Low frequency noise cancellation mp3

[ndolf]

Anyone knows where I can find a noise cancellation mp3 that made up of noises that is out of human's hearing range but would suppress the surrounding noise? Or is this even possible with regular earphone?

 

[PottedMeat]

I don't understand what you're asking. Noise cancellation works with a speaker + microphone + processing to create destructive interference. Using a speaker alone may work if you have a recording of the noise and the noise doesn't change. Suppressing noise out of hearing range isn't going to affect noise in your hearing range.

 

[Lord Banshee]

Use High Pass-filter to remove "low frequency" noise, but it will also remove low-frequency everything.

Also everything PottedMeat said is correct.

 

[Peter]

Sub-hearing frequencies on earphones? Dream on. Earphones can't even reproduce low frequencies /within/ our hearing range - you need WAY bigger membranes for that.

 

[Rubycon]

Originally posted by: Peter
Sub-hearing frequencies on earphones? Dream on. Earphones can't even reproduce low frequencies /within/ our hearing range - you need WAY bigger membranes for that.

This is incorrect. Large diaphragms are necessary to reproduce low frequencies in free air to be heard. When covering the pinna or in the case with IEM; a tiny diaphragm is very capable of reproducing movement well below 20Hz. Most loudspeakers low frequency section (reflex design) are tuned at a much higher frequency (45Hz typ) and when driven with material this low produce very audible first and second harmonics. (not to mention their drivers can be physically damaged from over excursion at excessive inputs at these frequencies)

My IEM's can produce infrasonic material down to single digits cleanly and the effect is scary. Using a sine tone generator it produces a sensation like a grub is past the eardrum trying to chew into my brain! :laugh: They can do this with inaudible distortion as they have separate drivers for high and low frequencies with independent ports. These shoot right into the ear canal only a few mm from the ear drum. The biggest benefit is the isolation - the reduction of background noise. Much lower (and safer) listening levels can be achieved in this manner which is important for stage use. (I'm a musician)

As far as cancellation goes it's possible to record a monotonous noise and set up destructive cancellation however with random background noise this will not work at all.

 

[BrownTown]

One might also point out that the mp3 codec uses perceptual coding to achieve data compression, this means that it throws away any signals that are not able to be heard.

 

[Rubycon]

Originally posted by: BrownTown
One might also point out that the mp3 codec uses perceptual coding to achieve data compression, this means that it throws away any signals that are not able to be heard.

High frequencies are affected mostly since they take up much more space. The lowest bitrate can still accurately reveal low frequencies. In some products (namely video recording gear) that have a severe roll off of low frequency the first thing that gets blamed is the compression methodology. This is simply incorrect for the reason just mentioned.

If one were to encode full spectrum pink noise (20-20k) they would see the output affected especially past 15kHz. How this looks depends on the codec and the bitrate chosen.

 

[BrownTown]

Its not even a matter of bitrate, its simply that the perception threshold at those frequencies is so high it makes no sense to encode them at all. Also like was said earlier the equipment might not even be able to produce sounds less than 20Hz. However all of this is moot since the OP asked about something which isn't even possible regardless.

 

[Special K]

Originally posted by: Rubycon

Originally posted by: BrownTown
One might also point out that the mp3 codec uses perceptual coding to achieve data compression, this means that it throws away any signals that are not able to be heard.

High frequencies are affected mostly since they take up much more space. The lowest bitrate can still accurately reveal low frequencies. In some products (namely video recording gear) that have a severe roll off of low frequency the first thing that gets blamed is the compression methodology. This is simply incorrect for the reason just mentioned.

If one were to encode full spectrum pink noise (20-20k) they would see the output affected especially past 15kHz. How this looks depends on the codec and the bitrate chosen.

Why do high frequencies use much more space? I guess a better question would be to ask would be - what is the general structure of an audio file?

 

[BrownTown]

The reason that the higher frequencies use more space can be seen for to reasons. First off if you are doing just a PCM representation of the wave like a ".wav" file then the bit rate is equal to twice the highest frequency you wish to represent. In this case the highest frequency is ~20kHz and the sampling frequency used is 44.1kHz, however in most sorts of sounds you might be recording the amount of information in the higher frequencies is very low, not to mention your ear can't hear that high of frequencies very well. So you could cut the sampling frequency in half and still get most of the information (for example all the voice information and most instruments although the sounds won't be as "rich" for loss of the high frequency harmonics). With the perceptual coding files like ".mp3" they compare the information from an input signal to a graph of humans ability to perceive sounds, they essential take the flourier transform of the signal for small windows and then create a "masking curve" based on human perception and the way that sounds of similar frequencies can mask each other. The energy at the different frequencies is then compared to the masking curve and anything below the curve is thrown out completely, anything above the curve is coded at a bitrate proportional to its intensity over the masking curve. What Rubycon is pointing out is because there is usually very little information at the high frequencies and because they are harder to hear to begin with you usually end up throwing them out all together. And like discussed before the bandwidth of this "high frequency" rage is very large, so you can GREATLY reduce the size of an audio file when they are removed (and lose no quality). Something like the mp3 can produce a sound file 1/10the the size of a 44100Hz 16bit PCM file that sounds just as good to most people.

