29 (0x1D)

29 (0x1D)

When considering other potential ceiling flaw....

The problem isn't so much what you have already done, but rather what you still have to do! In the end, you want to have a 2-leaf structure around, and ONLY two leaves. Not one, not three. not four, and not any other number. Only two. Those are your "outer leaf" and your "inner leaf". That is what builds your MSM resonant system. If you add more leaves, then you destroy that, and create an MSMSM system (or even worse an MSMSMSM system), which will ALWAYS give you WORSE isolation for low frequencies, all other factors being equal. The "outer leaf" is normally your exterior building shell, stripped of whatever would otherwise make it into a 3-leaf system once you build the inner-leaf. And that is the issue here... The problem is that your ceiling already has two leaves (the outer roof, or the floor above, or whatever else is up there inside that dark hole plus the ceiling that you put an extra layer on), but you have not yet built your inner leaf! So no matter what you do now, you are stuck with three leaves, period. In other words, no matter what you do now, you will DECREASE the isolation that you could have had for low frequencies. I'm surprised that the acousticians and studio designers you consulted did not notice this, and warn you about it!

The existing situation is that the ceiling is floated underneath timber roof rafters,

Floated? How did you float it? You didn't show any photos where you took off the existing original ceiling and installed RC or joists on isolation hangers, or RSIC clips with hat channel, or anything else. So how did you manage to "float" it, without taking off the original ceiling?

with a similar product to green glue / decibel drop applied inbetween the layers

Do you have a link to the specs for this "Decible Drop" product, to see what it is supposed to do?

I would considered possibly going the resilient channel route, however, aluminium resilient channels are also not available in my country. I have tried to locate some, but that would entail importing construction materials at a very large cost.

Can you get plain old hat channel in RSA? As far as I recall from when I lived there, you should be able to. If so, it would be cheaper to import a box of RSIC clips and just add your own hat channel.... But I digress: that will not solve your problem. Even if you did get RC or clips+channel, took of the ceiling, but them in, and put the ceiling back again, you would STILL have a 3-leaf structure! That's the problem: With your present ceiling "plan", you have backed yourself into a corner where you have no choice but to build a 3-leaf. So the question is: Did you take that into account when you designed the isolation plan for your studio? In other words, when you figured the TL and MSM calculations, did you use the equations for 2-leaf or 3-leaf? If you did use 3-leaf equations, what were your F- and F+ frequencies? And most important of all, what is your PLAN? What are you trying to accomplish, in terms of real numbers? How much isolation do you need? What NC (or NCB) curve are you aiming for? What RT figures? What room response curve? Etc. I think it would be a good idea if you could post all those details here, so folks can check them over and make sure that you are being realistic about them, because once again, based on your current situation, it looks like they are not realistic. If you post all your specs and goals, and your isolation plan, then we can take a look at them and see if you actually will be able to achieve something usable like that, because frankly, the way it looks now, it seems unlikely.

The locally suggested approach would be to layer a dense glasswool in between rafters, above the existing ceiling board layer as the more common "think pink" would be applied.

Well, that would be part of the spring and damping in the second cavity of your MSMSM ceiling, yes, but you can't just say "throw in some dense glass wool" and hope for the best! What do you mean by "dense"? In reality, density is NOT what you need. What you are looking for is insulation that has the correct gas flow resistance for an MSM cavity. There is a rough relationship between density and gas flow resistance, yes, but you still have to be sure that you are using stuff with the CORRECT characteristcs! Once again, a good acoustician or studio designer would have given you figures, not just "put in some dense stuff". He should also have told you not only the density range, but also how thick (deep) to pile it up, not just to "put some between the rafters". The depth of the damping isolation in the cavity affects which constant you choose for the MSM equations, so I'm assuming you used the right ones?

The roof construction is a pitched roof with dense cement ceiling tiles on a pitched construction.

Great! so we know that you have a gable roof up there. How did you seal that roof deck air-tight? How did you deal with the ventilation issue? I'm pretty sure that the building code in RSA hasn't changed that much in recent years, so the requirement is still there to ventilate attic spaces: How did yo deal with that, and did the inspector sign off on it? Sealing your MSM air cavities hermetically is essential for good isolation. But ventilating roof cavities is also essential. So it is important to see how you dealt with the two conflicting essentials.

