Ok. Dumb idea of the day.
As I understand it noise cancelation is achieved by sending the opposite sound wave against the sound wave of interest.
My computer is loud, and to my ear the fans make a fairly repetative noise. Could I record it with a mic then use a sound editing program to create an inverse wave?
Er, Then play it through my speakers in a loop.
I think with a recording, you’d never get the sounds to synchronize correctly. The timing must be perfect or you end up being lounder half of the time. There’s probably some fancy feedback or digital circuitry in manufactured sound cancellation equipment.
Do you lead a sheltered/isolated existence?
Have you done a web search for “Noise Cancellation”?
Do you know how many people make a living making and selling Noise Cancellation devices?
Any More Stupid Questions?
The OP asked thether the same effect can be achieved by a standard PC sound card. It’s a legitimate question.
But the answer is no - you need to detect the phase of the sound source and emit an equal sound with opposite phase (peaks of the emitted sound line up with the troughs of the original sound source, so they cancel each other). You’ve got to do the processing in real time.
If you wrote a program to sample the microphone input in real time and emit an opposite sound… Um, I’m not sure. I think the delays involved in sampling, processing and generating sound will make it impractical.
If the problem is fan noise, that’s at a very constant pitch and timbre, and the scheme could be made to work. The only problem would be that the microphone would have to be placed exactly where your head would be, and your head could never move while at your PC. If it did, the phase shift would cause the fan noise and the cancelling noise to double at times, with a beat frequency that would be even more annoying than the original noise. That could be overcome with noise-cancelling headsets that would hold a constant position relative to your head, and could pick up constant-pitch sounds with its own mike. Bose already makes those, and if you have a couple of hundred spare bucks you might look into it.
It doesn’t really bug me all that much. I don’t really want noise cancelation headphones either.
It was just a random dumb idea that I wanted to know if it’d work.
Just to further clarify. Noise cancelation doesn’t actually cancel out the sound wave right? It merely combines to create no sound in the ear right?
Not necessarily. If the sound has a single source, one can place a microphone at the source and then play the inverted phase signal at the same point. Then your head just has to stay inside the speaker pattern.
There are also various other tricks. If the sound you’re trying to cancel is a single frequency, you can do stuff with multiple sources to create phase interference. My boss and I played for a while with some subs and a digital signal processor in the park back in summer working on a way to solve a problem where an outdoor sound stage at a street festival was close to a residential area, and a particular homeowner was complaining about bass thumping. We tried various things - second sound source on a plane perpendicular to the direction of the house, seperated by 30’ and phase inverted, leads to a very deep null running at right angles from the midpoint of a line between the sources. Unfortunately, the null is pretty narrow, so it’s quiet at one point, and not quiet at all 6’ over. We were going to try placing a second source signficantly close to the house, then running the signal through a delay and then inverting phase, which theoretically should have made for a broader null in line with the second source, but before we scavenged up equipment with a delay, the organizers decided to just leave things as they were.
However, when it comes to something like computer noise, it’s almost certainly going to be cheaper and easier to make the computer quieter rather than try these active techniques. A good place for tips on this is www.silentpcreview.com. The absolute simplest thing, if it’s feasible, is to stick the computer into a closet and run the cables out to your desk. Otherwise, you’re looking at replacing, potentially, the cpu heatsink fan, power supply, and possibly altering the video card cooling.
Spingears.
You’re in GQ, not the Pit. Shape up, quit insulting posters.
The OP’s question is very imaginative and the idea has been an area of research. I used to work for Bose so I know a little bit about this. To the OP: First of all, it won’t work because timing is absolutely critical for this type of application. Noise canceling headphones will work for this type of sound but the circuitry is specially built for it and works in the sub-millisecond range. You don’t stand a chance of synching up the sound manually that well. Even if you initially got it right, any little slowdown in your computer would put it out of sync. Secondly, no one has really gotten the technology to work outside of headphones because of this synchronization problem. The cancelling wave must hit the ear at exactly the same time as the original sound. The fact that your head would be moving around would put the two signals out of sync. There has been research in this area however (with little result).
The military is very interested in this technology for tank and helicopter crews. There have been experiments with car stereos to cancle out the road noise but with very limited results. Seems the area of cancleation is small compared to the head and that any movement outside of this small zone is quite distrubing.
Your idea is not that stupid.
Or you could utilise the method of noise cancellation implemented by one of my colleagues; to wit: “The fan at the back of the computer was really noisy, so I stopped it by jamming a nylon cable tie into the blades”, although of course I can’t really recommend this approach.
OK, that’s your problem. Deal with it, rather than trying to generate more noise to cover it up.
