NMR and crystallography, or rather, crystallography and NMR.

I like the ultimate paragraph of the Highlight written by Burkhard Luy in Angewandte Chemie Int. Ed. (2007, volume 46, pages 4214-4216), it illustrates the complementarity of X-ray crystallography and NMR spectroscopy very well:

"With the NMR techniques developed over the past decade and the availability of corresponding crystal structures, molecular complexes of nearly unlimited size seem to be amenable to liquid-state dynamics measurements. These results are an important step in understanding the modes of operation of complicated molecular machines in biological systems."

Note that the crystal structure of the 20S proteasome complex was necessary in order to be able to interpret the NMR signals of the particle, and that the NMR results shed light on dynamic properties which crystallography had not measured.
Also note the NMR work was expensive, requiring extensive labeling, mutation to make a monomeric version - so it would presumably be only worth doing this for very important macromolecular complexes.