The Fast Fourier Transformation (FFT) is a variety of Discrete Fourier Transformation (DFT) which has been optimised in order to reduce the number of operations required in an implementation of the DFT. The procedure involves normalising the number of discrete samples or `points' to a power of two (a common figure is 
, resulting in a 5-bit or 512 point FFT). By a binary `divide and conquer' strategy familiar from other binary operations (e.g. binary search), the operation is reduced from 
to 
complexity.