Bounded-Magnitude Discrete Fourier Transform [Tips & Tricks]

Analyzing the magnitude response of a finite-length sequence is a ubiquitous task in signal processing. However, the discrete Fourier transform (DFT) provides only discrete sampling points of the response characteristic. This work introduces bounds on the magnitude response, which can be efficiently computed without additional zero padding. The proposed bounds can be used for more informative visualization and inform whether additional frequency resolution or zero padding is required.