This is one of the most pervasive "fake acoustics" myths out there! If you look at YouTube, you'll find a whole bunch of videos there of people who are trying to float their home studio floors with a couple of 2x4s resting on rubber pads (such as the ones in the picture above), with a plywood deck on top... thinking that they are "floating their floor"... AAAARGH!!!!
Surprisingly, you even see that recommended on some supposedly high-quality websites and forums about recording studio design and building, coming from people who should know better, saying that this is a good way of building your studio floor! SIGH!
So let's get this clear once and for all:
YOU CANNOT FLOAT A FLOOR WITH A LIGHT-WEIGHT WOODEN DECK, ON RUBBER PADS. IT IS IMPOSSIBLE.
Full stop. End. Fin. Finito. That's it.
So, you must be thinking: "Then all of those people who show their videos on YouTube, about 'How I built my wonderful floating floor, real cheap', are wrong?"
"They don't know what they are doing?"
"They are all stupid?"
Well, maybe not stupid, but certainly mislead, or ignorant.
It works like this: If you want to "float" a structure (any structure) so that it can't transmit vibrations into whatever it is sitting on (which is what you are trying to do with a "floated floor" for your studio), then you need to be sure that it really does "float"! By "float", I mean that it is decoupled from the base that it is sitting on, sort of like a boat is "decoupled" from the river bed under it by all that water, or a plane is decoupled from the ground by all that air under it. The boat can rise and fall on the waves, completely unaffected by the river bed, and the plane flies along fine, not even aware of the ground below it. If there's an earthquake, and the river bed vibrates wildly, the people on the boat would not feel that, because they are "isolated" from that by the floating boat. Ditto for the plane: People in the plane would have no idea that the ground below them was shaking wildly, because the plane is "floating" on the air, decoupled from the ground.
That's what people are are trying to do with their isolated studio floors: they want it to "float" in the same way as that boat or plane, so that any vibrations in the room cannot get down to the base, and vise-versa: no vibrations outside can get into the room.
OK, that's the concept, and it's simple to understand in principle.... but its a bit more complex in practice! Just to clarify: the concept is fine. and correct! It's the implementations you see that are silly. And ignorant. And sometimes illegal. And occasionally even dangerous.
Here's the thing: an object can only float if the thing that it is floating on is "soft" enough (water, for the case of the boat, air for the plane, and springy-rubber pads for those studio floors). This is basically just a weight sitting on top of a spring, just like a car sitting on top of it's suspension system:
That's the exact same concept needed for floating a floor, or a room. The spring must be "soft" enough that the car can "float" on top of it. If the spring is too hard, then all of the bumps and jolts from the road will be felt inside the car. And the same happens if it is too soft: If the spring is so soft that it is flattened out by the weight of the car, then it "bottoms out", and does not float. Once again, all of the bumps will be felt inside the car. So if the spring is not chosen correctly for the weight of the car, then it won't work: the car won't float. You need a "Goldilocks" spring: not too hard, and not too soft... it has to be just right!
The same applies to a studio floor (of the entire studio, if you are attempting to float the complete room, not just the floor). The "spring" that it sits on must be right for the weight of the floor (or room). If it is too soft or too hard, then the floor does not actually float, and all of the sounds and vibrations will get through, in both directions: Loud sounds outside will get into the room, and loud sounds inside will get out..
Now, to complicate things a bit: if you have a weight sitting on a spring, then that is a "resonant system": It wants to resonate at a specific frequency. It wants to vibrate.
It is actually easy to calculate that frequency: All you need to know is the amount of weight sitting on the spring (also called "mass"), and how "springy" the spring is: technically, that is called the "resilience" of the spring. If you know those two things, then you can calculate the resonant frequency. In simple physics, this is a "Mass-Spring" system.
Also, there's this issue of "resonant frequency". Your floor can only float for vibrations above a certain frequency. There is a lower limit to the frequencies that it can isolate, and that is the resonant frequency of the floor. Because: for vibrations at the resonant frequency, it does not isolate at all, and even worse, it can AMPLIFY those vibrations, making them much worse.
A classic illustration of this "resonant amplification" is the famous Tacoma Narrows bridge disaster, of 1940:
(NOTE: Some web browsers are not able to show this in motion: if you just see a still image, or even just a link, then then click here instead )
That animation shows a form of resonance. In the case of this bridge, it is aerodynamic flutter resonance caused by the strong winds blowing over the bridge, and you can clearly see how the resonance has greatly amplified the tiny original vibration, to a huge amplitude. That started out as just a slight little wobble in the bridge, but the resonance made it bigger and bigger. Clearly, the bridge is not "isolating" itself from that resonance! Here's the final result of that resonance: Not surprisingly, the structure of the bridge was unable to withstand such huge motion, and it failed spectacularly in November 1940, less that 6 months after the it was opened! During those 6 months, the bridge wobbled and resonated very often like this, so much so that it became known as "Galloping Gertie".
