Is there any way to synchronize phase of output of
inverse DFT in each buffer? When I send the output of
inverse DFT to the speaker it sounds nasty.
Of course I know the
windowing functions but it just gives me in/decreasing level at the beginning and the end of buffer. It sounds better but with tremolo effect. I’ve also heard about overlapping windowing, I haven’t tried that but I am pretty sure it also won’t give me clean signal. Probably it will give me some phasing/chorusing effect that will appear periodically on each beginning and end of buffer. So I wonder how to repair that?