The Simple Parallel analyser (which happens to be most used at least around me) has only two options for timing the sample taking: "Data valid on rising clock edge" and "Data valid on falling clock edge". Unfortunately I encountered many situations where this is not enough or leads to errors. An example that is already suggested is that /both/ edges can be important. What happens often here is "Data valid xxx nanoseconds /after/ the (usually) rising edge". There is no chance to set this and we get erroneous results. So the suggestion is to make it flexible with at least three, ideally four options: - rising edge, <time> after rising edge, falling edge, <time> after falling edge.