The standard explanation of resonance — push a swing at the right moment and the amplitude builds — is not wrong, but it leaves people unable to predict anything. Mould uses a flame tube, where gas escaping through a row of holes is modulated by the pressure pattern of a standing wave inside, to argue for a more useful picture built around how a system responds across a whole range of driving frequencies rather than at one magic number.
This matters more for cymatics than it first appears. Everything on a plate, in a water cell or in a bubble film is a resonance problem, and the swing analogy leads people to expect that a particular frequency has a particular shape attached to it, independently of the object. It does not. The object supplies the modes; the driving signal selects among them.
Thirteen minutes, one apparatus, and the best conceptual grounding on this list.