I've received a bunch of messages about door soundproofing and thought it would be helpful to share what I found in my research into the topic. I live near a well known acoustics testing lab and began exploring a few things. First was just the ability to buy a soundproof door. The cost was very high and the complications far greater than anticipated. Then I looked into manufacturing a custom door and having it tested at this lab (They were nice enough to deter this waste of money and give me access to existing test data). This will ensure that if I do ever go down this road of having a door maker manufacture such a custom door, I have good reason to believe it will have high TL to begin with before spending time and money having it tested.
Ok first some of the problems.
- Doors are nearly always the weak point in a soundproof wall
- Door seals are rarely the weakest point in a door that was not constructed for sound proofing
- The door jambs are a flanking path and often the weakest link in the door structure
- High STC doors installed in high STC walls require highly massive door frame installation, as in with cement and metal
- High TL over a wide bandwidth requires a large air space, a single door cannot achieve this
- Thick, massive, solid structures can transmit sound readily. Remember that sound transmits easily through a foundation
- Stiffness is the enemy of TL above the resonant frequency and below the coincidence frequency, undamped resonances further reduce TL at the top and bottom of the the curve
The above truths tell us a few things we need to know. First, that a door creates a hole in our walls that becomes the weakest link. Therefor, like any other penetration, the smaller the hole the better. Large doors, worse yet, double doors (french doors for example) are a bad idea. Small doors aren't as dramatic looking and make the movement of large equipment more difficult.
The next thing we can infer is that a communicating door assembly is the best option. There is no single right way to do a communicating door, but there are certainly wrong ways. Any distance between the doors is good, it creates that airspace I mentioned above and allows a door that is more similar to the wall (remember that a party wall has drywall on two sides typically and the TL of that wall is partially dependent on how thick the wall is). There is no official minimum distance, but anything greater than 12" will cause the two doors not to "see" each other acoustically anymore. This isn't a problem per say, but it does mean that 12" is a good maximum. Once you go to say 3 or 4 feet, you are creating a hallway. If you have a hallway, then you do introduce some other problems to consider, but for the most part this is the best solution. You want to ideally offset the doors so they are not inline with each other if possible. In addition, acoustic insulation on the walls, doors, and ceiling are helpful. Remember that this adds a few STC points to a wall structure, same with the doors. The difference is small, but something is better than nothing.
Door Frames, the Weakest Link
The door frame is an obvious weak link, and I'm often surprised that more people don't comment on this issue. Door frame's are typically shimmed in place and this leaves an airgap around the frame. The traditional construction technique would drywall the opening and insert the door afterwords. The space around the door is then covered with moulding. This is a big problem and creates an obvious flanking path for sound around the door. There are a few solutions for this. First is that you can install the door frame before drywall and then drywall over the door frame. If you do that you need to deal with an exposed drywall edge, but it allows you to seal the door frame to the drywall and puts a lot of mass. You need to caulk this as well. Another option is to cement the door frame in place once its shimmed, then "glue" the moulding around the frame with acoustic caulk.This is similar to the technique used by studies with steal soundproof doors. The metal frame is cemented in place. It is not uncommon to use a damped polymer cement, which is why I did. It's an additive that makes the cement more flexible, and of course, damps the cement acoustically. The Drywall approach is quite frankly cheaper and easier and I suggest that as a better option.
STC and TL for DOORS
Ok first let's talk about common STC values for doors. MDF has an STC value of about 25 for panels 16mm or greater. Oddly enough, little data is available for thicker panels and what data does exist suggests that thickness doesn't matter much, the STC doesn't increase. TL values are pretty constant. Mass law tells us this shouldn't be true, so whats the reason? The primary causes are speculated to be a result of the structural resonance of the slab assembly and the coincidence frequency. There is a dip in TL at those two frequencies and it doesn't change with thickness. The only change is the frequency, but making a door also makes it more rigid, so these two factors fight each other, and in fact thicker slabs often have a higher resonant frequency. Also keep in mind, plywood and solid wood both have STC values of closer to 15-20 points, quite a bit lower than MDF.