Brick Gables close off both sides of the studio, where it pitches down to overhanging eaves on the other two borders.

Great! The gables is usually where the vents are (unless you had a ridge vent, but I don't recall those as being too common in RSA).

I will draw up a few sketches regarding my thoughts for interrogation and post them asap.

Excellent! That's the best way of doing things around here: it means that the rest of us can "look over your shoulder" so to speak, and spot potential issues BEFORE you actually build them, or at least before you get too far along that ripping things out to re-do them becomes daunting.

The idea was to prepare the shell for the more specific / detailed work to follow, and saving some time while the design is finalized.

Well, the shell (outer leaf) is just as important as the inner leaf: they act together, to form the MSM system. Building one without having yet design the other is problematic! That's like trying to put the water in your swimming pool before you actually dig the hole.... :)

I do realize that the shell is just as important, and hope that we can resolve some of the issues going forward.

Hopefully they can be solved, but you'll need to post a lot more info than you have given so far! The original basic plan is the most important. - Stuart -

29 (0x1D)

http://www.sae.edu/reference_material/p ... 0Chart.htm -

If I recall correctly, John Sayers himself wrote that book for the SAE Institute, many years ago, and I think he was a bit upset that he didn't get any credit for it! :) In any event, the first diagram you posted (cavity wall) plus drywall on studs) basically shows a two-leaf system... And the second one clearly shows a plug designed for two-leaf wall, where the inner surface of that wall is the inner leaf of the room: that isn't your situation. You are adding ANOTHER inner leaf to the diagram, which changes everything, acoustically. Anyway, note what you are saying about the isolation: "example of STC ratings used for sound isolation". STC ratings. I imagine that you are aware that STC is NOT a good method for discussing or specifying isolation for studios? STC does not even consider any frequencies on the low end of the spectrum: no bass, no drums, no bottom few octaves of the keyboard.... STC is fine for talking about isolation of voices in houses and offices, but not much use at all for studios. It simply does not deal with the part of the spectrum where all of your REAL issues will be: below about 150 Hz. Those frequencies do not even enter the calculations for STC. You can indeed build a wall that has a great STC rating, but through which drums and bass are clearly audible. You should be looking at TL, not STC.

obviously higher indicates more isolation.

No, not necessarily with STC! Depending on what you are dealing with, a wall rated STC 40 could easily have better isolation than one rated STC 45! You should take a look at the clumsy method used to come up with STC ratings for walls: it's an education! Basically, they take readings at certain specific frequency bands, then nudge those numbers up or down until they get a curve that exceeds another curve... complex, confusing and not very valid for studios...

The best approach would probably have been to remove the window / frame, block up the window with a similar construction to the standard cavity wall and plaster / render to match.

Yes, that would have been the best approach. Second best is to build a plug that is part of the overall studio isolation design. Way down the list "take a wild guess, and throw something together without bothering to see if it will work or not". :)

I would like to have the option to possibly remove the seal / block in the distant future if the room is converted or used for a different use.

That's why people build plugs for their windows, if they can't brick them up for whatever reason. But the plug has to be built with the overall design in mind.

The block is loosely based on a double leaf construction.

No it isn't! It is already THREE-leaves! Count them: Glass ... Drywall ... Drywall (with the correposnding air gaps in between). That makes THREE not TWO. And you still have to add the inner leaf, which makes four.... You seem to be missing the overall picture of what "2-leaf" actually means: It means that between you, sitting at your console, and the world outside, there are only, exactly, precisely, TWO leaves. Your window plug construction is 3-leaves. The inner-leaf will add one more, making 4. The low frequency isolation WILL suffer: simple physics. If you don't believe me, do the math!

You have mentioned a possible issue with thermal properties / cracking etc, I doubt wether this will be a concern in this application. The window is very well shielded from any direct sunlight by the roof eave, as well as extensive tree vegetation in close proximity.

That's the construction issue that Brien brought up, but just as important is the 2/3/4 leaf issue, which you seem to be ignoring.

