Hi, I do not know what you really mean, but try to do the FT like this:

fftshift(fft(fftshift(cos(x))))

And check the phase.

Hello everybody,
>
>I have a problem in phase shift with Fourier transform: what I did is to create a cosine function with known period, in example T@ and perform an FT on that signal.
>
>FT give me correct period but phase is correct only if I perform FT on an integer multiple data points, in example T@ and num.points 400 or any x*40.
>
>I get max phase shift (180�) with T/2, so in previous example performing FT on 400 +(40/2) points.
>
>Is there anyone do know how can I get every time correct in phase no matter how many points I do perform FT?
>
>Or does anybody knows if there are links, algorithms, codes that fix this problem?
>
>Alberto
>
Hello everybody,

I have a problem in phase shift with Fourier transform: what I did is to create a cosine function with known period, in example T@ and perform an FT on that signal.

FT give me correct period but phase is correct only if I perform FT on an integer multiple data points, in example T@ and num.points 400 or any x*40.

I get max phase shift (180) with T/2, so in previous example performing FT on 400 +(40/2) points.

Is there anyone do know how can I get every time correct in phase no matter how many points I do perform FT?

Or does anybody knows if there are links, algorithms, codes that fix this problem?