What I've learned so far is, you don't have to write a prompt unless you what something specific done in the video. WAN pretty much knows what you want to do provided you give it a start and ending image. Different resolutions output different videos, more so than prompting (I think). You have to use interpolation frames to get smoother vids. There's no way around it. I haven't quit figured out sound and the best resolution to frames ratio yet, but I'm getting there.