point (1.414 x F0), the gain is unity, so it neither amplifies nor attenuates. However, above that point, for high frequencies, the wall starts isolating better and better. At twice the resonant frequency (2 x F0), it isolates quite well, and at at 3 x F0 it's doing very nicely. The higher you go, the better it gets. In fact, the isolation curve, or more correctly the "transmission loss" curve, always looks something like this:
typical-MSM-transmission-loss-curve-2.jpg
The shape is always the same. The tilt might change for different types of construction, but the shape of the TL curve is always similar to that, due the laws of physics.
You can see the "resonant dip" at around 100 Hz in this particular example. That's the point where the wall provides the least isolation, because it is resonating in sympathy with that frequency. It is "tuned" to that frequency. In this case, it is providing a little tiny bit of isolation (around 10 dB: about the same as a thin sheet of cardboard) because it is damped internally with abundant insulation. If it did not have that insulation, it would have "negative" isolation (less than o dB): in other words, it would amplify. And considering that things like drums and bass guitars put out a lot of energy around 100 Hz, this wall would be pretty lousy for a studio!
So, the curve always has the same shape: Then how do you get more isolation? You move the curve over to the left! If you can slide that entire curve to the left, such that the resonant dip gets down to, say, 60 Hz, then you can see that the isolation for all other frequencies gets better! If you could do that, then the wall would isolate for drums much better... but still not so much for a five.string bass, that goes down to 36 Hz. So then, how about if you slide that curve over to the left even more? Make the resonant dip occur at, say 15 Hz, where it is way below the frequency of ALL musical instruments? Yep, that is, indeed, the solution. Slide the curve over far enough, and the wall will isolate for ALL frequencies.
That leads to the next question: How do you slide the curve over to the left? What do you have to do to the wall, to make that happen?
Well, considering that the curve is really showing you how the wall is tuned, what you have to do to "slide the curve", is to re-tune the wall! And to do that, you need to adjust one of the factors that changes the resonant frequency. It turns out that there are only two factors that you can adjust here (OK, in reality three, but we'll just look at the two big ones for now). Those to factors are the mass of the wall, and the size of the air gap. Make the wall heavier (more massive) and the curve slides over to the left. Make the gap between the two "leaves" of the wall bigger, and the curve slides over to the left. It's that simple. There's a set of very simple equations that you can use to predict what that curve will look like, or rather, how far over to the left you have moved it. By choosing the size of the air gap and the mass (density) of the materials carefully, you can tune the wall any way you want.
There's another benefit to adding mass and increasing the air gap: the curve not only slides to the left, it also slides UP the page a bit. so you get more isolation from that point of view as well. In fact, as you add even more mass, the curve tends to "flatten out" a bit, at a higher level up the graph.
OK, so now that I've explained that, I can also explain why putting clips+channel+drywall directly over MLV is a really bad idea. Think about it: by doing that, you have created a very SMALL air gap, between the MLV and the drywall. Much smaller than the air gap in a normal wall. So, if INCREASING the air gap moves the curve to the left, guess what happens when you DECREASE the air gap? Yup: the curve moves to the right.....
And for very small air gaps, it doesn't just slide a bit to the right: it makes a mad dash to the right! The resonant frequency goes way up, and the isolation is lousy all across the bottom of the spectrum.
In fact, for a scenario where you have a sheet of 1/2" drywall on clips over a sheet of MLV; the resonant frequency is about 180 Hz, so the wall does not isolate until about 250 Hz, isolates reasonably above about 360 Hz, and isolates well starting at about 530 Hz. In other words, it isolates terribly across the entire low end of the spectrum, and also the lower part of the mid range!
would insulation behind the channels change this?
That's the "third factor" I mentioned above, in addition to mass and air gap. Adding suitable porous absorption inside the cavity, such as insulation of the correct density, does increase the isolation. It does that in several ways, but basically by acting as a "damper" on the resonances inside the wall. So yes, putting insulation in there would help. However, the situation is so bad that the "help" isn't very useful. Filling that entire thin cavity with excellent insulation would decrease the resonant frequency down to about 125 Hz, so the wall would not isolate until around 175 Hz, isolate somewhat at 250 Hz, and isolate well at around 375 Hz. You improved it a lot, yes, but it is still lousy. Consider that a huge chunk of the energy put out by drums, bass guitars, keyboards and electric guitars (among others) is below 200 Hz, and you'll see that this wall is no use for a studio.
In short, you need a much larger air gap. A thin gap, such as the depth of the clip plus the hat channel, is way too thin.
would using resilent channels work better or is that essentially the same thing?
It's essentially the same thing. Worse, in fact, in your scenario of having it directly over MLV since RC is thinner than the combination of clip+hat channel: the air gap is even smaller!
RC (or clips+hat) is only useful when you put it directly on the studs, with nothing behind it. That increases the air gap to be the full depth of the studs, plus the depth of the clip+hat (or RC). With that much air gap, the isolation improves greatly.... But still not as much as you can get by building a separate frame for that leaf.
regarding the window, i dont really have the option of drywalling over it as the city inspector will be giving final approval. getting creative, perhaps the isowindow treatment (removable for the inspector) with some MLV layers glued on the back in coujunction with a removable barrier over the egress hole outside (something sand filled or concret slab?)
I wound suggest just building a wall across there, with a sliding glass door in it. That does not block your egress path (which is probably a legal requirement), so it should pass inspection. It also adds mass (glass has plenty of mass) and an air gap...
so i guess the question of how quite i want to be is difficult, i dont have measuring equipment and even if i did i wouldnt be able to use it as the drum set has yet to be purchased.
That's fine: You can estimate how loud drums are: it's a known factor. Consider a level of around 115 dBC as typical for drums played normally. That's not too much of a problem. But you DO need the meter to figure out the other side of the equation: "How quite o I need to be?". That's not something I can tell you, since it varies greatly from person to person, situation to situation, and legal code to legal code. What my family considers "quiet", your family might consider "really loud", or vice versa. You have to actually measure the level that YOU feel is acceptable, in YOUR situation. And you need a decent Sound Level Meter to do that. Set it to "C" weighting, "Slow" response, and take several measurements at various points in the house, when you think the ambient level is good. Get someone to turn up the radio or stereo or TV in another room, playing bass-heavy music with lots of drums, until YOU think that the level is no longer acceptable, and see what that measurement is. Turn down the noise until it is just on the very edge of being acceptable. Do that in several locations, average the results, and that answers the question. Subtract that number from the 115 dB number I gave you earlier, and that's roughly the amount of isolation you need. Based on that, you can design your wall for the level of isolation you need, by "sliding the curve" over until it gets where you need it to be.
outlets and switches. this is tricky, my wife probably would not sign off on external conduit with surface mounted outlets.
I didn't mention external conduits! I mentioned a "surface mounted electrical system". Such as those made by Legrand, Kalop and others. They look pretty good, actually:
surface-mount-electrics-01.jpg
surface-mount-electrics-02.jpg
The raceways come in many sizes and shapes, to fit all tastes and all needs:
surface-mount-electrics-03.jpg
And you can even get some types that are disguised to look like crown molding:
surface-mount-electrics-04--crown-molding.png
For high isolation, you can allow yourself exactly ONE single penetration of the wall, to bring in the power feed. From there, it is distributed around the room using this type of surface-mount system, or if you build your walls "inside out", then it can be concealed behind the visible surface of the wall, which is not solid, massive or rigid...
more questions to follow,
- Stuart -
Worth it for safety and to pass inspection though!