MSP includes an fft˜ object that computes the FFT of an input signal.
The input signal can be complex, in which case the imaginary part of the input signal is connected to the right inlet.
The result of the FFT operation is normally complex, with the real and imaginary parts being output from the left and right outlets, respectively.
By default, a 512-point FFT is computed. Optional arguments control the FFT size, the FFT “hop” or period in samples (the number of samples of input between successive FFT computations), and “phase” (described as an offset from the beginning of the period when the FFT is computed).
MSP also has an ifft˜ object that computes the inverse FFT (or IFFT) from a complex pair of inputs.
Note that the cartopol˜ object converts a cartesian complex number representation () to a polar representation ().