What does that have to do with your floating floor? Hang in there... you'll get the point soon... (Don't worry: your floor is probably not going collapse from resonance, like the bridge did! Your floor just wont work.)
Just like the bridge, your floor will have a "resonant frequency", and at that specific frequency, it will not isolate the sounds in your room at all. In fact, just like the bridge, it will amplify those sounds, making them LOUDER, not quieter. For mathematical reasons, it's not just exactly at the resonant frequency that you get amplification, but also for all frequencies around it, for about half an octave above and an octave below. So, for example, if your floor happens to resonate at 100 Hz, then it would fail to isolate for half an octave above that (up to about 150 Hz.), and an octave below that (down to about 50 Hz). Thus, for the entire range of 50 Hz to 150 Hz, your floor would make things louder, not quieter. (To be more precise, the upper limit is not exactly 1.5 times the resonant frequency, but rather 1.414 times the resonant frequency. But 1.5 is a good approximation). So, any note that you played in your room in that range, would be heard outside the room LOUDER than if you had not floated the floor...
Don't believe me that a badly floated floor can amplify sounds? That's a graph from actual tests done in an acoustic lab, of an OSB deck assembly sitting on various types of resilient pads. Notice that the horizontal line is at 0 dB, indicating neither amplification nor attenuation of the sounds, and the vertical scale is "reduction in sound pressure level". So all the positive numbers on that scale (everything above the 0 dB line) show isolation, and the negative numbers (below the line) show "negative" isolation... which is amplification. You can see that all of the things they tested here show gross failure to isolate at all, for all frequencies below 500 Hz, and that all of those floors have resonant frequencies between about 100 Hz and 250 Hz, so they actually AMPLIFY sound in that region, some by as much as 10 db! Thus, if you built your floor like that, it would be twice as loud as if you did nothing at all... Not one of those floors shows any useful isolation until over 1,000 Hz.
Therefore, it would be a good idea to design your floor such that the resonant frequency is much lower than the lowest frequency you need to isolate! That makes sense. In fact, the general rule of thumb here is to design it so that the resonant frequency is not just half an octave lower, but rather at least one full octave lower (dropping by one octave means that you are going down to half the frequency). So: if you plan to play six-string bass in your room, where the lowest frequency is 31 Hz, then you should design your floor to have a resonant frequency not higher than 15.5 Hz.
So far so good! This is not hard to understand: your floor is going to be resonant; at resonance it amplifies; therefore design it such that it won't amplify any frequency that might happen in your room. Simple!
As I said before, there are very simple mathematical equations for calculating that resonant frequency, and the only two things you need to know are how heavy the floor is (mass) is, and how "springy" the springy is (resilience). That's what the term "spring constant" is all about. Its the technical term for "springiness". Because if the spring is too "hard" (not springy enough), then the boat won't float! And also if it is too "soft".
Think of this: if you have a heavy weight sitting on a spring, and it is so heavy that it squashes the spring completely flat, then obviously, that isn't going to float! Its not going to isolate. Also, if the weight is so light that the spring isn't compressed at all (maybe with just a feather on it), then it also won't float... and it won't isolate.
Simple illustration: If you take the suspension spring out of a truck and try to use it for a small car, such as a Mini or a Smart Car, then it wont float: the car is just not heavy enough to compress the spring at all: that spring is too hard for the job. And also if you take the spring from a Smart Car's suspension and try to use it for the truck's suspension: the weight of the truck would squash the spring flat: that spring is too soft. You have to choose the right spring in each case, so it is "just right". Not too hard, and not too soft. (Just like Goldilocks! )
Here's the spring from a Smart Car: Its the yellow coiled wire thingy. Trucks don't use coil springs, like the one above: the weight would squash it flat. Instead, trucks use leaf springs, like the one below. A leaf spring is basically just a stack of thick steel plates, with the ends bolted to the truck body, and all of that sitting on top of the wheel axle, which is bolted to the middle of the spring: From those two pictures, its pretty clear why the weight of a Smart Car wouldn't budge a truck spring at all, not even slightly.... and also why the weight of a truck would totally flatten a Smart Car spring.
So: You have to use just the right type of spring for the weight, in order to make it "float". In fact, it turns out that you need to use a spring that will compress by about 10% to 30% when it is fully loaded with all the weight. If it compresses more than 30%, then it doesn't float. If it compresses less than 10%, it doesn't float. And that range also depends on the type of spring: for example, neoprene rubber has a different range than EPDM rubber, Sorbothane rubber yet another range, and a steel spring is different again.
As you can see, it's not just as simple as putting any old spring under any old weight! (such as putting a plywood deck on rubber pads....) There's math and physics to think about, equations to do. It is very important to get the right spring for the weight of your floor. If not, then either your floor will over-compress the spring, and it will "bottom out", meaning that it doesn't isolate... or it will "under-compress" the spring, so it "tops out", and it won't float... and won't isolate. You have to get it right, if you want your floor to actually be isolated.
It would be rather sad if you spent all that time and money and effort to build your floating floor, then when it is finished you find that it does not actually isolate, because the spring you used was too hard, or too soft...
So this seems very simple then! All you have to do is to figure out how much mass you need to put on your floor, and what type of springy rubber you need, and all will be fine! What's the problem? Surely, you just need to buy some of those rubber pucks, put some 2x4s on top of them, then nail down some nice heavy plywood, and the whole thing will float nicely! Right? Just like in the YouTube videos? It MUST be like that, because I saw it on the Internet! Just like this!!!!!
If everybody is doing it like that, then why do I keep saying that it is impossible, and it won't work?
The sad truth is, that in order to float a floor properly, you need a huge amount of mass, and the exact right spring. 2x4 joists on rubber pucks is way, way short of what is needed. The graph below shows the actual, real, true, tested, situation, not YouTube fantasy. You can see for yourself why bits of rubber and 2x4's don't work.
That graph is for the case where you use air as the spring. Air is actually a great spring, believe it or not: far better than a coil of steel, or a piece of rubber. It is much "softer", and that's a GOOD thing! You want the spring to be as soft as possible (within reason). The harder the spring, the less it isolates.
So, that graph above shows four different curves, for four different weights of the floor deck ("psf" means "Pounds per Square Foot") over an air spring cavity. On the left side, it shows the resonant frequency that you would get for various combinations of those two. As I already said above, the goal is to get your resonant frequency to at least on octave lower than the lowest frequency that you'll have in the room. As I also mentioned, "one octave lower" means half the frequency (musicians talk about "octaves"; acousticians and engineers talk about "frequency", but it's the exact same thing in different words: For any give note, the same note one octave higher just means that it is twice the frequency. Middle C in music is about 261.3 Hz, that is also known as "C4". If you go up one octave to C5, then the frequency is twice as high, so 522.6 Hz. If you go DOWN one octave from Middle C, to C3, then that's half the frequency: 130.6 Hz.).
OK, so we need to go one octave lower than the lowest tone you will play in your room. Let's use the graph above to figure out what floor you would need in order to achieve that. I used the example of 31 Hz before, which is the lowest note on a 6-string bass guitar, so let's go with that: if you want to isolate frequencies at 31 Hz, then you need to make sure that the resonant frequency of your floor is less than half of that, or less than 15.5Hz. So look at the left side of the graph, locate the spot where 15.5 Hz is (call it half way between 10 Hz and 20 Hz), and draw an imaginary line across the chart there. You can now see that no matter how deep your air cavity is, the top two dashed lines are no use: you can never get a low enough frequency if your floor only weighs 5 PSF (pound per square foot) or 10 PSF. Not possible. However, at 30 PSF it is possible (the dotted line, third from the top): it looks like you would need to have an air cavity that is at least 4.5 inches deep, so you can't do it with 2x4's, as they are only 3.5" deep. You'll need to use 2x6's (which are 5.5" deep). Your other option is to go with a heavier floor: the bottom curve on the graph, labeled 60 psf (solid line, not dashed). With that option, you can get a frequency of 18 Hz. with a cavity about 2" deep, so you could use 2x4s there.
So those are your options: you can build up your floating floor on 2x4s with a 60 PSF floor, or 2x6s with a 30 PSF floor.
And that brings up the obvious question: What would you need to do, to get a 60 PSF floor? Well, let's consider OSB: the density of OSB is roughly 610 kg/m3, which works out to about 3.2 PSF for every inch of thickness. So to get 60 PSF using OSB board, you'd need to make it about 19 inches thick! In other words, you'd need to have 31 layers of 5/8" OSB on your floor, to get enough mass. But if you wanted to go with the 30 PSF option, you'd "only" need 16 layers of OSB to get there....
Now, think about this: if the floor area of your room is 100 square feet, at 60 PSF the floor would have to weigh six thousand pounds!