Additionally, there are many solid core doors in which the solid core material is made from light weight wood products or fiber products that have low STC values. As a result, these assemblies have STC values in the low to mid 20's.
Notice that DorCor and Particleboard have similar STC numbers. DorCor is a low density product and I suggest avoiding it for theater doors. I almost fell for this myself.
These are commonly used as theater doors, but they are not good sound proofing doors. They have a relatively low STC rating. The core is noted as "low density."
Notice in the first link that the highest STC door value is obtained with a mineral core raised panel door that is 1 3/4" thick. The STC value is 35, and in fact, for cheap doors, its about as good as I could find. I have one of these for my doors and one flush series 1 3/4" particleboard door, which actually has a lower STC of about 32. The mineral core is actually a core similar to that of drywall, it is denser than fibercore or MDF and it has better internal damping. That is why the door has the highest STC and in fact, this is how most true high STC doors are made. With a thick damped mineral core, not MDF.
This is an interesting study done by the BBC on door slabs and its helpful to see that thicker wood doors really don't help the situation much at all, and in fact, damping is only marginally helpful. Their best door had a lead lining originally but they found that damping MDF worked about as well, as did an air gap, i.e. a hallow core door. Again, this shouldn't be shocking, especially when you see the Masonite data that shows a hollowcore door and a solidcore having similar STC values. Also notice that the bitumen vs just same MDF aren't that different, that the damping layer is reducing the coincidence dip, and that is the primary advantage. In fact studies into CLD have shown that in many cases damping does not add any TL below coincidence or above resonance. That its only benefit is damping the negative effect of the resonance. In some cases ,which well damped materials, there is no advantage at all, the only difference can be attributed to the change in mass.
This may lead you to think that the clear winning option is to add a sheet of MDF with green glue to a slid core door and be done with it. This is actually not a terrible option, but a few things to note. Some here have done this. The gain in transmission loss over the standard solid core doors is modest but meaningful, at low frequencies there will be no difference, but at mid and higher frequencies the difference will be about 3-6 db's more loss. You may see as much as 9db's of loss gained at the coincidence frequency of 3.8khz. Any damping of the door would make a difference however and in fact this sandwich only adds about 3dbs extra loss at the coincidence dip and nearly nothing at other frequencies. Consider that most people have a door that has an STC value of 25, a typical wall will be STC 40 to STC80 (Depending on construction). That means the door is not STC 28, the wall is still twice or more as soundproof. The door is still a huge weak point.
Notice that the winning door they tested was a hollow core door constructed of two sheets of 18mm and 9mm MDF with bitumen damping and wool insulation in the hollow core. Clearly not what we are used to seeing. This is a better door option than simply adding a big sheet of MDF to an existing door with green glue as it provides the best sound isolation. This appears to follow an STC curve that may be in the high 30's to low 40's. This is better than most commercially available doors.
Exterior Entry Doors ARE NOT better than interior doors for soundproofing
There is a common belief that exterior entry doors provide better sound isolation than interior doors. This just isn't true. They actually roughly equal an interior door, and in many cases are worse. They do have a benefit, many come with a threshold and gaskets pre-installed, but these gaskets are not designed for the best sound isolation. A lot of exterior doors are insulated and low mass, with poor damping, and this leads to relatively low STC ratings.
Here is Masonite test data on their entry doors. I've looked around and found comparable numbers from others. Some have doors with much higher values, but often these same companies make similar interior doors with much higher numbers as well. For those who noticed two entry doors that show much higher STC values in the 40's, note that is tested with a storm door. That is a communicating door assembly with a very low mass second door. That should tell you something. An airspace and second door makes a huge difference.
Mohawk has an STC rated door, it is about 10 points higher than their other options:
but we are still looking at just STC35. Note it is a mineral core door.
Communicating Door Assemblies are the Best!
There is just no getting around this. Two doors is always better than one. This site gives the best approachable information on the topic I could find.
This specialty door manufacturer has high STC rated doors and does so only with communicating doors.
