Loop Forge

A Bucket of Blood, 1959

Works, from LTX 2.5 Creative Lab LoRAs, tested locally

Frame from the 1959 film result
1959 film
Frame from the Colorization LoRA result
Colorization LoRA
Watch it in the video at 2:17
LoRA
Colorization at 1.0
Output
1152×640, 121 frames at 24fps
Sampling
8 steps, euler_ancestral
Seed
42
Generated in
5 min
Peak VRAM
11.3 GiB
Ran on
RTX 3060, 12GB (local)

What was wired in

  • A Bucket of Blood
    Guide video
    A Bucket of BloodRoger Corman, 1959. Public domain, via Wikimedia CommonsDownload MP42.8 MB
Works

This is a real black-and-white film, so there is no colour original to check against. Every colour I named in the prompt lands, from the dusty-rose cardigan to the yellow lampshade, and the faces hold through the shot. The cane comes out red without being asked for.

Reference shows a cluttered 1950s bohemian apartment where a smiling young woman with wavy shoulder-length hair, in a fitted cardigan, a wide belt and a dark skirt, holds a wicker handbag beside a grinning young man in a dark sweater and jeans, while a man in a dark suit with a cane stands by the doorway behind them and a tall sculpture draped in a sheet fills the right of the frame, rendered in monochrome film. Edited shows the same scene with natural colors restored. COLORIZE the woman with strawberry-blonde hair in a dusty-rose cardigan, a black leather belt and a charcoal skirt with a pale straw handbag, the young man with light-brown hair in a navy sweater and faded blue jeans, the man by the door in a black suit, an off-white canvas sheet over the sculpture, pale sage-green walls and cream doors, a warm yellow lampshade glowing above a dark wooden dresser, a worn red patterned sofa and warm natural skin tones. Subject identity, framing, and background geometry are identical to the reference; only color information differs between reference and edited.

Run it yourself

Download everything for this test (.zip, 2.8 MB). It holds the workflow, every input it loads and a README. Copy the inputs into ComfyUI/input, put the models below in place and drag the workflow onto ComfyUI.

The Lightricks repositories (the LTX 2.5 base model and the LoRAs) are gated. Log in to Hugging Face and accept the terms on each repository page before downloading.

Custom nodes: ComfyUI-LTXVideo at commit 3bf3ca6. Everything else is part of ComfyUI 0.37.0 (commit 0f74f7fb, frontend 1.53.6).