OK, so OSB isn't gonna work here . Maybe plywood? A sheet of really thick 3/4" plywood weighs about 3 pounds per square foot... it's a bit lighter than OSB. Thus, You would need TWENTY LAYERS of 3/4"plywood to get enough mass on your floor!
So how do you get such a high mass? If you can't do it with OSB or plywood, then what do you need? Simple: Concrete. The density of concrete is around 2400 kg/m3, which is roughly 12 PSF for each inch of thickness. So a concrete slab just 3 inches thick (36 PSF) would let you do it with a 3.5" air cavity, and if you went up to 5" thick concrete slab, you could do it on a 1.5" air cavity.
Now you can understand why those folks on YouTube don't have much success... And why I keep on saying that it is impossible to float a floor with a light-weight deck. The only reasonable way to get enough weight here, is by making a 5" thick reinforced concrete slab.
That's the plain, hard, cold facts. You cannot float a light-weight deck and expect to get good isolation for low frequencies. The laws of physics prevent it. It is physically impossible to float a light-weight deck consisting of just a couple of sheets of OSB or plywood on 2x4 studs. If you did that, the resonant frequency would be around 42 Hz, so the floor would amplify kicks, toms, bass guitar, electric guitar, and keyboards! It would only isolate from about 84 Hz upwards.
Andre just posted a great document in another thread (click link here) It shows the test results for a proper floated floor, as tested in a proper acosutic test lab. Summary: that floor weighs "only" 125 PSF! So a 100 square foot floor would weight 12,500 pounds!
That's what it takes to do it right. That's why we laugh at the poor guys in the YouTube videos, with their rubber pucks and plywood decks. Especially this little gem, which I found yesterday on an acoustic consulting web-site, where they supposedly teach you how to "soundproof" your floor:I kid you not! That really is the published advice of a so-called acoustic expert, on how to use U-boats. It doesn't take much imagination to figure out just how silly that is... They didn't even read the manufacturer's instructions on how to do it.... That's not only acoustically dumb, but structurally unsound, and likely illegal. I doubt that any building inspector would sign off on such a ridiculous setup.
Anyway, back to reality....
Another important point: that graph I showed above assumes that the "deck" is fully isolated from the underlying sub-floor, and that the only "spring" in there, is the air in the cavity. In real life, that is not possible (unless you can figure out how to make bags full of compressed air to support your floor): you need some type of resilient mounting to decouple the deck: it might be those famous "U" shaped rubber pads you see in all those YouTube videos, or metal springs, or something else, but there has to be something that disconnects the deck from the sub-floor, mechanically. Which makes things worse! The rubber or metal spring works in parallel with the air spring, and that REDUCES the total "springiness". So you actually need a deeper cavity to get the same frequency...
But there's more to this insanity, so let's take it one step further. Let's go back to the light-weight YouTube deck (pretending that the above graph does not exist, and imagining that it might be possible to magically get the right frequency with just two layers of OSB). We already know that two layers of OSB weighs about 6 pounds per square foot, so let's say we do some calculations for magical rubber pads, made of purest snake oil and pixie dust, and arrive at the conclusion that we need four pads that measure two square inches each, for every square foot of floor, and with a load of 6 PSF, that will float just fine, with exactly 15% compression. Great! Amazing! The floor floats! ... Until you stand on it.... Assuming you weigh about 180 pounds, and that your weight will be spread across four square feet of floor, just by stepping on that floor you increase the loading from 6 PSF to about 50 PSF Gulp! I think you see where this is going.... You just flattened your rubber pads into oblivion! They are now squashed flat, and don't float.
So you think creatively, and decide that you don't need the floor to float when you are not in the room, it only has to float when you ARE in there, so you re-design it to float when the load is 50 PSF. Fantastic! Wonderful! It floats! .... until you bring in your guitar, amp, a couple of pizzas and a crate of beer... now the load on that spot on the floor, where you are standing, is 65 PSF, and the floor doesn't float....
So you wrack your brains, and re-design the rubber pads yet again, so they float at 65 PSF.... But then you invite your buddy over to join you for a jamming session, and he brings his girlfriend, another amp, more beer, and a suitcase, since he's going to stay the night.... and now you have a load of 80 PSF....
OK, so I'm exaggerating a bit here, but I can keep on adding scenarios here, such as the desk, chair, couch, your DAW, other gear, etc. etc., ... however, you can see the problem: The load on a light-weight deck varies so enormously that it just is not practical. But with a 5" concrete deck, that has a much, much higher density and much higher mass, this is not a problem. Putting all that extra load on the floor, or taking it off, only changes the total mass by a few percentage points, and the floor still floats: the springs are still inside their optimal range.
