The start frame is a product of multiple evolutions.
I am still in the process of learning how to do videos, and of course the video models, like any other AI, do weird things of their own. (Did I ask for this music? No, I did not. Each animation had a different prompt, but every one asked for sounds of rippling water.) And I know that minimizing camera movement will help you keep a consistent style throughout your animation, but I don't wanna do things that way 'cause it's booooring me. Anyway, one day I'll spend some credits on Veo to see if it's worth it, but today is not that day.