I want to create this kind of videos in which it endlessly zooms into the image and morphs.

https://www.youtube.com/watch?v=OrNTyPAAxuE

I experimented a while ago with disco diffusion workbooks (e.g. https://www.patreon.com/posts/i-used-ai-to-66518281) but tbh. I don't remember it quite and the results haven't been really good.

I guess the idea is probably to start with a seed image, and then render out diff. sections of the video and use the last frames a s new starting images to control the journey and to have the parameters set a bit diff. e.g. changing zoom / movement through the video. E.g. like rendering 15s bits / or a musical phrase until you have the full video done. Or maybe you need some kind of checkpoint images where it should morph to inbetween?

I'm quite a rookie and would be looking for tutorials you can recommend or even better example workbooks I can run, or online platforms which simplify the process. I'm happy to join a Patreon or buy a prompt if that makes things easier/faster. But my issue is a bit that I'm not even sure what I shall google for, and whats the right process. I mostly landed on tutorials which link to civitai, but I guess this is more for morphing a real video into a certain model e.g. a comic character.