Signal Processing Applications

Complex Analysis · Axiom Academy

Your phone turns a wave of sound into numbers, then asks which frequencies are in it — and the answer comes from the same roots of unity you have been studying. Here is how, in three moves. A sound is a wave — really the real part of a rotating exponential e iωt . Your device cannot store a smooth curve, so it grabs N evenly spaced samples. Set the frequency, hit Go, and watch the samples land. Those N sample slots are not random — they are the n-th roots of unity W = e −2πik/n , spread evenly around the unit circle. The FFT calls them "twiddle factors." Drag N and watch them snap into place. Now the payoff. Multiply the samples by each root of unity and add them up — that is the DFT, X k = Σ x n e −2πikn/N . Slide the bin k and watch its size |X k | spike exactly at the signal's hidden frequency. Three moves: sample the wave, recognize the sample slots as the roots of unity , and sum against each one to read off the frequencies. The Fast Fourier Transform does exactly this billions of times a second — anywhere a signal is decomposed into frequencies: audio (equalizers, noise reduction), communications (WiFi, 4G/5G), and medical imaging (MRI, CT). Same roots of unity, every time.

This is the written version of the interactive lesson above. See the full Complex Analysis course.