In the current state of the code, the RFs use a zero-order-hold interpolation, which must be changed to a linear interpolation instead.

The idea is that the waveforms of both RFs and Grads must have the same kind of interpretarion.

This issue should address the problem of #200 and it is related with #167.