aoleg/photographic-styles-and-wildcards-for-krea-2 — explained in plain English
Analysis updated 2026-05-18
Apply a realistic vintage film look to AI generated images.
Combine lighting, composition, and color grade prompts for consistent photo styling.
Load the wildcard file into a dynamic prompting tool for randomized photographic styles.
Write better sentence based prompts for language model style image encoders.
| aoleg/photographic-styles-and-wildcards-for-krea-2 | 000madz000/payload-test-api-route-handler | 0marildo/imago | |
|---|---|---|---|
| Stars | 3 | 3 | 3 |
| Language | — | TypeScript | Python |
| Setup difficulty | easy | easy | easy |
| Complexity | 1/5 | 2/5 | 2/5 |
| Audience | general | developer | general |
Figures from each repo's GitHub metadata at analysis time.
Just copy the CSV or YAML file into the appropriate folder of an already installed image generation tool.
This repository is a library of writing prompts for AI image generation, not a piece of software. It is meant to be dropped into an image generation tool called Forge Neo (and similar tools like Comfy or SwarmUI) to give users a large menu of photographic looks to apply to their generated images, such as specific old film stocks, decades, lighting setups, and camera angles. The author explains that most existing style packs were written for an older kind of image model that only understands loose lists of keywords, like "masterpiece, best quality." Newer models such as Krea-2 use a text encoder that is closer to a real language model and actually reads full sentences. Feeding that kind of model old style keyword lists does little or nothing, and can even confuse it. So instead, every entry in this pack is written as a plain, descriptive sentence explaining what a look actually produces, for example describing the warm skin tones and fine grain of a specific film instead of just naming it. The styles are organized into thirteen independent categories, covering things like film stocks, photography decades, lighting, composition, camera position, color grading, and depth of field. Because each category only affects one aspect of the image, several of them can be combined at once on the same photo, for instance stacking a film stock with a lighting style and a camera angle. The README notes a few categories do conflict if combined carelessly, and explains which ones. The pack is provided in two formats: a CSV style file, and a YAML wildcard file for the "film" folder used by dynamic prompting tools. There is a newer, more compact version of the styles alongside an older, more descriptive version, and the author found the newer one causes less unwanted change to the overall composition of an image. This is a prompt and data library rather than a program, aimed at people who already use AI image generators and want more precise control over photographic style. The README does not state a license.
A library of writing prompts and style presets for AI image generators, letting users apply realistic photographic looks like specific film stocks, eras, and lighting to generated images.
The README does not state a license.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly general.
This repo across BitVibe Labs
double-check against the repo, no cap.