1) when
you are doing C-NMR, you have the possibility that the spins of the
carbon and that of the protons will couple, and the spectrum would
then be very hard to understand. In this method, you get rid
of any coupling by following the initial pulse with a pulse that randomizes
the spins of the protons; in this way, the signal averaging
gets rid of any spin coupling between them. It does make it
easier to see how many carbons there are, although it makes it impossible
to integrate peak heights to figure out the abundances of each equivalent
carbon.