1. In signal processing, any effect resulting from taking a Fourier transform of a signal that is not periodic. Discrete Fourier transformations work on the assumptions that the signal is periodic and that first and last values of the signal samples will be equal. If this is not the case, there is a discontinuity in the signal and this will show up in the transform as extra spurious frequency components. A variety of techniques are available to increase the number of samples used, including padding with zero data or reflecting the signal to make it periodic. Typically, the number of samples is doubled.

2. Deviation from parallel in the lines of force representing the electric field at the edges of parallel-plate capacitors, resulting in a field that is nonuniform at the edges.