.Chris@19: Next: The 1d Real-data DFT, Chris@19: Previous: What FFTW Really Computes, Chris@19: Up: What FFTW Really Computes Chris@19:
The forward (FFTW_FORWARD) discrete Fourier transform (DFT) of a
Chris@19: 1d complex array X of size n computes an array Y,
Chris@19: where:
Chris@19:
.FFTW_BACKWARD) DFT computes:
Chris@19: 
.FFTW computes an unnormalized transform, in that there is no coefficient Chris@19: in front of the summation in the DFT. In other words, applying the Chris@19: forward and then the backward transform will multiply the input by Chris@19: n. Chris@19: Chris@19:
From above, an FFTW_FORWARD transform corresponds to a sign of
Chris@19: -1 in the exponent of the DFT. Note also that we use the
Chris@19: standard “in-order” output ordering—the k-th output
Chris@19: corresponds to the frequency k/n (or k/T, where T
Chris@19: is your total sampling period). For those who like to think in terms of
Chris@19: positive and negative frequencies, this means that the positive
Chris@19: frequencies are stored in the first half of the output and the negative
Chris@19: frequencies are stored in backwards order in the second half of the
Chris@19: output. (The frequency -k/n is the same as the frequency
Chris@19: (n-k)/n.)
Chris@19:
Chris@19:
Chris@19:
Chris@19: