> As the probes twinkle on and off, computational models estimate exactly where each molecule is located — and reconstruct a high-resolution image of the sample.

How do time and motion fit in with these techniques? I'm dimly aware that the molecular machinery inside cells moves pretty fast, and that a lot of things move around randomly. In normal size ranges that kind of thing would naturally make it hard to get a clear picture. Do these imaging techniques require that stuff be frozen or specially prepared? Or do the techniques themselves work so fast that they can get a snapshot regardless?