Decibels

In your search for the quietest computer parts (or even in your search for a well-designed, reasonably quiet pc), it is important to know something about bels and decibels (a decibel, or dB is a 10th of a bel).

When you double the decibels, it isn't twice as loud.

The measurement of noise (bels and decibels) is logarithmic. In other words 40 decibels is not just twice as loud as 20 decibels. And, 40 decibels is actually about nine times as loud as 20 decibels. The reason for this is that loudness increases approximately three times for every bel (or 10 decibels). So 30 decibels is three times as loud as 20 decibels. 40 decibels is three times as loud as 30 decibels. This makes 40 decibels nine times as loud as 20 decibels. Loudness itself is quite subjective. What sounds twice as loud to one person may not sound twice as loud to another. To the average human ear, however, an increase of 10 decibels means that the perceivable increase in noise is triple the noise.

How can I use decibel ratings to tell how much louder one device is over another?

In the real world, using decibels to determine how loud your computer will be is far trickier than a simple algebra equation. There are many factors that play their part in the overall noise produced by your computer which make decibel ratings hard to use when determining PC noise.

First of all, there are many factors inside your computer that will influence noise, including your case's potential either to dampen (quiet) noise or to increase the vibration of a particular device. It is wise to search for a case that hinders vibration, dampens noise, and yet still provides for ample airflow. (Our Nexus Quiet iStyle Case is one example of a well-designed case that works well for anyone in search of a quiet computer.)

The second factor that makes it difficult to accurately pinpoint how much louder one component will be in comparison to another is the fact that accurate and consistent decibel ratings are hard to come by. Not all manufacturers give scientific or accurate decibel ratings at all. Many decibel ratings are simply estimates obtained by comparing one product to another. Other decibel ratings are nothing more than over optimistic marketing slogans, which don't turn out to be accurate. And, even when manufacturers give a standard by which their measurement was obtained, those standards vary widely.

How We Note Decibels

While we have no affordable means by which to scientifically test the decibel levels of all the products that we sell, you can rest assured that we independently test each product and confidently use them in our own quiet systems. Every product we sell is the quietest available in its class. We also make an effort to make clear the advantages and disadvantages of the products we carry (price and performance versus decibel level). Additionally, we make every effort to not use misleading decibel ratings. When a manufacturer has given us a decibel rating along with a decibel rating standard (such as ISO07779), we reference that decibel rating standard. If we are at all approximating the benefit of using a particular product (using less accurate standards such as the use of a decibel rating obtained with the use of a decibel meter in an uncontrolled environment, or with the use of subjective listening tests in a quiet but uncontrolled environment), we use words like "about" or "approximate". Bottom line...we carry a wide variety of products and search widely for the quietest computer solutions. We only carry products that we feel are the very best.

The Decibels To Noise Equation

So, assuming you live in a perfect world, and that you have decibel ratings which are accurate and consistent (assuming that we are comparing apples to apples), here is the decibel to noise equation you would use to see how much louder one component is in comparison to another component.

x = bel rating of fan #1 or other component (louder component)
y = bel rating of fan #2 or other component (quieter component)
z = how much louder fan #1 is in comparison to fan #2

(3^x)/(3^y) = z

How Decibels Convert To Noise In A Real Life Example

So if fan #1 is 20 decibels (like the Zalman 80 mm case fan's decibel rating) and fan #2 is 12 decibels (like the Papst 8412NGL case fan's decibel rating), then you take (3^2.0) and divide it by (3^1.2). Remember 10 decibels equals 1 bel.

In a spreadsheet this would look like:

=SUM(3^2.0)/(3^1.2)

The answer to this comes out to be 2.41. So a fan which is rated at 20 decibels is approximately 2.41 times louder than a fan which is rated at 12 decibels. If you needed more airflow, you could get more airflow by using two 12 decibel fans and still have both fans be less loud than one 20 decibel fan (two 20 decibel fans would be approximately twice as loud as one 20 decibel fans, perhaps less is there is some noise dampening effect taking place with the fans).

In this particular case (again if this were a perfect world and there were no confusing factors), one of our Papst 12 dB fans pumps 19.4 cubic feet of air per minute, whereas one of our Zalman 20 dB fans pumps 28.0 cubic feet of air per minute.

So, two Papst fans pump a total of 38.8 cubic feet of air through per minute and create 2 times as much noise as one Papst 12 dB fan. Whereas, one Zalman 20 dB fan pumps 28.0 cubic feet of air through per minute and creates 2.41 times as much noise. You can see that ideally, it would be better to go with two Papst fans as their combined power pumps more air through and at the same time creates less noise than one Zalman. The down side of this is that the Papst 8412NGL 12 dB fan costs quite a bit more than the Zalman 20 dB fan (especially if you are using two of them). Another confusing factor to throw in there is the potential of two fans to create a bit more vibration in your case than one fan will. However, this is probably a very insignificant factor in terms of perceivable noise.