ErixenScenario1-ProjectNoraima

This work is incomplete, but I used a locally run AI model via ComfyUI to experiment with image-to-video generation, which helped animate parts of my original hand-drawn artwork. The process is extremely time-consuming, requires a capable GPU/CPU setup, and is not suitable for anyone seeking fast results. However, once configured, it becomes much easier to produce high-quality “template” art.My workflow involved animating a few seconds at a time, selecting frames that best matched my original designs, and redrawing over them in Clip Studio Paint EX while creating alpha channels. These assets were then imported into DaVinci Resolve for timeline editing. I also animated some backgrounds and guided motion using positive and negative prompt filtering to achieve specific results.I believe this setup could yield strong outcomes with sufficient time. I briefly tested using a Unity 6 scene snapshot to extend the animation beyond my initial ideas, but the results were unsatisfactory and would require heavy per-frame correction. Even so, this workflow could reduce development time from years as a solo animator to months—or even weeks—with enough caffeine.

http://www.nicovideo.jp/watch/sm45723735