# Fixing for windowing function in FFT

I've been learning about the FFT and about the usefulness of windowing functions. A problem I'm running into is the affect of windowing functions on the amplitude of signals in the FFT. For example, lets say I have a signal that is equal to:

y = 5.0 * sin(50.0 * 2.0 * pi * x) + 1.0 * sin(80.0 * 2.0 * pi * x)


And then I want to perform an FFT and receive the following plot:

Then if I perform a kaiser windowing function of the 14th degree, I receive the following FFT:

What I'm wanting to do is the following: Given a signal, I want to extract a wave out of it of a frequency I approximately know. To do this, I'd like to apply a windowing function to make it clearer what the frequency is I'm trying to extract but not lose information about the amplitude of wave I'm trying to extract. Is this possible? Like a way to reverse the amplitude decrease from the windowing function.

• Some useful search phrases include "scalloping loss compensation" and "flat top window". – hotpaw2 Dec 26 '15 at 17:59