When the sampling frequency is increased by twice in case of the interpolation, the signal has one zero inserted after each sample. This zero stuffing introduces high frequency noise which is exactly the mirror of the low frequencies. To filter out this high frequencies a lowpass filter is needed. If the sampling frequency is 2Hz, then the lowpass filter should pass everything below 0.5Hz and stop everything above 0.5Hz. The TransBW property defines the width of the transition region of the lowpass filter centered on 0.5 Hz. The narrower the transition region, the longer the filter and more CPU demanding computation is required.

public