Friday, 22 April 2016

Discrete Fourier Transform

We had to compute a signal's [x(n)]  Fourier Transform and observe the Magitude Spectrum. Two cases were taken, first was a signal with length equal to 4 and second was the first signal with zero padding to make its length 8. DFT was calculated using the C code, run through Linux terminal. It was observed that DFT always give periodic results and increasing the length by zero padding of signal gives better approximation of signal and resolution of spectrum increases.As N increases frequency spacing reduces, approximation error decreases and resolution improves.
Link:-
DFT
IDFT

6 comments:

  1. MENTION THE EFFECT ON RESOLUTION

    ReplyDelete
  2. MENTION THE EFFECT ON RESOLUTION

    ReplyDelete
    Replies
    1. Resolution increases because frequency spacing reduces since length increases.

      Delete
  3. Spacing between the values reduces as the value of N increases. By appending the input sequence by zeros the resolution error reduces.

    ReplyDelete
  4. The magnitude spectrum is periodic due to the characteristic of twiddle factor

    ReplyDelete
  5. by appending more zeroes, the missing values in less point DFT are present in the DFT with more point.

    ReplyDelete