Zero filling increases the size of an FID by adding zeros to the end. It slightly increases digital resolution after Fourier transform and improves line shape by effectively interpolating between collected data points. It also allows the imaginary data to be reconstructed via Hilbert Fourier transformation.
Typically zero filling is used to double the number of points and is done after apodisation and linear prediction
NMRPipe manual pages for are available for zero filling.
Previous: Linear prediction
Next: Phase correction