What would have been a better solution to seal this cavity, but still retain the window / frame?

Build a plug properly! IE, for that specific case that you are dealing with, build a plug that acts as a single leaf.

The existing wall construction surrounding most of the space is standard cavity brick wall construction.

... which very likely is only acting as a single leaf, since it is most improbable that it was built as a proper two-leaf MSM isolation wall. From what I remember of average South African bricklayers and the way they work, I can almost guarantee that your cavity brick wall is acting much more like a single leaf than a 2-leaf, despite the cavity... :) - Stuart -

29 (0x1D)

"As a professional in the field of architecture, I am familiar with the concept of a cavity wall along with rudimentary mathematics (usually able to count to four on a good day :) ). Based on Mass | Cavity | Mass which offers thermal insulation as well as some acoustic isolation, this was the premise for the window block. " Forget Architecture, those guys use tables...we do not care about tables. Let's talk about the difference between what we care about, isolation, and what you get as a by product of just doing an acoustic build...thermal. Forget thermal...it happens just due to what you will do in a build, since insulation is a big part of the project between the boundaries, and on the interior of the room as well...the thermal part is, or should be an after thought. Mass/spring/mass is what you developed in this window plug. Now I ask you, do you intend to install another hard boundary, another leaf? This is where you get multiple small hard air spaces that reduce the TL of the assembly, not that you have not already introduced them already. You can trust me, I deal with Designers, Architects, Owners, etc, every day, so stature, while commendable, does not mean that a person with a degree knows what they are talking about in or as it pertains to the acoustics of a building or room. My experience is that they do not... You have to be ready or at least open to suggestion...and we read a lot around hear, so this is not an attempt at hypnosis. :)

29 (0x1D)

29 (0x1D)

At this point, with the existing window block, it has made a considerable impact.

Have you tested the LOW END of the spectrum? For example, set up a drum kit and a bass player with a good amp inside, and let them belt out their favorite beat while you measure the spectrum outside? That's where you will see the difference. I think you need to take a look at the theory here, and understand that a 3-leaf wall will ALWAYS give WORSE isolation than a two leaf, all other factors being equal, in the LOW END of the spectrum, which is where it really matters. If the low end is getting good isolation, then you can bet your bottom dollar (Rand) that the high end is getting fantastic isolation. But the reverse is NOT true. Just because you have good high end isolation does NOT mean that you have good low-end isolation. You can talk and sing and shout and yell all you want inside, and run power tools and radios and the TV, and the isolation might sound fantastic... until you run the REAL test, which is low frequencies. In fact, a 3-leaf wall will have BETTER isolation in the voice region and above, so if you are getting great isolation there, then that might well be a sign that you do, in fact, have a 3-leaf system. One more time: Your wall is a tuned filter. It isolates terribly at the frequency to which it is tuned. In fact, it doesn't isolate at all at that specific frequency. lets call that "RF" for "Resonant Frequency". But at 1.4 times RF the wall starts to isolate. at 2 x RF, it isolates reasonably well. At 3 x RF it isolates great, and it just gets better and better the further you go up the spectrum, until you hit the coincidence dip, or course, where it goes down again for a bit. But even there (in the coincidence dip) the isolation should be damn good if you built your wall correctly, since the frequency of the coincidence dip is high enough that you are already well up in isolation, so the loss you get there isn't a big deal. The "big deal" is in the low end. So, once again. The isolation that really matters is in the low end. That's why when you did the math for your isolation plan, you should have selected RF to be half of the lowest frequency that you need to isolate, and designed for that. That's for a 2-leaf wall. But for a THREE leaf wall, you have two resonant frequencies, F+ and F-. In that case, you should have selected F+ to be no more than half of the lowest frequency you need to isolate. Did you do that? I don't think so, from what I can see. In ALL CASES for an equivalent mass and wall depth, F+ on a 3-leaf wall ALWAYS, WITHOUT EXCEPTION be higher than RF for the equivalent 2-leaf wall. That's the point that you seem to be unable to grasp. The resonant frequency of a 3-leaf wall will, without the slightest doubt, always be HIGHER than the resonant frequency of a 2-leaf wall, all other factors being equal. So, also without a doubt, a 3-leaf wall will ALWAYS have worse isolation in the low end than the equivalent 2-leaf: This is not a debatable point: it is simple physics: There is nothing to discuss about it: it is just plain fact. So, to repeat, regardless of what you might expect from a 3-leaf vs. a 2-leaf, the 3-leaf will always lose out in the low end. Maybe this will help you understand: Take a careful look at that. Notice that, going from left to right, doubling up an ordinary wall with insulation (moving from the one marked STC-36 to the one marked STC-40) only adds 4 dB of isolation! That's less even than what mass law predicts, so clearly something is wrong right there. Move to the next one, where the total mass has been REDUCED by 25%, but the isolation GAINED ten points: Why? Because the wall went from 4-leaf to 3-leaf. Move across to the next one, and the wall loses another 33% of its mass, yet once again GAINS 7 points! How can that be? Because it went from a 3-leaf to a 2-leaf. On the final image, adding that mass back again, but in the right place this time, and the isolation jumps once again, by 6 points, which is what you would expect from mass law (but actually isn't due to mass law in this case...) Take a close look: the wall marked STC-40 has TWICE the mass of the one marked STC-57, yet has a whopping 17 dB LESS isolation. In other words, 4-leaf isolates about 50 times WORSE (in terms of acoustic power) than 2 leaf, even with twice the mass! Pretty surprising the first time you see it, and totally non-intuitive if you don't understand the principles, but absolutely true. If that doesn't convince you then try this: That's from the famous Wyle report of 1973, but just as valid today as it was back then. They call it "double panel" and "triple panel", but that means the same as "2-leaf" and "3-leaf" respectively. Note the curves: 2-leaf has BETTER isolation in the low end than 3-leaf, ALWAYS, and they both beat mass law hands-down, except in the very, very low end. No exceptions. Simple physics.

