2nd year computer engineering student here.
This explanation really isnt for 5/yos but i think its interesting nonetheless.
Basically, very smart humans have mathematically defined what a “system” is – takes an input and returns and output. Moreover, they found that some systems – most systems in nature actually – behave in a way that makes them very easy to predict – and reverse.
The useful part is that almost everything is a system, especially everything regarding sound. (for example, noise is a system: takes what you say and return what you say, but with some background noise)
Thats how noise canceling algorithms work – they find what “system” the noise is and reverses its effects.
To be exact, when i say “reverse”, i mean putting the output of the “noise system” through another system – the “inverse” system. Basically a new system that can take the output of the previous system and give the input – the noiseless sound – which is exactly what we want.
If you want to know how the math works, look up “convolution systems”, and more precisely “LTI systems”. I can explain it further but idk if anyone would want to read that tbh
2nd year computer engineering student here.
This explanation really isnt for 5/yos but i think its interesting nonetheless.
Basically, very smart humans have mathematically defined what a “system” is – takes an input and returns and output. Moreover, they found that some systems – most systems in nature actually – behave in a way that makes them very easy to predict – and reverse.
The useful part is that almost everything is a system, especially everything regarding sound. (for example, noise is a system: takes what you say and return what you say, but with some background noise)
Thats how noise canceling algorithms work – they find what “system” the noise is and reverses its effects.
To be exact, when i say “reverse”, i mean putting the output of the “noise system” through another system – the “inverse” system. Basically a new system that can take the output of the previous system and give the input – the noiseless sound – which is exactly what we want.
If you want to know how the math works, look up “convolution systems”, and more precisely “LTI systems”. I can explain it further but idk if anyone would want to read that tbh
Two reasons:
The first is that the microphone and speakers point in the same direction. This means that for the mix to pick up the sound being produced by the speaker that sounds needs to project or away from the phone, reflect off of something, then bounce back in the direction of the mic. The sound generally spreads out and reduces in volume enough here that it isn’t picked up by the microphone effectively enough (as the mic is tuned to detect the volume of a human voice pointing directly at it).
The second is that we can also use active sound cancellation. Because the phone knows what sound it is playing through the speaker, it can detect this same signal being detected again and cancel it out using some electronic trickery.
Two reasons:
The first is that the microphone and speakers point in the same direction. This means that for the mix to pick up the sound being produced by the speaker that sounds needs to project or away from the phone, reflect off of something, then bounce back in the direction of the mic. The sound generally spreads out and reduces in volume enough here that it isn’t picked up by the microphone effectively enough (as the mic is tuned to detect the volume of a human voice pointing directly at it).
The second is that we can also use active sound cancellation. Because the phone knows what sound it is playing through the speaker, it can detect this same signal being detected again and cancel it out using some electronic trickery.
Latest Answers