Larger values for step_size about crepe HOT 1 OPEN

The choice of step size depends entirely on your application/needs, and is just a way of controlling the temporal resolution of the output. Since many audio analysis tasks (e.g. speech and music analysis) typically require such small step sizes (because the pitch of human speech and singing evolves very rapidly), default is small. If, however, your application only requires a pitch estimate once a second, you can definitely set the step size to 1000 ms, there's no logical flaw in using a large step size assuming it makes sense for your downstream application.

from crepe.