 

[Born2bwire]

Originally posted by: BrownTown
The reason that the higher frequencies use more space can be seen for to reasons. First off if you are doing just a PCM representation of the wave like a ".wav" file then the bit rate is equal to twice the highest frequency you wish to represent. In this case the highest frequency is ~20kHz and the sampling frequency used is 44.1kHz, however in most sorts of sounds you might be recording the amount of information in the higher frequencies is very low, not to mention your ear can't hear that high of frequencies very well. So you could cut the sampling frequency in half and still get most of the information (for example all the voice information and most instruments although the sounds won't be as "rich" for loss of the high frequency harmonics). With the perceptual coding files like ".mp3" they compare the information from an input signal to a graph of humans ability to perceive sounds, they essential take the flourier transform of the signal for small windows and then create a "masking curve" based on human perception and the way that sounds of similar frequencies can mask each other. The energy at the different frequencies is then compared to the masking curve and anything below the curve is thrown out completely, anything above the curve is coded at a bitrate proportional to its intensity over the masking curve. What Rubycon is pointing out is because there is usually very little information at the high frequencies and because they are harder to hear to begin with you usually end up throwing them out all together. And like discussed before the bandwidth of this "high frequency" rage is very large, so you can GREATLY reduce the size of an audio file when they are removed (and lose no quality). Something like the mp3 can produce a sound file 1/10the the size of a 44100Hz 16bit PCM file that sounds just as good to most people.

Mmmmm... flourier transform. That's the magical transform that engineers use to convert your basic grains into alcoholic beverages. Let's hear it for liquid bread!

A little more clarification on mp3, they store the frequency domain information as opposed to time domain like PCM. So with mp3's, they have only so much data that can be used to describe the frequency response within a given time window. So that is why they care about dropping or decreasing the bitrate of frequency information that cannot be intelligibly perceived. Instead, they devote more data to the portions that more greatly affect one's perception. This also means it's rather difficult to make a judgement on the quality of a mp3 given it's bitrate. The frequency and word length of PCM directly controls the maximum frequency that can be reproduced and the dynamic range possible (the difference between the loudest and softest sounds and the difference between the two closest loudnesses). For mp3's, the quality depends greatly upon the encoder and decoder but given the maturity of them these days there is less variance in this then there once was.

 

[bobsmith1492]

So, interesting discussion but what the heck was the OP talking about?

 

[silverpig]

Originally posted by: bobsmith1492
So, interesting discussion but what the heck was the OP talking about?

It sounds like he wants an mp3 that plays sounds that you can't hear that will cancel out noise that you can hear. That doesn't exist though.

 

[1prophet]

Darpa's Ultrasonic Cloaking Device

This effort will use directed ultrasound technology to enable the capability to significantly reduce sound emissions from large scale tactical military hardware. Theory predicts that nonlinear effects of high-power acoustic radiation on the atmosphere can cause acoustic energy to dissipate rather than radiate. This theory has been confirmed in some limited experiments; this program will apply it to reducing acoustic emissions of U.S. equipment. Reduction in noise levels by at least 30dB would enable U.S. forces to effectively operate considerably closer to enemy forces without being detected aurally

 

[0roo0roo]

Originally posted by: silverpig

Originally posted by: bobsmith1492
So, interesting discussion but what the heck was the OP talking about?

It sounds like he wants an mp3 that plays sounds that you can't hear that will cancel out noise that you can hear. That doesn't exist though.

i think he thinks that noise canceling headphones just play a generic noise suppressing noise when its really made in real time and only works in real time.

 

[bendixG15]

Originally posted by: Peter
Sub-hearing frequencies on earphones? Dream on. Earphones can't even reproduce low frequencies /within/ our hearing range - you need WAY bigger membranes for that.

I agree with you, Peter, because you are right.

 

[soccerballtux]

OP, your best bet is to get some pink noise and turn that on. Helps your brain tune out any specific noise you're hearing. There are pink noise generators for this.

 

[SonicIce]

Ok what if you had a nice pair of open headphones, a good microphone, and a processor. The mic would pick up sounds coming to your ears from your environment, and be processed to be equal and opposite to the environment sound so you would effectively hear nothing? If the ear peices were pointing the opposite direction maybe?

 

[bobsmith1492]

Originally posted by: SonicIce
Ok what if you had a nice pair of open headphones, a good microphone, and a processor. The mic would pick up sounds coming to your ears from your environment, and be processed to be equal and opposite to the environment sound so you would effectively hear nothing? If the ear peices were pointing the opposite direction maybe?

That's the concept behind noise-canceling headphones except the idea is to put the microphone exactly where your ear is and then generate an output signal using negative feedback to so that the amplitude of signals picked up by the microphone at your ear is zero.

 

[SonicIce]

Originally posted by: bobsmith1492

Originally posted by: SonicIce
Ok what if you had a nice pair of open headphones, a good microphone, and a processor. The mic would pick up sounds coming to your ears from your environment, and be processed to be equal and opposite to the environment sound so you would effectively hear nothing? If the ear peices were pointing the opposite direction maybe?

That's the concept behind noise-canceling headphones except the idea is to put the microphone exactly where your ear is and then generate an output signal using negative feedback to so that the amplitude of signals picked up by the microphone at your ear is zero.

But would it be possible to write software for windows to do that? I wonder if it would actually work. All you would need to do is walk around with a laptop wearing headphones and mic's. Except most soundcards only accept a mono mic.

 

[bobsmith1492]

You'd have to put the mic in your ear for it to work properly. I don't know if you could process quickly enough, either, to cancel high-frequency components. Maybe, who knows?