I could not find any test data on communicating doors made with standard MDF core doors. Based on some proprietary test data (Not proprietary to me, but I can't share details) and various links I have found, two doors back to back would give a 3db addition to STC. Adding 6 inches will add about 3db's. Each additional 6 inches adds about 1-3 db's. At some point you hit a limit, no more gains to be had without upgrading the doors. My own door assembly has a 12" air gap and a roughly 45 db average transmission loss. It should be noted that at 45 db's of loss or more, I am at or below the measurement limit that is possible. I cannot measure below 40db's reliably at most frequencies, and the ambient noise outside my theater is at or above 45 db's. That means that I must play my stereo at or above a constant 100db's, and while the system can handle it, that puts high stress on the system. The transmission loss at high frequencies is so great that I cannot measure it, I am at my noise floor.
Gaskets for Doors on Soundproof Walls
First, look at this test data
and notice the benefit gained from the various gasket sets for a given door's STC rating. Now look back to the values we found for typical doors we use. Standard silicone stick on gaskets with acoustic thresholds work just as well as any other option on doors up to STC 42, and in fact, there is little to be gained up to STC 45. The adjustable gaskets and adjustable automatic door bottoms do not provide benefit until you hit about STC 49 and above. Notice that not even the custom BBC door met that criteria. That tells us that unless you bought a custom high STC door, these gaskets are not a huge benefit.
I think people should buy what they like, the kits are easy to use and you know they won't be the weak point. However, they are also expensive. I suggest that people consider the stick on gaskets and thresholds first and if looking at two different options, one more expensive than the other, don't assume you get any acoustic benefit. There is no acoustic benefit to the expensive gaskets and bottoms over much cheaper products.
High STC ratings above 50 always are hurt by even the best gaskets, and so the only option becomes more advanced solutions like multiple gaskets and vault like closures. Often these doors are large, thick, metal, and use double or triple seals that close with magnets.
Myths I wish would die
Key Take-away, what should you do
- Outside doors are better for sound isolation than inside doors
- Simply adding more mass increases transmission loss
- Door gaskets are the weak link
- The door jamb doesn't matter
- You can achieve equal or better results to a basic communicating door assembly with a single heavy and thick door with quintuple gaskets
- A thick solid core door with 1/2" MDF and Green Glue can provide huge benefit over a standard door (As mentioned earlier the benefit is actually small)
To be honest, probably the best option for most will be either a)Cheapest bang for buck is the communicating door approach with large as possible air gap and high as possible STC rated doors, or b) buy a professional high STC door and install properly. For those doing the Communicating door approach, damping the door assembly will increase transmission loss at coincidence frequencies. Doing so with a single door is ok, adding another layer of MDF and green glue will certainly be better than nothing, but it's important to know that this is offering only 3-6db's of additional transmission loss and only at certain frequencies. It will not stop any more bass, and in fact, because of the added stiffness may block less bass.
Possible DIY Option
My idea for a possible ultimate DIY door would be to create one using the approach the BBC tested, but enhanced with additional materials. I would consider creating a door with half inch MDF faces and framing. Total door thickness should be as thick as possible, but a minimum of 2" and preferably 3"-4". The core should laminate 5/8" drywall and lossy adhesive (like green glue) to the MDF faces. The core should have a 1" air gap with fiberglass insulation. Both layers of MDF should be damped with greenglue and Drywall if possible. That of course creates a problem, as the door would end up being more than 3.5" in this scenario. It is possible that only 1 layer of green glue and drywall is used, and that 1/4" masonite and green glue are adhered to the other layer. Bitumen damping could also be used for the MDF damping. The next hard part will be the door frame. It should be made of MDF and will need to be installed in a manner that is acoustically sealed, damped, and massive. As I said, I used a damped cement adhesive product for my two door frames, again, with an airgap. I further glued the moulding in place to ensure a good seal.
I would like to try building such a door as mentioned above. I plan to try making one that fits in my existing 1 3/4" slab door frame and so will need to use thinner material. I also want to have it tested if possible as I think it would be great to know that it works.