Liminal Phantom

Please consider supporting me on Patreon, where the free tier unlocks everything, to keep up with models, workflows, tools, process documentation, and artwork.

***Recommended LoRA Strength: 4.5***

Liminal Phantom

Recommended Settings

Base Model: Flux.1-dev

Resolution: 1138x640 (horizontal) or 640x1138 (vertical)

Trigger Word: "synthesia"

LoRA Strength: 4.5

Flux Guidance: 1.2

Max Shift: 1.

Base Shift: 8.0

Sampler: res_2m

Scheduler: beta

Steps: 35

The feeling of being connected to another without having ever known them. Making eye contact with a passing wild animal in the forest. Seeing a familiar stranger in a dream.

The images from this film were created with the Flux.1-dev LoRA "Liminal Phantom" and animated with WAN2.2-I2V-A14B.

This project is a further study in using generative ML tools to create new aesthetics. Building original visual languages from old data is both the challenge and the promise of these tools. The Liminal Phantom model is twice distilled and deliberately overfitted. The first synthetic dataset was built from a multi-LoRA pipeline in ComfyUI, used to train the initial version of the model. That model then fed a second image pipeline to generate the training data for the final version, the one that produced this look.