.cannam@167: Next: The 1d Real-data DFT, Previous: What FFTW Really Computes, Up: What FFTW Really Computes [Contents][Index]
cannam@167:The forward (FFTW_FORWARD) discrete Fourier transform (DFT) of a
cannam@167: 1d complex array X of size n computes an array Y,
cannam@167: where:
cannam@167:
.FFTW_BACKWARD) DFT computes:
cannam@167: 
.FFTW computes an unnormalized transform, in that there is no coefficient cannam@167: in front of the summation in the DFT. In other words, applying the cannam@167: forward and then the backward transform will multiply the input by cannam@167: n. cannam@167:
cannam@167: cannam@167:From above, an FFTW_FORWARD transform corresponds to a sign of
cannam@167: -1 in the exponent of the DFT. Note also that we use the
cannam@167: standard “in-order” output ordering—the k-th output
cannam@167: corresponds to the frequency k/n (or k/T, where T
cannam@167: is your total sampling period). For those who like to think in terms of
cannam@167: positive and negative frequencies, this means that the positive
cannam@167: frequencies are stored in the first half of the output and the negative
cannam@167: frequencies are stored in backwards order in the second half of the
cannam@167: output. (The frequency -k/n is the same as the frequency
cannam@167: (n-k)/n.)
cannam@167: