Automotive Sound Deadening with RAAMmat - Pseudoscience on the 2008 Subaru Forester XT
It is well known that Subaru vehicles are poorly insulated. Slamming the door of one is often likened to dropping a bunch of pots and pans on the ground - it's very 'tinny' to say the least. My forester was no exception, but I was recently doing a full stereo install, which meant that I had the doors taken apart to install new speakers anyways. Thus, I made the decision to 'deaden' the doors.
Sound deadening is a term that refers to adding mass to a resonant surface (like a car door) to reduce vibrations and hence reduce the transmission of sound waves entering the vehicle from outside sources like road noise and wind. Most sound deadening products available on the market are sheets of thin (~1/8"), but dense material. The easiest products to work with are those which have an adhesive backing, like the product that I used - RAAMmat. RAAMmat is a product of raamaudio.com and, after a bunch of research, seems to offer the best bang-for-the-buck. I bought their pre-packaged bundle #1, which came with 20 sheets of RAAMmat BXT II (for the primary sound deadening) and 1 yard of PS Ensolite (a closed-cell stick-on foam used to reduce higher frequencies).
Before I began any of the stereo installation, I found a road and took some sound measurements. Using just a couple of iPhone apps ('Decibel 10th' and 'Spectrum Analyzer'), I took measurements of both the frequency spectrum (more later) and the overall sound pressure level (SPL).
Screen of Decibel 10th app:
Screen of Spectrum Analyzer app:
Just a couple of pictures of one of the doors:
The picture below shows the rear door with the RAAMmat BXT II installed. It is also installed on the inner door skin as well (behind the sheet that is visible in the picture).
On top of the final layer of RAAMmat is a layer of PS Ensolite.
In addition to deadening all 4 doors, I did the hatch door as well as the hatch floor (beneath the spare tire). I still had a few sheets of RAAMmat leftover when all was said and done.
After the install, I immediately noticed the car felt more solid. Even before I drove the car, the doors now felt much more dense. Slamming the doors now offer a very satisfying 'thud.' After driving the car for a bit, I did notice a difference in the cabin. The noise was clearly reduced, but it was difficult to remember exactly what it sounded like before. So, in the name of pseudoscience, I took the same sound measurements after the install to see what the quantitative before-and-after difference actually was.
I'm calling it pseudoscience because I took a somewhat scientific approach using some relatively crude tools.
Sound Pressure Level (dB):
The following table shows the overall sound pressure level (SPL) in decibels (dB) for three different driving conditions:
At idle, there was no discernible change, which isn't too surprising. My car has only a TurboXS axleback exhaust, so it is relatively quiet at idle. During in-town driving conditions, simulated by going 35 MPH in 4th gear, there was a 2dB reduction, and while driving at highway speeds in 5th gear, I recorded a 4dB reduction. These numbers are the result of averaging several measurements together, in an attempt to eliminate erroneous data. Still, with a small data set to average and a relatively crude measurement (iPhone), the trend is present.
Since the decibel is expressed on a logarithmic scale, a 4dB reduction is actually quite large. To put it into perspective, a 3dB reduction corresponds to reducing the sound pressure level by half.
Spectrum Analysis (Frequency Response):
The overall SPL reduction in dB is only a part of the full story, however. Recall that sound can be thought of as many many different sinusoidal components summed together. We can look at the power spectrum, or power spectral density (PSD) to visualize where in the frequency spectrum a particular signal resides. The x-axis, or independent variable on a PSD, is the frequency. Thus, a perfect sine wave would show up on a PSD as single spike at that particular frequency.
I used the Spectrum Analyzer app available for free in the Apple App store to measure the spectrum, again under three different driving conditions:
At idle, this is essentially the noise floor. There is almost no discernible noise to reduce, so the before and after spectrum looks nearly identical.
At 35 MPH in 4th gear, there is some noticeable reduction. The plot below shows the spectrum before (green plot) and after (blue plot) the sound deadening. The most reduction occurred between approximately 210-420Hz, while there was actually some small amplification in the 420-600Hz range. Overall, this corresponds to a 2dB reduction, as indicated in the table above.
At 70 MPH in 5th gear, there is pure reduction in the spectrum (no amplification). There is a discernible reduction in nearly all frequencies from 200-900Hz, likely as a result of using BOTH the RAAMmat BXT II and the PS Ensolite. Presumably, the higher frequency elements (500-1,000Hz) are due to wind noise at high speeds, since these frequencies were not as prevalent in the 35 MPH case. This is only speculation (again, this is pseudoscience). In any case, this measurement seems to agree with the 4dB reduction measured by the dB app.
Despite what the numbers say, this was truly one of the most satisfying installs I have done on any one of the 5 cars (3 of which were Subarus) I've owned. I read somewhere that one of the key differences between economy cars and luxury cars is the amount of sound deadening material used. Don't get me wrong, I'm not comparing my Subaru to a high-end Mercedes Benz, but I do enjoy driving my car even more than before.
I must also say that using the RAAMmat products were extremely easy to use and the cost was very reasonable for what you get. It's noticeably easier to hold a conversation in the car while driving at speed, but that fact is often diminished since this was an entire stereo installation, which means the music is turned up louder now. But that's a topic for another post.