I'm specifically trying to use an inverse Fourier Transform to find $h(t)$, but I'm finding it difficult to get $H(\omega)$ in the first place.

I'm under the impression from my textbook that $H(\omega) = |H(\omega)|e^{j\arg\{H(\omega)\}}$, and that $h(t)$ would just be the inverse Fourier Transform of that.

  • 1
    $\begingroup$ You're right about all that, so I guess there's no question, is there? $\endgroup$
    – Matt L.
    Nov 13 '18 at 21:29

Yes you are right. Let's practically implement it with Matlab / Octave.

N = 8;
h = [1,2,3,4,4,3,2,1];    % just an impulse response (FIR)

H = fft(h,N);             % DFT H[k] of h[n] as samples of H(w) at w = 2*pi*k/N

Hm = abs(H);              % extract the magnitude of H[k]
Hp = angle(H);            % extract the phase angle of H[k]

Hr = Hm.*exp(j*Hp);       % Reconstruct H[k] from its magnitude and phase

hr = ifft( Hr, N);        % reconstruct h[n] from the reconstructed Hr[k]

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.