This study represents one of my first successful AnimateDiff workflows. The subject is a vintage, golden Coty lipstick—a classic physical object reimagined in a digital state of flux.

The Technical Pipeline

As one of my earliest experiments with neural motion, this was a foundational test of ControlNet and early AnimateDiff nodes. I focused on:

• Maintaining Form: Using the golden casing as a structural anchor.

• Simulating Heat: Forcing the AI to hallucinate a liquid state that respected the lighting of the original object.

A Digital Fossil

Because this was created during the early, "experimental" stages of the software, the specific combination of nodes and weights used here is no longer functional. In the fast-moving world of AI, this workflow has become a "legacy" piece—a specific look from the beginning of the neural motion era that cannot be replicated with today's more polished, stable tools.