"One use for white noise is in the field of architectural acoustics. In order to dissemble distracting, undesirable noises in interior spaces, a low level of constant white noise is generated."

Could this be used to dissemble fan noise in any way?

No. Fan noise *is* white noise (the moving air part, actually).

This is more for "hiding" chatter, like people talikng in a reverbant space. Introducing a higher noise floor by using white noise to MASK the chatter, would give the listener the psycoacoustic impression of less distraction and an overall smoothness precieved as quiet, yet acctually louder.

But it is true that if you produce exactly the same sound and play it out of phase the two will cancell each other out. In theory you could cancell your fan noise by recording it and playing it back. Maybe you could install a small mp3 player and speaker inside your PC and just get phase cancellation by moving it away from the fan until the distance is great enough?
Potsy

consult this thread for more information:

http://www.silentpcreview.com/forums/vi ... ling#95520

briefly, you have to cancel the noise at the receiver (your ears), not at the source. you can get noise-cancelling headphones for long-haul flights/noisy train journeys etc. they work (not very well mostly!) in exactly this way.

In-ear headphones work the best to block out all external noise. Its like an ear-plug with speakers.

The nature of white noise probably makes it just about impossible to use phase cancellation -- it's supposed to be random noise. How could you create an out of phase replica of random noise as it's being generated? It would have to be not so random noise.

I work in a cube farm with white noise pumped in and if you put on a pair of those Bose noise canceling headphones, flip the power switch, the white noise disappears. So white noise is not totally random or the response time of the anti-noise generation system is sufficient to detect and emit before the sound wave hits the ear drum. I, along with many colleagues, travel internationally and we've found a broad range of quality (and cost) for those headphones. That makes me lean towards the response-time solution.

Yeah, I've used noise canceling headphones on planes before, and they do work. I agree with jaganath that it's probably only possible to achieve good noise cancellation at the receiver.

I think in principle you could cancel noise at the receiver or at the source, but only if the source is a point source and the noise-cancelling emitter is co-located with that source (or with those sources). Otherwise the very finite speed of sound is going to mean that you'll get noticeable phase shifts in the signal and the cancelling signal (and some very entertaining spatial interference patterns going on too!)

On a plane a lot of the white noise will come from the engines (pseudo-point-sources), but I'm guessing that quite a lot of it will come from the airframe too, both through vibrations transmitted from the engines and also due to aerodynamic stuff happening to the airframe itself. In such a case there really isn't a point source you could attach a cancelling emitter to, so cancelling at the receiver is the best approach.

That's my theory anyway

your theory seems pretty spot-on (although im not sure engine noise is white noise, as its mostly periodic and a lot of the energy is concentrated in a narrow frequency band):

http://en.wikipedia.org/wiki/Noise_cancelling

Noise cancelling must occur at the hardware level. Any higher and the time delay between the device's reception of sound to the rebroadcast of the inverse to the time it reaches your ear will be too great. Software solutions ie. mic to fan and speaker out would be too slow. That's also not taking echo effects into account. It's much more practical to silence your computer with traditional means. Stopping the noise at the source is much better than trying to fight it with either its inverse or with more "white noise".

That's interesting that white noise can also be phase-cancelled!

BTW do you think it's possible to cancel the whine/buzz of lamp dimming inverters in LCD monitor?
(I mean by some other way than headphones)

That noise has a pretty definite spectrum consisting mostly of frequencies 50, 100, 200, 400, 500 Hz and such.