If $\omega$ is the frequency of a continuous time sinusoidal signal, $-\infty < \omega < \infty $.

But when it comes to discrete time sinusoidal signals, the range becomes limited : $-\pi \le \omega_{discrete} \le \pi $.

Does that mean discrete sinusoidal signals cannot represent all frequencies possible ? What is the true picture here ?


It is important to realize that in the discrete-time case, $\omega$ is normalized by the sampling frequency $f_s$:

$$\omega=2\pi f/f_s$$

The inequality in your question is basically the sampling theorem, which says that a discrete-time signal can only represent frequencies lower than half the sampling frequency. All higher frequencies will be folded back into the frequency band $[0,f_s/2]$ (this is called aliasing). If you need higher frequencies then you need to increase the sampling rate.

  • $\begingroup$ Essentially, once we select a sampling rate, it constrains the representable frequency range. Correct ? $\endgroup$ – curryage May 19 '14 at 9:59
  • $\begingroup$ @curryage: Right. $\endgroup$ – Matt L. May 19 '14 at 10:00

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.