FAQ: "How loud is my fan?"

FAQ: "How loud is my fan?" and other noise questions

What I'm doing here is basically trying to put together a FAQ on noise levels and decibels.

So, “How loud is my fan?”

Well, that all depends on the decibel level.
The decibel scale is based on powers of 10, or in other words, it’s a logarithmic scale. For the few of you who understand how the pH scale works, it’s similar.
The lowest level of sound audible to a human being is defined as 0 decibels. Every time the sound wave gets 10 times more intense, you add 10 decibels. Here is a good link explaining it.

“How does this compare to my stock heatsink?”

Here is a list of heatsinks made by coolermaster. These are generally what AMD and system builders use. Note that they range between 33 and 42 decibels.
Intel does not note any specifications, but note that here, a sound meter with a minimum of 50db does not detect the stock heatsink at 8”. This means that it is at MOST 36db.

“How is fan noise measured?”

Most companies state that they measure the noise level at 1 meter, or about 39 inches, from the fan.

“What if I use two quiet fans?”

Note that if you double the intensity of the sound wave, the decibels increase by about 3. This is NOT a good indicator of how loud it seems. To the ear, an increase of 10 decibels is twice as loud.

“Will I be able to sleep in the same room as my computer?”

Maybe. At about 45 decibels, a person cannot sleep. Fortunately, sound intensity decreases with distance. The decrease does not depend on the initial noise level, and it only depends on the relative distance. That means that something will be X deceibels quieter if you double the distance, no matter whether you start from 1 foot away or you start from 10 meters away. Here’s a quick calculator to tell you how much quieter it will be. For those of you who use a 57 decibel tornado, that means you’ll have to be 13 feet from your computer if you even hope to sleep.

“So, wait a minute… if the decibel drop doesn’t depend on the initial noise level, then couldn’t I get negative decibels?”

Sure. And it would be a good way to measure the sound. Keep in mind, though, that you can’t hear anything under 0 decibels.

“How loud can it be before it damages my hearing?”

Huh? Did you just say something?
Well, a sustained noise of 85 decibels can damage hearing. You’d have to have 645 tornado fans running at the same time to reach that! Or you could just hold your ear 2 inches from the fan. Of course, the latter is more likely to damage your hearing by cutting off your ear.
Keep in mind, though, that if you can’t hear that delta or tornado fan over your speakers/headphones, you may be damaging your hearing. You won’t feel pain until about 130 decibels.

I always thought a decibal increase of 3 was our perception of something being twice as loud, but i guess it was confused with what you said.

Also that 45 dba and a person not being able to sleep is innacurate, it really varies, i can sleep with a 20in box fan on high 18in from my bed. it probably produces 80dba (assuming an 80mm tornado produces 70dba which it does in most cases, the box fan seems twice as loud.) And going by that chart we cant hear anything less than 3dba, example: we cant hear anything that is less then zero right? then we cant hear anything that is 1 or 2dba either because it is less than what we are able to percieve.

3 decibels is typically the smallest change is volume that a person can detect when they aren't expecting it. 1 decibel is what most people could detect if they are expecting it and 10 decibels seems to be twice as loud.

The decibel scale is really a very counterintuitive thing. I remember a kid in my physics class standing up and going "That's bulls**t!" during a lecture on the decibel scale

I know some of it is tough to believe/understand but what caffinehog said is correct.

I dont really doubt it, but it could be revised, it contradicts quite often. Ook so 1 dba is possible if your expecting it, the link he reffered to said it was not perceptible, sheesh.

All my teachings and experience with IASCA and other sanctioning bodied soundoffs has taught me that 3 Db is when sound doubles, unless you are running a different Db scale , which I highly doubt, percieved sound will double every 3 Db's

I've spent hours. There's enough contradiction out there to drive you nuts. I put together what I think is the most consistent between sites.

3db is TWICE AS INTENSE of a sound wave. The ear does NOT perceive it as twice as loud, though.

schismspeak said:

I always thought a decibal increase of 3 was our perception of something being twice as loud, but i guess it was confused with what you said.

Also that 45 dba and a person not being able to sleep is innacurate, it really varies, i can sleep with a 20in box fan on high 18in from my bed. it probably produces 80dba (assuming an 80mm tornado produces 70dba which it does in most cases, the box fan seems twice as loud.) And going by that chart we cant hear anything less than 3dba, example: we cant hear anything that is less then zero right? then we cant hear anything that is 1 or 2dba either because it is less than what we are able to percieve.

Click to expand...

And I'm innaccurate?

Zero decibels is the threshold of hearing, not three.
As for three being the least we can detect, I won't even touch that... there's going to be too much "I can hear it, even if you can't." I would think 3 decibels is pretty detectable, though.
A tornado is 55-58db, depending on the model. That's measured at 1 meter.

P.S. The first link is pretty darn accurate. Remember, the INTENSITY of the sound wave is NOT equal to the decibel level. The intensity increases by a power of ten every time the decibel scale increases linearly. (In other words, to ADD ten decibels, you have to MULTIPLY the intensity by 10.)

Caffinehog said:

3db is TWICE AS INTENSE of a sound wave. The ear does NOT perceive it as twice as loud, though.

