The Fourier transform, in plain words

The Fourier transform answers one question: which frequencies is my signal made of, and how much of each? That is it. Everything else is bookkeeping.

A signal is a recipe

Take any signal in time. Fourier’s claim is that you can rebuild it by adding together pure sine and cosine waves of different frequencies, each with its own amount. The transform hands you that recipe:

X(j\omega) = \int_{-\infty}^{\infty} x(t)\, e^{-j\omega t}\, dt

The complex exponential $e^{-j\omega t}$ is a “probe” for the frequency $\omega$ . Multiply your signal by it and integrate, and you measure how much of that frequency is hiding inside. $X(j\omega)$ large means lots of that frequency; near zero means almost none.

A cosine is two spikes

The cleanest case: a pure cosine has exactly one frequency, $\omega_0$ . So its spectrum is not a smear, it is just two sharp impulses, at $+\omega_0$ and $-\omega_0$ :

A\cos(\omega_0 t) \ \longleftrightarrow\ A\pi\,\bigl(\delta(\omega - \omega_0) + \delta(\omega + \omega_0)\bigr)

(The negative frequency is a bookkeeping partner that keeps the maths real; do not lose sleep over it.) A sine looks the same, but the two spikes are imaginary and point opposite ways.

Time and frequency are two views

The same signal, seen two ways. A key habit: narrow in one domain means wide in the other. A short, sharp pulse in time spreads across many frequencies; a long, slow signal is concentrated in a few. That is why a perfect box in time becomes a sinc that rings on forever in frequency.

Series vs transform

If a signal repeats with period $T_0$ , it only needs frequencies that are multiples of one base frequency, so its spectrum is a set of discrete lines: that is the Fourier series, with coefficients $c_k$ . Let the period grow to infinity ( $T_0 \to \infty$ ) and those lines crowd together into a continuous curve: that is the Fourier transform. Same idea, one periodic and one not.