//
Generative tools: Stable Diffusion 1.5
Source material: A custom LoRA trained on black-and-white grids assembled from found, appropriated, and artist-created imagery, including photographs and paintings.
Process: Grid generation → grid upscaling → selection and cropping → upscaling → AI-assisted grid colorization (ChatGPT) → iterative upscaling → post-production.
Upscaling: Local tiled diffusion workflow using Juggernaut Reborn, SDXLRender LoRA, and ControlNet.
Post-production: Color correction, digital painting.
Execution: Generated and finalized in mid 2026.