An integral transform that maps a function or tempered distribution in the time/spatial domain to a complex-valued function of frequency, decomposing the original into sinusoidal (exponential) frequency components.

Principle

Linearity, superposition, and the dual pair: integration against e^{-iωt} (or the chosen sign convention) converts convolution to multiplication and differentiation to polynomial multiplication in the frequency variable; inversion and Plancherel/Parseval link time and frequency energy.

Demonstration

The Fourier transform of a Gaussian e^{-at^2} is another Gaussian in frequency; solving constant-coefficient linear PDEs reduces to algebraic manipulation in frequency domain and inversion to obtain the solution in time/space.

Misapplication

Applying the transform to functions without ensuring appropriate integrability, ignoring convergence or distributional interpretation, or treating pointwise products as products of transforms (convolution vs multiplication confusion).

Consequence

Access to spectral analysis: frequency-domain solving of differential equations, spectral filtering and signal processing, energy conservation across domains (Parseval), and characterization of smoothness and decay through transform behavior.

Reversal

The reverse perspective is reconstructing exact time-domain localization from global frequency information alone; extremes emphasize either pure time localization (impulsive behavior) or pure frequency localization (steady tones).

Boundary

Defined on L1 functions and extended to L2 via Plancherel, tempered distributions via Schwartz theory; different sign and normalization conventions exist; some functions require distributional interpretation and not all operations commute without hypotheses.

Semantic Tension

Tension between Fourier series (periodic, discrete spectrum) and Fourier transform (nonperiodic, continuous spectrum), and between pointwise convergence versus convergence in norm or distribution sense.

Synthesis

The Fourier transform converts problems into frequency algebra by decomposing signals into exponential basis elements, coupling linearity and inversion to enable spectral methods across analysis, PDEs, and signal processing.