So that's the issue. Floating a light-weight floor is not a viable solution. You need huge mass to float a floor successfully. It is certainly possible to float a floor, and companies like Mason Industries and Kinetics make devices to do that, but it only works with very high mass for the floor deck, such as 4 or 5 inches of solid concrete.
Here's what a properly floated floor looks like:Those are prefabricated floors: thick concrete on metal frames, resting on proper steel springs. Here's a close-up of the springs:
I mentioned Mason Indistries above: here's some of the devices they sell that allow you to float your floor properly: Those are rather nice, actually, because they allow you to put down a sheet of plastic on your current slab, spread a few of those devices around at the right locations, then just pour the concrete, right there. After the concrete has cured, you then "jack up" the entire floor, by using large screwdrivers or wrenches on those things. As you turn the mechanism, they slowly lift the floor to get the right air gap under it, and also allow you to level the floor. Smart.
A couple of years ago, Mason published a white paper about this subject, but on an even larger scale: how to float an entire building! Such as the Carnegie Hall, for example: they had to float ALL of that complete performance space, because the subway runs very close by, and they needed to stop the vibrations getting into the hall. Interesting read! (click here to view it) )
Galaxy Studios in Belgium is another great example of how to do this right: They went to extreme measures with their isolation. They floated the whole room! The complete facility is basically a concrete bunker, and each room is a concrete box floated inside that bunker, on both steel springs AND neoprene pads (to maximize isolation at all frequencies), and with a huge air gap underneath. Here's what that looks like:And here's the interior of one of the live rooms at Galaxy: That entire room is floated on multiple steel springs, like the ones above. Galaxy gets an incredible 101 decibels of isolation for their studio.
That's how you do it, if you want it to actually float for real! Now you can see why I just have to laugh at the YouTube videos of kids proudly "floating" their rooms on rubber "U-boats", with plywood decks on top of 2x4s...
(Aside: You might be thinking; "But I've seen those u-boat things advertised! You mean they don't work? Don't the manufactures have to test them or something before they sell them?" You would think so, but here's a statement from the website of one such manufacturer: "U-Boats have been placed under floors supporting quite a bit of weight. However, testing for compressive tolerances is expensive and to keep U-Boat costs down, we have never verified performance." So: no, they don't test them. That statement seems to have been removed from their website recently, but you can still find it on the "Way Back Machine" Internet Archives. In other words: by their own admission, the manufacturer never even bothered testing these things, to see if they work!)
Yes, it's complex. Floating a floor, or a room, is a Big Deal. It CANNOT be done successfully by just putting down some random bits of rubber thingies with a wooden platform on top! There's no way that can work. People who do not know how to float a floor or how to do the math, make these silly videos using rubber pucks, tennis balls, bicycle inner tubes, old car tires, U-boats, and other various bits of random rubber: that is just plain silly. Especially when the floor on top of that is merely a layer or two of plywood with some laminate flooring, or carpet, or vinyl tiles on it! It's a joke in bad taste. Anybody who actually did that in real life will end up with a failure: a floor that does not float, and is probably illegal anyway. There's a thing called "building codes", which are laws and regulations that govern how you are allowed to build something. Putting unapproved materials under your floor is going to cause you major problems with that.
Finally, there's another issue here: the structure of your house is probably not even strong enough to be able to do it! Most likely, your house, apartment, or high.rise office can't handle it. What type of floor do you have at present? Is it a concrete slab? A wooden floor? Something else? And what is underneath that floor? Does it rest directly on the ground, or is there another room down there below you?
If your existing floor is wood boards over wood joists, then it is pretty much impossible to float your studio on that, as a wooden floor would not be able to support the huge weight of a fully floated studio. Even if you have a concrete floor, it still might not be possible, especially if you are on an upper floor of a building, with other rooms below you: even concrete has limits. You would have to hire a structural engineer to come take a look at your place, and tell you how much additional weight you could safely put on that floor. The only time you would probably be OK, is if your floor is a concrete slab that sits directly on the ground ("slab on grade"). And if that is the case, then you don't need to float the room in any case! A concrete slab on grade is an excellent studio floor, with good isolation all by itself, so there would be no need to float your room.
And to end off this discussion of floated floors, it's important to know that you don't even need to do it anyway! (See here: why you do not need a floating floor .) The vast majority of home studios do NOT have floated floors... Firstly because it is usually not needed, secondly because it is really, really hard to do right (as you can see from the very simple explanation above), thirdly because you need a huge amount of weight.... that your current floor probably could not handle anyway, and would collapse... and fourthly, because it is insanely expensive to do it.
- Stuart -