Mathematics
Fourier Transform
Quick fact
The Fourier transform is so powerful that it makes modern JPEG image compression possible—by separating an image into low‑ and high‑frequency components, it can discard invisible details to shrink file size.
Why this is interesting
When you hear a chord on a piano, your brain instantly recognizes the blend of notes. But how can we mathematically separate a complex sound into its pure musical tones?
Read the full explanation
Understanding Fourier Transform
Imagine a prism: white light enters, and out come the rainbow colors. The Fourier transform does something similar for signals. Any signal—like a sound wave, an electrical pulse, or a stock market trend—can be thought of as a mixture of sine waves of different frequencies. The Fourier transform takes this mixture and reveals exactly which frequencies are present and how strong they are. For a periodic signal, you get a discrete set of frequencies (Fourier series). For a non‑periodic signal, you get a continuous spectrum. The transform uses the idea of correlating the signal with pure sine and cosine waves (complex exponentials) to measure similarity at each frequency.
A deeper explanation
Mathematically, the Fourier transform of a function f(t) is defined as F(ω) = ∫{-∞}^{∞} f(t) e^{-iωt} dt, where ω represents angular frequency. The complex exponential e^{-iωt} acts as a probe: by multiplying f(t) by this oscillating function and integrating over all time, the result is large only when f(t) contains that specific frequency. Why does this work? Because complex exponentials are eigenfunctions of linear time‑invariant (LTI) systems—they pass through such systems unchanged except for amplitude and phase. This property makes the Fourier transform the natural language for describing LTI systems. The inverse transform reconstructs the original signal from its frequency content. The transform is invertible, meaning no information is lost. This duality—representing a signal in both time and frequency—unlocks powerful techniques: convolution becomes simple multiplication, differential equations become algebraic equations, and we can selectively filter out noise by removing unwanted frequencies. The fast Fourier transform (FFT) algorithm made real‑world applications possible, from medical MRI to audio compression (MP3) to solving partial differential equations.