I Designed a Soundproof Wall That Couldn't Dry Out
SOUND ISOLATION DESIGN · SPYS DESIGNS
Sound Isolation Assemblies Are Moisture Traps by Design
What a roof detail on a Tallahassee project taught me about vapor, and why the sound isolation designer has to be the one who asks the question.
Two weeks ago I was on the phone with the builder on a project we designed in Tallahassee, Florida. I asked him how the roof assembly was going to handle vapor. He said, honestly, that never crossed my radar.
I told him the truth, which is that it had not crossed mine either until about a week before that.
What follows is what I learned, what I got wrong, and what we changed on a live job site before the walls were closed up. I am writing it down because I could not find this material anywhere in our field. Acousticians do not write about vapor. Building scientists do not write about sound isolation. The overlap is where the failures live.
What I got wrong about OSB
The wall as originally drawn had one layer of ⅝” drywall and one layer of 23/32 inch OSB on the interior face of a double stud isolation assembly. I drew it that way because OSB is would help us nail acoustic treatment to the wall without having to find studs or use drywall anchors. This is something many of you have probably thought of doing. Makes sense. The problem is that OSB does not let water vapor pass through.
The Engineered Wood Association puts OSB below one perm when the surrounding relative humidity is under 40 percent, which is where a conditioned studio lives nearly all of the time. Below one perm is a Class II vapor retarder. OSB only opens up into the two to four perm range once ambient humidity climbs past roughly 60 to 80 percent, and at that point the panel is already in the moisture range where things grow on it.
Gypsum board behaves almost in reverse. Bare gypsum is wide open, measured in the tens of perms. It is the latex paint doing the retarding, and latex painted gypsum board sits in the Class III range of one to ten perms.
So the wall I had drawn carried roughly the same low permeance on both faces, with a sealed cavity in between that nobody could inspect, in a climate where vapor drives inward eleven months of the year.
That wall would have met its acoustic targets. It also had no way to dry.
"The corrected wall. Semi-impermeable outside, vapor open inside, drying inward."
The only rule you actually need
Every assembly has to be able to dry in at least one direction.
That is the whole concept. You do not need to memorize perm tables. You need to look at a section and answer one question: if water gets into this assembly, where does it go?
The International Residential Code sorts materials into three classes by permeance:
- Class I, 0.1 perm or less. Polyethylene sheet, sheet metal, foil facers. A true vapor barrier.
- Class II, greater than 0.1 and up to 1.0 perm. Kraft facing, some smart membranes, and OSB at interior humidity.
- Class III, greater than 1.0 and up to 10 perms. Latex paint on gypsum board.
There is one more rule that matters enormously in our work and almost never gets said out loud. Permeance is inversely proportional to thickness. Two layers of a material is roughly half the permeance of one layer.
We build assemblies with four to six layers of sheet goods. Every one of them makes it harder for vapor to dry out through that material.
Why sound isolation makes this harder than a normal wall
Four reasons, and every one of them is a consequence of doing the acoustic job correctly.
Mass. Isolation performance comes from surface density, so we add layers. Each layer cuts the assembly permeance roughly in half.
Air sealing. We seal every penetration because air leakage is the largest flanking path in an isolated room. This is also, as it happens, the single most effective moisture control measure in building science, because air transport carries far more water than diffusion does. But it removes the assembly ability to forgive a mistake. A leaky wall dries in spite of itself. A sealed wall does exactly what you designed it to do, including the wrong thing.
The decoupled cavity. A double leaf wall or a room within a room creates a cavity that nobody conditions, nobody vents, and nobody can inspect once the finish goes up. That cavity can only dry through one of the two leaves.
Membrane products. Mass loaded vinyl is a filled vinyl sheet. Vinyl sheet goods are the material family building scientists have spent thirty years telling people to keep off the interior face of walls in hot humid climates, because they create a condensing surface exactly where you least want one. I have argued against MLV on acoustic grounds for years. This is a second and entirely independent reason, and it is one acousticians do not raise and building scientists do not connect to studio construction. If you are considering MLV anywhere in an exterior assembly, get the manufacturer permeance data before you specify it.
"Same acoustic assembly, three different correct answers."
The climate changes the answer. The acoustics do not.
We are designing in Florida, Arkansas and Colorado at the moment. The acoustic assembly is close to identical across all three. The vapor strategy is not remotely the same.
Florida
Arkansas
Colorado
Climate zone
1A to 2A, hot-humid
3A to 4A, mixed-humid
5B to 6B, cold-dry
Vapor drive
Inward, nearly year-round
Reverses seasonally
Outward year round, strongly in winter
Assembly must dry
Inward
Both directions
Outward
Interior side
Class III only. Never Class I or II.
