what does 10 dBA sound like?

[twilyth]

I'm getting a Sigmatek 1283 for a quad (either q6600 or 9450) and I want to use Scythe fans in place of the the one that comes with it. I might go push-pull (or blow-suck you could say) but from what little I've read, it doesn't seem to make huge difference over having just one high cfm fan. I have 2 Scythe fans in a case and they are amazing - SFF21F S-flex (http://www.newegg.com/Product/Product.aspx?Item=N82E16835185006). They are rated at 63.7 cfm and noise at 28 dBA. But they also have the DFS123812-2000 (http://www.newegg.com/Product/Product.aspx?Item=N82E16835185053) at 87.63 cfm and 32.91 dBA and the DFS123812-3000 "ULTRA KAZE" 120 x 38 mm (http://www.newegg.com/Product/Product.aspx?Item=N82E16835185054) at 133.60 cfm and 45.90 dBA.

The s-flex is $20 but has a fluid bearing so it should last forever right? And 63 cfm is pretty damn good, especially with that noise level. The other 2 - the 2000 and 3000 - are both sleeve bearings and both are $10 but have higher cfm numbers - especially the 3000 - I don't think my 20" box fan moves that much air.

But how much of a SUBJECTIVE difference is there between 28 dba and 33 or 46 dba? Is most of the noise going to be in the upper range where I'm not likely to hear it anyway or is it very noticeable?

I'm inclined to go with the s-flex just because I know how amazing they are and use them both for the case and the heatsink.

Any insight on this?

[RyderOCZ]

http://www.gcaudio.com/resources/howtos/loudness.html

10dB increase is twice as loud.

[Hyperhorn]

You have to understand, that dB works not linear, but logarithmic.
The link posted shows that. 40dB sounds twice as loud as 30dB, not 1/3.

(But it depends from the characteristic of a sound, too.)

[twilyth]

http://www.gcaudio.com/resources/howtos/loudness.html

10dB increase is twice as loud.
Thanks, that was interesting - even a library registers 30dba.

The rating on the Xigmatek fan is only a little higher than the s-flex and moves nearly 50% more air. But I guess the Kaze is the way to go. One review on Newegg noted using it with the 1283. It's a lot louder and probably won't last as long, but it will be in a sealed case where ambients my get up to 100F or more, so I want to make sure that ventilation is nowhere near being an issue.

I'll add in 3 or 4 s-flex's for the case and that should do the job.

[64dragon]

you wont notice much going from the 28dba s-flex to the 33dba but you will notice a huge increase in noise going with the 46dba

[spookmineer]

Sound and noise is always hard to get to grips with.

First off, a 6 dB(A) increase is twice as loud. Not 10 dB(A). Logarithms...
(Source Decibel (http://www.prorec.com/Articles/tabid/109/EntryID/51/Default.aspx) )

With a quiet PC in a quiet room, a 25 dB(A) fan will be heard at 1 meter, while a 20 dB(A) fan at 1 meter will be hard to measure at all.

Also, almost all manufacturers lie about their noise specs. I've got a couple of S-flexE fans, undervolted to 800 RPM, which can still be heard (supposedly 20.1 dB(A) @ 1200 RPM).

[edit: sorry, I got it wrong first time - it's actually 20log(p1/p2) instead of 10log(p1/p2)]

[twilyth]

you wont notice much going from the 28dba s-flex to the 33dba but you will notice a huge increase in noise going with the 46dba

Sound and noise is always hard to get to grips with.

First off, a 6 dB(A) increase is twice as loud. Not 10 dB(A). Logarithms...
(Source Decibel (http://www.prorec.com/Articles/tabid/109/EntryID/51/Default.aspx) )

With a quiet PC in a quiet room, a 25 dB(A) fan will be heard at 1 meter, while a 20 dB(A) fan at 1 meter will be hard to measure at all.

Also, almost all manufacturers lie about their noise specs. I've got a couple of S-flexE fans, undervolted to 800 RPM, which can still be heard (supposedly 20.1 dB(A) @ 1200 RPM).

[edit: sorry, I got it wrong first time - it's actually 20log(p1/p2) instead of 10log(p1/p2)]

Interesting. So going from 28 to 46 will be a difference of 18 which 2 x 2 x 2 = 8 times as loud? Damn. So Xigmatek is probably full of **** with their noise measurements since the cfm isn't a lot less than the Kaze.

I guess I'll try the Xigmatek fan and keep the Kaze as a back up. It will be inside an Antec P180 which has the laminate sides so that might help to attenuate the noise if I have to use the Kaze. I might also try a push-pull arrangement.

Thanks a lot guys.

[64dragon]

Also, almost all manufacturers lie about their noise specs. I've got a couple of S-flexE fans, undervolted to 800 RPM, which can still be heard (supposedly 20.1 dB(A) @ 1200 RPM).

this is why Vapor does the fan roundups. the 2nd one is stickied here in the air cooling section and he tested some s-flexs' and they scythe was pretty good about their measurements. i believe he will be testing the kaze series in roundup 3 but not sure when that will be out but he has made a few statements at the end of the 2nd roundup about what to expect.

i'm running 2 s-flex E's and 2 F's in my tj07 and yes they can be heard but they are still very quiet, my ceiling fan is louder than my whole case. in my previouse case i had a generic red led 120mm fan and it was extremely loud, compared to these s-flexs' i'd have to guess in the 43dba range and was deffinatly not worth it, the s-flex F's push as much air as it did and create much less noise

[spookmineer]

i'm running 2 s-flex E's and 2 F's in my tj07 and yes they can be heard but they are still very quiet, my ceiling fan is louder than my whole case.

My wall clock is louder than my case...

[weston]

um do your math, it only takes 3dB difference to double the sound intensity. B1-B2 = 10log(I1/I2) I think you would be able to tell the difference of as small a difference as 1 dB if it was the only sound present. With the background noise of the computer though, they are both pretty low noise and would likely be drowned out by the background noise.

[spookmineer]

um do your math, it only takes 3dB difference to double the sound intensity. B1-B2 = 10log(I1/I2) I think you would be able to tell the difference of as small a difference as 1 dB if it was the only sound present. With the background noise of the computer though, they are both pretty low noise and would likely be drowned out by the background noise.

You are referring to power (which is twice the voltage) instead of sound pressure.
Did you read the link?

20log (http://www.cds.caltech.edu/~murray/wiki/Why_is_dB_defined_as_20log_10%3F)

sound pressure (http://www.phys.unsw.edu.au/jw/dB.html)

Lojzek and Nick89's posts (http://www.physicsforums.com/showthread.php?p=1743656)

[s1nykuL]

Another link which may clarify

http://www.brcacoustics.com/noisedescriptors.html

[weston]

Oh my mistake! I thought that the sound intensity was the same as loudness I guess. So 10dB is a double in loudness but 3dB is a double in power. Thanks for the link. I have had several Physics classes and I had always thought otherwise!

[spookmineer]

First off, a 6 dB(A) increase is twice as loud. Not 10 dB(A). Logarithms...

10log for power, 20log for volume.