Fred Langa’s computer column (www.langa.com/) has had several items on making your computer quieter. Most are quite easy, and most cost just a few dollars.
To quote part of it:
“In “Cool and Quiet— Part One” ( Langa Letter: Cool And Quiet, Part 1 we started with the basics of quieting a PC’s noisy fans, and saw how PC noise reduction can actually be rather easy and cost just a few dollars. A high-quality, nearly silent PC fan can cost under $10, for example.”
I’d try something like that first.
Noise cancellation doesn’t “generate more noise to cover it up”, it actually eliminates the noise altogether. At least in theory. It’s difficult to execute in practice, as the prior responses have made clear.
as I stated before. My computer doesn’t actually bother me. I just had an idea and wanted to see if it was at least feasible.
The idea as stated in the OP wouldn’t work, for reasons related to synchronizing in time but actually much more subtle. In this case it has to do with information theory, of all things.
If you were trying to cancel out a single frequency sine wave, which is a very pure and dull tone to the ear, you could do it with an excellent recording and careful synchronization. It would be difficult, but they do much harder things when they synchronize cell phone equipment and other such stuff.
But it’s pretty much white noise coming from the computer fan. There’s only a little bit of a pitch or any other specific pattern to the sound.
To cancel it you’d have to put a microphone inside the headphones and give the right feedback so that the microphone got zero total sound, I think. In answer to another question, typically the noise cancellation doesn’t cancel the entire source, it just cancels it in a tiny space next to your ear. That’s where this microphone would be.
White noise doesn’t have any special pattern to it. In this sense it is like what your old fashioned modem would send down the phone line. If there were a constant pattern to the sound, that component of the information could be separated out and sent as a message like “add the following pattern back into the sound at the other end”. Just like if you wrote a letter and inserted the Gettysburg Address at every 1000th character space, obviously a file compression mechanism could figure out how to replace it with a marker and make the file so much smaller. So, when a modem is operating in a highly efficient mode where compression has been used to eliminate superfluous patterns, and therefore use all of the signal to represent the nonpattern part of the information, it makes white noise. Your computer fan makes white noise because of the tiny eddies of air swirling around into ever smaller and smaller vortices, but the result is the same. There’s no organization to take advantage of to compress the information content it any further.
Eventually white noise and fine air turbulence turn into heat, which is a special statistical distribution of molecular motion that is completely disorganized with respect to patterns. The idea of entropy comes up in two places - information theory and thermodynamics - and this area is where the two are actually common with one another.
So any recording of the white noise coming out of a fan would not match the white noise coming out at another point in time, any more than random number generators create duplicate sequences for extended times.
I think this is a very subtle point, and think that before about 1900 or so nobody in the world would have been able to explain why the OP idea wouldn’t work.
If fan noise really all white noise? My PowerMac G4 a.k.a. Quicksilver was nicknamed “the wind tunnel” by irate purchasers when it first came out. For good reason, too, because it’s louder than hell when the fans really get going. There’s definitely that “white noise” kinda “phffffft” element to the sound, but also continual tones, I think of perhaps more than one pitch, relatively high up on the register, and quite annoying. I’d be happy to get rid of noises like that and just keep the dull “whooshing” of air sound.
I think it’s a really clever idea. As many other posters have said, you have to do it in real time rather than as a recording. The trick would be to use the virtual machine capabilities of newer processors to run a little RTOS. Comapnies like AMD and Intel are always trying to think up things to do with all those extra CPU cycles they have.
WTF. All noise cancleing headphones I’ve had cancal noise is the sub-bass region (i.e. 20 - 80 Hz). It is much easier to cancel low frequency noise than mid-range noise. It has nothing to do with “white noise”. Fans are easy to cancle out with headphones, they typically are low frequency. Cancle them out without headphones, that requires high order diffetential equations to match the opposing frequencies and a controled enviroment. Not easily achieved or explained to people not immearesed in the technology.
In sound recording noise cancellation is often used when recording someone who doesn’t want to listen to foldback through headphones. You provide two speakers wired out of phase and position the mic where the maximum cancellation effect occurs. It is easy to find by ear. You can test the same effect yourself if your stereo lets you change the leads around. Simply reverse positive and negative on one and stick your head between the speakers and notice how the sound level drops particularly on the bass. If it doesn’t and sounds better (which is common) you had your stereo wired out of phase anyway and that’s why it’s so gutless.
I have always assumed that all the commercial noise cancellation devices use the out of phase effect which would be simple to replicate in some form. A few times doing live sound I used the trick with the foldback speakers to eliminate feedback squeal. It always worked and no-one on stage ever noticed.