Anti-noise sub/bass/mid speaker

[paulspencer]

Sounds a little like Star Trek doesn't it?!

Here is the idea. It has come from another thread in which wintermute suggested a noise cancelling headphone circuit to quiet the fan noise in my amp. It got me thinking. Could this work to keep things quiet for the rest of the house?

Last night I went into the space between my sound system and where others sleep. I could hear bass (of course) and some muffled vocals, most of the energy below 500 Hz.

I know active noise cancellation works. But does it work in 3D space with speakers, rather than just headphones. With headphones, the mic is right at your ears. Also it is easy to get the cancelling sound much louder than background noise. In my case the speaker is at least 20 db ahead at 20 Hz.

Thoughts on the concept anyone?

If it works, it's a great way to crank the music at night and not disturb anyone. It has potential for people in the house, as well as neighbours getting disturbed by subwoofers! Much cheaper than turning your room into an acoustic treated bomb shelter!

For those interested, here's the link to the schematic of the device:
http://www.headwize.com/projects/noise_prj.htm

[thoriated]

Problem is, the area where noise cancelling is effective is smaller than the wavelengths to be cancelled, at least for a single point system.

[paulspencer]

Care to elaborate a little?

I'm thinking so long as it can be done so the phase is fairly close, it should work ... not totally sure how it will work in a 3D space (headphones are much simpler)

[thoriated]

Well, as a thought experiment, let's suppose you have a system that is successfully canceling an ambient sound field at some point in space by producing an antiphase sound. Let's further suppose that some particular noise source that is being canceled is from a specific direction and has energy concentrated around 1Khz. If you move only three inches from the null point toward that sound source, you'll get an addition of the two sources, not a cancellation, at 1Khz.

There will be other, partial nulls at that frequency as well as areas where the two sounds reinforce each other as you continue to move in that direction, with diminishing amplitude from SPL attenuation due to increasing distance from the ambient cancellation sound source.

[paulspencer]

The bass is the main thing however. At 100 Hz the wavelength is 3.4m and 17m at 20 Hz. Perhaps it would be effective for subwoofer cancellation only. This is the real problem area.

[thoriated]

You might get some useful results at lower frequencies. Maybe more than one cancellation point would help here also.

I once built some resonant bass traps (adjustable helmholtz resonators) for a listening room (not quite the same thing, I know) They worked, but I was a little surprised how restricted the area of effectiveness was if they weren't right near the source of the sound they were modifying.

[paulspencer]

Well, I have a few applications for this in mind. One is for a noisy fan on an amp - in this case the cancelling speaker will be right near the source of noise, so I think it should be quite effective.

So I can build this thing and try it for different uses.

I'm wondering what sort of output I'll need to make it effective. My subs have very high output down to 20 Hz. As I understand, the sound used to cancel out sound needs to be a lot louder to really work, since if you only match the level, you will only get 6db attenuation.

[wintermute]

Hi Paul,

I've read about the use of active noise cancelling in industrial environments to cancel out the noise from machinery. I also have a friend who worked for Hazletons (regional airline in Australia), and he told me that the were using noise cancelation technology in the turbo prop planes to make the interior quieter for the passengers. So I think that it's possible to do on a larger scale.

I think for the fan it should be "relatively" straight forward (as yu said it's close to the source and you want to cancel all of the noise), but for the other applications I'm thinking you have a whole lot of other problems to deal with. biggest one I see is that you only want to cancel the noise in particular directions, but not where you want to listen. with the sub I'd say it is your entire wall(s) that is radiating the noise, could be a bit tricky.

I think the technology works best if you are canceling the noise as close to the source as possible, stopping it from getting a chance to radiate in the first place (or in the case of the headphone as close to the listening position as possible). It's a long time since I read about it, and they weren't giving away all of their secrets, but I suspect that the were multiple speakers involved surrounding the noise source.

But go for it with the fan, and if you succeed there's no harm in experimenting further

Tony.

[RHosch]

That thought crossed my mind once as well, and I came to the conclusion that the most beneficial application would be some sort of "bass shakers" on the walls that in part cancelled the bass transmitted through them from inside the room. This would only work for closed rooms... wouldn't be much help in keeping the livingroom sound from spreading loudly into the kitchen, for example.

[AndrewJ]

I worked for many years in the field of active noise control, and my brother is still in this field, primarilly designing noise cancelling headsets.
Unless you can get close to the source of the noise, then noise cancelling is effective only over a very small area.
The amplitude of the two noises must match within better than 1dB, and the phase better than a few degrees to get decent cancellation. Clearly in a room where the source of the sound is unknown, then the best you will get is maybe 10dB of cancellation in a region of only a foot or two around the cancelling microphone, and only up to a hundred or so Hz.
If you can get to the source, then you can achieve much better results. I have fitted speakers to a ships funnel, which was radiating over 130dB at 35Hz, and achieved better than 20dB of cancellation.
Of couse with headphones the situation is easier. The enclosed volume of the region where cancellation is required is much much smaller, so cancellation is feasable up to around 1000Hz.
As for the fans, the noise spectrum from these is likely to be in the few hundred Hz range, so not a good candidate for free space noise cancelling.


Andrew