1
$\begingroup$

I'm going to do feature extraction using a speech signal.

The first step is the windowing of the signal using Hamming window.

I'd like to know if I should apply one single window to the whole signal or if I should apply several windows to different frames of the signal?

What are the differences between them?

Thank you.

Improve this question
$\endgroup$

2 Answers 2

Reset to default
1
$\begingroup$

Each frame to be FFT’d needs to be individually windowed to avoid rectangular window artifacts (sometimes called “leakage”) due to the finite length of the FFT.

The frame length you choose depends on the time-frequency resolution trade-off needed to extract the features in which you are interested.

Improve this answer
$\endgroup$
0
0
$\begingroup$

Windowing is used for multiplying each time frame of the signal with a window function to avoid discontinuity at the borders of the frame.

According to Haytham Fayek:

There are several reasons why we need to apply a window function to the frames, notably to counteract the assumption made by the FFT that the data is infinite and to reduce spectral leakage.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.