Congratulations to Joel for the publication of his paper Signatures of shear-layer unsteadiness in proper orthogonal decomposition in Experiments in Fluids. This paper addresses what has been a long-standing personal bugbear of mine: We use Proper Orthogonal Decomposition to produce reduced-order models of resonant flows, and it performs this task admirably. Typically you can represent the resonant behaviour with only two of the orthogonal modes. However a third mode often shows up, sometimes with similar energy to the first two, that does not appear directly involved in the resonant process. In this paper, Joel finally provides a clear explanation for the existence of this mode, as well as demonstrating how its presence or absence can offer information about variance in the resonance process.