Why is it important to match frequency to timescale of events?

I understand that too long of a time scale won't resolve an event clearly, and too much has the risk of exciting whatever you're trying to measure. Is there some sort of resonant effect when you match, say, a radiation frequency to the time of an event taking place? Specifically I'm applying this to detecting fluctuation phenomena.