Inverse FFT is a function which converts complex spectrum in a time-domain signal, i.e. the function which reverts FFT result back in the origin signal. To apply this function, you need to provide complex spectrum with real and imaginary components. Applying inverse-FFT directly to spectrum derivate like magnitude spectrum or power spectrum is not possible because phase information is not available anymore.
Inverse FFT in SIGVIEW is included in the Signal Calculator as a binary operator. It should be applied to a real and imaginary spectrum part as shown in the screenshot below:
The result would look like this in the Control Window:
Real and imaginary spectrum parts can be obtained in SIGVIEW in two ways:
- Calculate FFT twice for a signal and select "Real part" and "Imaginary part" in the "Show result as" field for these two spectra respectively. For both spectra, windowing and zero-expand options should be turned off if you wish to get exactly the same signal after inverse FFT. After calculating inverse FFT as shown above, the resulting signal should be the same (up to the calculation precision) as the origin signal. Please note that the DC component of the origin signal will also be lost, i.e. the result of the inverse FFT will probably be translated for a constant offset on the Y-axis from the origin signal. To avoid this, you can use the Remove DC offset function on the origin signal before calculating its spectrum.
- SIGVIEW distribution includes a ready-to-use custom tool named "RealAndImagFFT" which will automatically calculate real and imaginary FFT parts in two separate windows, as described above. The tool is available from the main menu option "File/Use custom tool...". You can use it to prepare data for the inverse FFT function. Since this tool uses no zero-padding when calculating FFTs, we recommend using power-of-2 signal length for best performance.