Residual dipolar couplings: are multiple independent alignments always possible?