You mention that the existing cavity wall will probably act as a single leaf (115mm brick | 50mm air cavity | 115mm brick). It probably depends on the accuracy of the actual construction and to what degree the two brick skins are connected / isolated.

Did they use ties when they built that wall? If not, then it should not have passed inspection. If they DID use ties, then it is coupled, no debate. Even if they did NOT use ties, I can guarantee that you have multiple mortar bridges in there, and numerous points: that wall is flanking, for sure. If not, then it was built by the only bricklayer in RSA who ever bothers to clean out the cavity, and take special precautions to ensure that no mortar ever squeezes out the back, and no mortar ever drips inside the cavity! I'd bet that your wall is very well coupled, and flanking all over the place, unless it was specifically built as an acoustic isolation wall by a brickie who really knows what he is doing.

My thinking in the window plug was, the first layer fixed to the window frame, in conjunction with the glass / frame collectively will form the first leaf,

No, that forms two leaves: the glass is one, then there is an air gap, then drywall: that's two leaves separated by an air gap. And from the photos you sealed it well, so that's a really good resonant chamber you have in there. Resonant frequency: somewhere around 100 Hz, give or take a bit (assuming 15mm drywall and 8 mm glass).

rockwool insulates the cavity, and the internal plasterboard leaf, caulked and sealed forms the internal leaf

Actually, the third leaf. Two leaves of 15mm drywall over an insulated 50mm cavity gives you a resonant frequency of roughly 56 hz. Using the 3-leaf equations, I'm getting F+ = 1,446 Hz, and F- = 1022 Hz, but that doesn't make sense so I think I dropped a zero and screwed up the calculations by an order of magnitude somewhere, and they should be 144 Hz and 102 Hz. That sounds about right. That being the case, your window will not isolate reasonably until roughly 290 Hz, and isolate well at about 433 Hz. That's really lousy, any way you look at it! On the other hand, if you would have put all the mass as close to the glass as you can get it, and make the second leaf the as-yet unbuilt inner-leaf with the frame about 3" away, you would have gotten a 2-leaf MSM resonance of something like 13 Hz, meaning your wall would have isolated at 26 Hz, and isolated well at 37 Hz. Now THAT is a wall worth building...

The recess of the internal will be plastered to offer a flush wall surface.

