In my recent study of signal processing I came to know that we can compress images/signal using DFT/DCT. If we consider the lossy compression of audio or any image then it follows the same steps:

  1. Apply DCT/DFT on N samples of data.

  2. Collect coefficients which contains most of the signal energy i.e., a signal's DCT representation tends to have more of its energy concentrated in a small number of coefficients when compared to other transforms like the DFT.

But I have doubt that how to get these coefficients. In DFT we have $$ {\tt ORIGINAL:}\ X(e^{jw})=\sum_{n=0}^{N-1} x[n]e^{-j2{\pi}/Nn}\\ {\tt CORRECTED:}\ X[k]=\sum_{n=0}^{N-1} x[n]e^{-j2{\pi}/Nnk} $$

Here, $ \omega $ is a continuous variable. Then how to extract coefficients when $ \omega $ is a continuous variable?

  • Welcome to DSP.SE! What you have in your equation is the Discrete-Time Fourier Transform. That is not usually what is used: note that this requires infinite data... which is hard to store. The DFT is something else. – Peter K. Dec 2 '15 at 17:37
  • sorry for that..I have edited it – Virange Dec 2 '15 at 17:41
  • Well, let me correct what you've done and you'll see. – Peter K. Dec 2 '15 at 17:50
  • 1
    Note that you missed out a $k$ (discrete frequency variable) in your edit. – Peter K. Dec 2 '15 at 17:51

Your Answer


By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Browse other questions tagged or ask your own question.