Class III, or a responsive retarder
Class I or II required by code (variable permanence not poly)
Why
Air conditioning makes the interior the cold side almost every month.
Anything tight on either face traps moisture in one season or the other.
Dry outdoor air year round means the drive never truly reverses, but air conditioning makes the interior the cold side in summer.
The code language reflects this. The IRC requires a Class I or Class II vapor retarder on the interior side of frame walls in climate zones 5 through 8 and Marine 4. In zones 1 through 3 it does not require one at all, because in a cooling climate an interior retarder is the thing that causes the problem rather than the thing that prevents it. Variable permeance membranes are now the right answer in zones 4 through 6, where the vapor drive runs in both directions across the year. In our Denver example, the membrane opens in summer, letting a damp cavity dry inward toward the drier air conditioned space, and closes in winter, resisting the outward drive that would otherwise carry interior moisture into cold sheathing where it would condense.
The same detail is correct in Denver and a failure in Tallahassee. Nothing about the acoustics changed.
The project: the wall
The correction on the wall we were designing in Florida was a single substitution. The interior leaf went from one layer of ⅝” drywall and one layer of 23/32 inch OSB to two layers of 5/8 inch gypsum board with a latex paint finish.
Here is what makes that a design decision rather than a lucky guess. 23/32 inch OSB runs about 2.4 pounds per square foot. Five eighths inch gypsum runs about 2.2. That is a difference of roughly half a decibel under mass law, which is inaudible. Permeance went from somewhere near 0.3 perm to somewhere near 2.5.
We gave up half a decibel to get five to ten times the drying capacity. That trade is the job.
One detail that came out of this and is now going on every drawing we produce: the paint schedule is an envelope decision. If someone specifies a vapor retarder primer, an oil based finish, or a vinyl wallcovering on those walls, the building loses its only drying path. In a studio there is a real chance somebody later staples up MLV or a foil faced product because that is what studio people do. The interior finish is a permanent constraint, not a decorating choice, and the client needs to be told that in writing.
"The actual redline. This is a live set, not a diagram."
The project: the roof, which was worse
Having caught it in the wall, I went looking at the ceiling and found the same material in the interior ceiling layers. Except in the roof, the assembly above the rafters already had rigid foam and a self adhered membrane over it, both of which are effectively impermeable.
That meant the rafter cavity was sealed above and sealed below. No drying path in either direction, with fibrous insulation in the middle to hold whatever water found its way in, whether from construction wetting, a fastener leak, or humid air moving through a penetration.
Same correction. The OSB came out, a second layer of 5/8 inch Type X went in. The cavity now dries inward, which is the only direction available and the correct one in this climate. It also picked up a fire rating it did not have.
"The roof as originally drawn. Sealed above, sealed below, nothing in between but insulation and time."
This is not in my scope. I flagged it anyway.
Vapor management is not sound isolation design. I want to be precise about that, because the distinction matters more than the technical content above.
My position is that sound isolation design changes the hygrothermal behavior of the envelope. We add mass, we seal the assembly, we introduce a decoupled cavity. Every one of those moves reduces the ability of a wall or roof to dry. That makes it my obligation to flag the consequence and coordinate the answer with the architect or the envelope consultant. It does not make it my obligation to own it.
On this project there is no architect. The structural engineer stamped the structure, which is what a structural stamp covers. Plan review passed it. Nobody else was going to ask the question, so the recommendation went to the client in writing, along with a recommendation that the engineer of record or a building envelope consultant confirm the revised assembly before enclosure.
That is the whole difference between a design professional and a vendor. A vendor builds what is on the drawing. A designer notices that the drawing cannot work and says so, even when the contract does not require it.
The formula is easy. Software can run the formula. What decides whether a room works is whether somebody on the job is asking the question nobody assigned them.
Planning a sound isolated room?
We design sound isolated rooms across North America, and coordination across structure, envelope and mechanical is part of the work rather than an afterthought. If you want somebody asking these questions before the walls close instead of after, start with a Sound Isolation Site Assessment.
Request a Sound Isolation Site Assessment
Sources and further reading
APA, The Engineered Wood Association. Published permeance data for oriented strand board.
International Residential Code, Section R702.7, Vapor Retarders. Class I, II and III definitions and climate zone requirements for frame walls.
Building Science Corporation. Published guidance on vapor control by climate zone, including the treatment of low permeance interior finishes in hot humid climates.