It doesn't need to be plastered or flush: as long as it is sealed well, that's all that matters.

3. Add insulation (if you don’t have insulation)*greenglue.com

You should ALWAYS have insulation, regardless of whether or not there is GG in there, or any other factor. The insulation is the damper on your spring: it absorbs acoustic energy and converts it to heat. If you leave the insulation out, it will cost you a lot of isolation, perhaps as much as 16 dB.

The assumption to consolidate window / small void / first leaf into a collective first (external leaf) seems to be a major issue.

It's not a major issue at all: you cannot "consolidate" (whatever that means). Rather, you have to do the math, and see what effect it REALLY has, not the effect that you maybe guess it possibly might be that you would like it to have!

Hypothetically, removing the hinged windows and just leaving the frame in place would then reduce the system to a two leaf system?

Correct.

Is the air volume in between the window frame and first leaf significant enough to have a major impact on sound transmittion at that specific resonant frequency?

The volume doesn't matter: it doesn't even enter into the equations. The DEPTH matters and the MASS matters. If you decrease the depth (ie, make a thinner gap) then the resonant frequency goes up, which is not good. For very thin gaps, it goes up faster. ie, the relationship is not linear. Also, as the mas goes down, the resonant frequency goes up: If that is very thin window glass, and thin drywall, then the actual resonant frequency is much higher than what I quoted above: I'm assuming 15 mm drywall and 8 mm glass.

I suppose the bigger question is, to what extent could a system collectively be considered as a leaf? eg. cavity wall acting as single leaf? (I'm thinking the level/degree of coupling?)

Coupled is coupled, period. To give you an example, on a normal MSM wall one single nail bridging the air gap destroys the entire effect, and costs you many dB of isolation. It's the same with seals: a 1 square inch hole in a 100 square foot wall passes the same amount of energy as the entire wall, and destroys the isolation.

When I mentioned loosely based on a two leaf system for the window block, the logic was to have this correspond to the existing wall (cavity wall = MSM).

Then you should have kept the surface density on the leaves the same as the brick: In other words, you should have put roughly ten layers of 15mm drywall on each leaf. Drywall is about 60% less dense than brick so to get the same surface density as a 4 inch thick brick wall, you'd need about 6 inches of drywall, which is roughly ten layers. But you only put one layer on each leaf, so you only have about one tenth of the mass you needed to maintain uniform surface density across the entire wall surface...

Based on Mass | Cavity | Mass which offers thermal insulation as well as some acoustic isolation, this was the premise for the window block.

Thermal insulation and acoustic isolation are two different things. You don't have to deal with concepts such as tuned systems when dealing with heat transfer. You are right that the wall has some acoustic isolation as is, but it isn't due to it being a 2-leaf MSM system (which it isn't). It is simply due to the high mass. Fortunately, with two layers of brick in there, you already have plenty of mass, so the wall would have isolated pretty darn good once you built the inner leaf.... if only it wasn't for that faulty window plug... :)

My preliminary goal is to prepare the shell for phase 2, the implementation of a finalized studio design.

Like I keep on trying to explain: the inner leaf and the outer leaf act TOGETHER as a SYSTEM in order to isolate. You cannot design one without taking the other into account. They interact with each other to create the tuned MSM filter: Change one thing, and you change everything. Change the characteristics of the inner leaf wall, and that automatically changes the overall isolation. For example, if you change the design of one of the inner leaf walls from normal to inside-out, that might reduce your air gap from 6 inches to 2 inches, thus destroying your low frequency isolation again. So a simple design change on the inner leaf totally blows the entire isolation plan.... The entire room is a system: it needs to be designed all together, to work as one, not as a series of disconnected, independent bits and pieces.

The alternative would have been to rip out the existing ceiling, and completely replace this with the new ceiling.

An even better alternative would have been to rip it out and leave it out! :) End of story.