Click to expand...

thats true, twice as intense does in no way what so ever mean it's twice as loud... maybe a certain some1 on this thread should check his/her sources before calling some1 else's hard work, and i quote " full of bull"... not to mention any names *cough* schismspeak *cough*

Also, most people cannot sleep at 45 decibels. Every website that I check says that somewhere between 40 and 50 decibels is the level at which most people cannot sleep. Anything as low as 40 decibels can disrupt sleep, meaning it doesn't do as much for your body as deep sleep. Also, some people can fall into a shallow sleep up to about 70 decibels, but it doesn't give anything like the benefit of a good night's rest.

On the other hand, familiar noises up to 40 decibels can help you get a good night's rest.

heh, i second on that, i personally think 35dB is too loud to sleep on, which is why im going w/c

schismspeak said:

That is exactly what i thought too, 1dba is perceptible, and 3more seems to be twice as loud.
So I have read.
This "FAQ" is full of bull. (No im not trying to offend, but before you put something like this together maybe you should look it up at several places and not just 1 or 2 to see if anything contradicts.) That first link of yours seems to be full of bull.

Click to expand...

You guys don't understand the concept of the decibel system. The link in question is just about 100% correct. Open up your mind and try to understand it, instead of discounting it when it doesn't make sense to you.

The decibel system is a bridge between the actual intensity of sound and the perceived hearing. It's purpose is to compact the wide range of intensities that we can hear; it goes from .00000000001 to 1 w/m squared (1 is the threshold of pain). Those aren't very user friendly numbers.

In the decibel system, a sound 10 times more intense will increase by 10 decibels. That way, you can fit that very large range in a scale of 0 to 130. Those two numbers are the threshold of hearing and pain in decibels.

Now, a 3 decibel increase WILL sound twice as loud. For simplicity's sake, I'll assume that 10 w/m squared equals 10 decibels. (It really doesn't!!) If you have a decibel level of 13, you can just divide 13 by 10 to get 1.3, then take 10 to the 1.3 power to get 19.952623. In other words, increasing by 3 decibels will almost double the intensity. (It isn't that easy to do real conversions from w/m squared. I simplified it).

Three dB is the half power point. If you don't know what that means, don't ask. If you know what it means you understand that electrical power and acoustic power are different and are measured differently using different instruments.

LOL, this reminds me of my physics class. hahaha, I'm just waiting for the swearing to begin.

A lot of you seem to have certain points right and others TOTALLY wrong.
So, I will explain citing equations from my physics text:

Due to the way that the human ear and more importantly the brain interpret variations in frequency and sound intensity, a scale was developed to allow people to more easily see how sounds relate based upon how loud they seem.
There are several different ways to "weight" the decibel scale. The most commong weighting that must of us will ever see is dBA. the dBA scale is weighted based upon the audible range of frequencies for the human ear. The faintest sound which the
typical human ear can detect has an intensity of 1 X 10-12 W/m2 This level is called the threshold of sound, which is also 0 decibels.
A sound that is rated at 10 decibels is 10 times more powerful than a sound rated at 0 decibels, however, the human ear only interprets it as 2 times as loud. This is why having 10 people talking in a room only seems around twice as loud as one person talking and not 10 times as loud.
Now, here is a summary:
0 dbA also 1x10-12 W/m2 = faintest sound that can be heard
3dbA - smallest change in sound intensity that can be detected when not expecting it, also typically represents a doubling in amplification power
10dbA - 10 times more powerful than 0 dbA, but only 2X as "loud"

ok, for the formulae:
dBA = 10 log ( I / Io ) also
dBA = 10 log ( P2/ P1 )
so, let's plug and chug for a moment, shall we?
I have two speakers playing the same tone at the same distance from my body. The amplifier for the first speaker is set to 1W power output. THe second is set to 10W output. so:
dBA = 10 log ( 10/1) or 10 log(10) = 10dBA so the difference between the two speakers is 10dBA or 10X the power. However, my ear/brain only interpret this 10X increase as being twice as loud.
Now, here's a real kicker (very similar to the problem that resulting in my friend getting kicked out of class, lol!)
the first speaker is pumping out 1W, the second is pumping out one million (1,000,000) Watts.
dBA = 10 log (1,000,000/1) or 10 log 1,0000,000 = 60dBA
So, the difference between 1W and 1,000,000W is 60dBA which is interpreted by my ear/brain as being 64 times as loud. (ie 2^6 power)

Good topic Caffinehog, and very understandable. In accounting, there is a principal of comprehensibility - it means that the material you present is possible to understand for those willing to take the time to understand it.

This is a common problem and this is a good summary of it's explanation, thanks for takeing the time to write it up. This thread will get it's share of linkage from me.

Lord_Mil you should come around more often BTW. You seem to have plenty of worthwhile input; I think your holding out on us.

Hmmm...maybe add this to another sticky?? A jumble of all of the stickies because this seems usefull for some people who may venture into Silent Air Cooling. I'm trying to do it also, and this sorta helped interperet a few things

Great guide.
Fold and Frag on
Brian

IMOG, thanks for the compliment. I'm not necessarily "holding out" on you, just keeping with the timeless advice of "always keep your audience wanting more" lol
In all reality, I try to give my input only when I feel very confident that it is accurate. There are lots of members of this board who have many more years of schooling/experience than I do, and I have to give them credit for the warehouse of information they have passed on to me, and I do my best to pass that on to others.