Recursive Moving Average Filter

Python. I implemented the recursive moving point average filter to clean background noise of test some audio files. It basically works in time domain. That’s way, it is more faster than FFT algorithms when we keep point number is small enough. It is very basic and works very fast. It is very optimal filter for some audio types such as telephone or normal speech to clean background noise. However, it is not good for musical instruments with many harmonics such as strings. When I tested this implementation with a string sound, it lost its some of properties. I will do more some experimental implementations based on this technique later on.

[1] Advanced Digital Signal Processing and Noise Reduction, Saeed V. Vaseghi

[2] The Scientist and Engineer’s Guide to Digital Signal Processing, Steven W. Smith

[3] Signal Smoothing Algortihms