Based on some questions I had in talks about how we dedisperse pulsar data and why it is harder to detect fast pulsars rather than slow ones, I made these movies: a slow strong pulsar, the same slow pulsar but much fainter (as per a real search), and a weak MSP. Note how much narrower the significance peak (i.e. reduced-chi2) is as a function of Dispersion Measure for the MSP! Here is the python script that made the movies if you are curious.