A silicone based material was applied in order to compensate for possible vibrations between the boards after fixing, and offer a even fixing surface under compression.

so basically you glued the board together with a product whose acoustic properties you do not know, have never been tested, and never been published? :shock: Why would you do that? Silicon caulk is not Green Glue. GG is specifically formulated to do one thing: constrained layer damping. That's all it does. It is not glue, never dries, and is not there to "compensate for possible vibrations between the boards". It is there to provide proper constrained layer damping. Silicon caulk, sealant, adhesive or whatever it was that you put in there has totally different properties, and will NOT accomplish the same thing. In fact, it most likely will make things worse, since you have trapped a very thin layer of air between two masses on resilient mounts: To me it sounds like you have created a series of panel traps in your ceiling, tuned to random unknown frequencies! :shock: Not only that, but according to your latest diagram, you used thin, flimsy 12.5 mm drywall on one of the layers, which introduces yet another issue: higher flexibility, lower mass, higher resonant frequency, lower surface density.... It's really hard to say WHAT is going on up there, but one thing is for sure: it ain't pretty! - Stuart -

29 (0x1D)

The wall / filter is tuned to s specific frequency and this is a function of the: Mass of leaf system (eg. plasterboard) Flexibility / Rigidity (of the plasterboard / membrane)? Insulation (Cavity insulation damping the spring effect) Depth (From inner leaf to outer leaf) Accuracy of seals (each leaf 100% sealed, providing a isolated cavity spring)?

Of those, only the mass and depth count. The characteristics of the spring (air), are assumed to be constant, and for most studio purposes that's fine (unless maybe you live at very high altitude, or something, where the air density / speed of sound would be different). Rigidity is only important below MSM resonance, so doesn't come into the MSM equation. Insulation simply changes the effective cavity depth, thus changing a constant. Seals don't come into it, since the MSM system must be sealed in order for the equation to be valid. Not sealed=equation won't give you valid results (and isolation wont work either). The simple form of the 2-leaf equation is: F = 60 / (SQRT (m x d) ) where: m = mass of the leaf d = depth of air cavity The constant changes from 60 to 43 if the cavity is insulation-filled (without insulation, the air acts adiabatically, but with insulation it acts isothermally. Another way of thinking of it: it makes the effective path length of the air gap greater). All figures are metric. The above equation assumes a single leaf over an infinitely massive second leaf. It is valid if you have, for example, a couple of layers of drywall n studs next to a thick concrete wall. The concrete wall isn't "infinitely massive", or course, but as compared to the drywall it is good enough for most purposes. If you have two similar leaves (ie, both leaves are drywall), then the long form of the equation is applicable: F = 60 [ (m1+m2) / (m1 * m2 * d) ]^.5 m1 = mass of leaf #1 m2 = mass of leaf #2 d = depth of air cavity Same rule applies to the constant, and all figures are metric. (In both equations the terms "mass" should really be "surface density", since that's the units you use.)

When aiming at 60Hz, in reference to what you mentioned about isolation in multiples of tuned frequency, you should be looking at a resonant frequency of 20Hz?

General rule of thumb is to aim for MSM resonance that is half of the lowest frequency that you need to isolate. So if your lowest frequency is 60 Hz (which is really high! :shock: Kick, toms, bass, keyboards are all in there...), then you would shoot for MSM resonance of around 30 Hz. The normal range o human hearing goes down to 20 Hz, so good professional studios aim to isolate down to 10 Hz. Not easy to accomplish for very high isolation levels.

I would like to calculate as an exercise the resonant frequency of the resultant plug to test the design before implementation

You'll need to know the surface density of the glass in that window, or at least the type of glass and thickness, to estimate it. - Stuart -

29 (0x1D)

how would you close off the residual gap in the window frame / door frame? I gather the idea is not to introduce a coupled wall, but you also would like a flush appearance and something to contain the rockwool from just sitting between the frames. My thinking a material with a very low sound transmission value? Light fabric spanning the width?

That's the normal way of doing it, yes. Or you can wrap a piece of 703 (cut to the right size) with cloth and fit that into the gap. Pretty much any kind of cloth that "breathes" well will work. Hold a sample over your mouth and blow gently: if you feel practically no resistance, then that should work, but if you notice that it is a bit hard to blow through it, then look for something else. Speaker grill cloth is ideal, but not cheap. Most reputable manufacturers of acoustic products sell some kind of acoustically transparent cloth, but basically anything that breathes is fine. - Stuart -