Comfy Manager is SLOW to upload. sometimes a day or so. If you get ANY missing nodes, just get them from my repo Here: https://github.com/lonecatone23
All LC Notes now translate to the following:
English, 中文, 繁體中文, Русский, 日本語 ,한국어, Français, Español, عربي, Türkçe, Português (BR), فارسی, עברית, Italiano
Works with any model. Just plug it in:
LC Vision Smart Prompt Batch is a wildcard processor with a brain. It reads your whole prompt, finds the parts you picked, and rewrites only those, as real description.
You pick what changes: Location, Time of day, Weather, Lighting, Season, Hair color, Hairstyle, Outfit, Pose, Expression, Camera angle, Framing. Or type your own.
Everything else stays word for word. The model never touches it.
The details tied to the old version adapt. This is the part a wildcard can never do.
Here's a real one. The base prompt has her on a rooftop at night, with a railing and string lights above her head and a city skyline behind her. Vary Location and one of the results reads:
... standing in an ice cave at night, framed to include the frozen ledge and aurora glow above her head; ... she leans casually against the frozen ledge with one arm extended; behind her, the blurred ice walls glow with bioluminescent veins under a deep black sky pierced by northern lights ...The railing became a frozen ledge, in both places it was mentioned.
The city skyline became ice walls. The string lights became an aurora.
Her face, her two-tone hair, the parka, the necklaces, "at night" and the film look are untouched, word for word.
note: a wildcard would have given you ice cave and left the railing, the string lights and the skyline exactly where they were. In an ice cave. 🤔
One prompt in, eight out. Same girl, same parka, a new place, weather, light, pose and angle in every picture.
Side by side 🥊
Same workflow, same model, same LoRAs, same seed. The only thing that changes is where the prompts come from.
Model: Krea 2 Turbo (Krealism V3.1)
Steps: 10, cfg: 1, Sampler: euler, Scheduler: simple, LC Speed Boost on
Seed: 424242 for all 8
Wildcards: the person written out, then
__location__, __weather__, __lighting__, __pose__, __camera__picked from plain wildcard lists.Smart Prompt Batch: the Filter caption of the same girl, vary Location, Weather, Lighting, Pose, Camera angle.
Things to notice:
Framing:
Top row: straight on, same distance, four times. "close-up", "walking" and "sitting" were in the prompts. The model didn't care.
Bottom row: the angle and the distance actually change, because the prompt describes them instead of naming them.
Light:
Top row, third picture:
beach, storm, neon lighting. Neon on a beach in a storm, so the model dyed her parka hot pink. 😏Top row, fourth:
library, fog, golden hour. It's just a library.Bottom row: every light comes from something that is actually there: a sodium street glow, the desert sun, morning beams through the trees, light through water.
The prompts:
A wildcard prompt, word for word:
a young woman with freckles, long wavy blonde hair, olive-green parka with a fur-lined hood, white lace top, beach, storm, neon lighting, standing, close-upThe same girl from the Smart Prompt Batch (second picture, trimmed):
... sun-burnt eyes scanning the horizon, lips slightly parted in quiet awe ... under natural, blazing overhead sun bleaching shadows from skin and hair ... in a canyon of ochre sandstone under a merciless blue sky.note: it isn't perfect. The fourth one went underwater and gave her goggles nobody asked for. That's the one I'd skip, and with Image with Prompt I'd grab the second one and branch off it instead.
Instagram: https://www.instagram.com/synth.studio.models/
Buy me a☕ https://ko-fi.com/lonecatone
This represents many hours of work that I give away for free. If you enjoy it, Please make sure to 👍like and post to my workflows. That is how you support me. Also, feel free to ⚡tip 😉

