When the frequency domain's independent variable is labeled as a fraction of the sampling rate, the values along the horizontal axis always run between $0$ and $0.5$, since discrete data can only contain frequencies between DC ($0$) and half the sampling rate.

But why is the highest frequency pegged at exactly $0.5$ of the sampling rate? Isn't it possible to have a scenario where the highest frequency is less than half the sampling rate?

E.g., consider a speech signal that has been filtered to remove all frequencies above $3.3$ kHz and has been sampled at $10$ kHz. The highest frequency ($3.3$ kHz) is less than $0.5$ of the sampling rate ($5$ kHz). How would the frequency domain of this signal (labeled as a fraction of the sampling rate) look like?


The DFT calculates spectral components up to $f_\mathrm{s}/2$, no matter what the input signal is. If the input signal to the DFT contains (absolutely) no spectral components between $0.33 f_\mathrm{s}$ and $0.5 f_\mathrm{s}$ the corresponding DFT outputs are zero (assuming that there is no spectral leakage).

| improve this answer | |

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.