
There are 4 messages in this thread.
You are currently looking at messages 0 to 4.
hello. i got some kind of problem with acoustic feedback. I need to create an algorithm in C language, which fights the acoustic feedback. It means, for instance, when i put the microphone device close to the speaker the feedback occurs, and it is supressed at once. Could anybody help me? thanks and greetings______________________________
Why does it have to be in C? Why could it not be in machine code or assembler? Or, if you wanted to illustrate that you were a tyro, in Forth?) I suggest that you read up on Echo Cancellation, a topic well-covered in the literature. (I suggest, rather than explain, because it's not yet an area of expertise for me, and you can do the reading just as effectively as I can) "decyL" <d...@poczta.fm> wrote in message news:9...@posting.google.com... > hello. i got some kind of problem with acoustic feedback. I need to > create an algorithm in C language, which fights the acoustic feedback. > It means, for instance, when i put the microphone device close to the > speaker the feedback occurs, and it is supressed at once. Could > anybody help me? thanks and greetings______________________________
"Airy R. Bean" <a...@lycos.co.uk> wrote in message news:c1na4q$1kttuu$1...@ID-217727.news.uni-berlin.de... > Why does it have to be in C? Why could it not > be in machine code or assembler? Or, if you > wanted to illustrate that you were a tyro, in Forth?) > > I suggest that you read up on Echo Cancellation, a topic > well-covered in the literature. (I suggest, rather than explain, because > it's > not yet an area of expertise for me, and you can do the > reading just as effectively as I can) > Why should he read up on echo cancellation? The problem and mechanism is quite different. What he needs is a automatic notch filter. This is like a project I used for one of my students last year. (Why should he not use C? With modern hardware there is rarely a need to use machince code or assembler to get the required speed. ) How are you getting on answering my questions about the unit sampling pulse? Chimera______________________________
The most effective algorithm for killing feedback fast is to reduce gain! This however has the side effect of reducing the gain. :-) You can also employ automatic notch filters that reduce the gain only at the frequencies are feeding back. But if you have too much loop gain (e.g. the microphone is too close to the speaker) this won't help because basically all frequencies are feeding back at once. You'll get rid of one and another will pop up. At that point, cutting overall gain is the only answer. Do some reading on the existing hardware devices to see how they work. Sabine is a good place to start looking. "decyL" <d...@poczta.fm> wrote in message news:9...@posting.google.com... > hello. i got some kind of problem with acoustic feedback. I need to > create an algorithm in C language, which fights the acoustic feedback. > It means, for instance, when i put the microphone device close to the > speaker the feedback occurs, and it is supressed at once. Could > anybody help me? thanks and greetings______________________________