|
|
--- |
|
|
tags: |
|
|
- text-to-image |
|
|
- lora |
|
|
- diffusers |
|
|
- template:diffusion-lora |
|
|
widget: |
|
|
- output: |
|
|
url: images/1764464512096__000005000_0.jpg |
|
|
text: A sunset on a beach, daramatic cloud on the sky and bright sun. |
|
|
- output: |
|
|
url: images/1764464517272__000005000_1.jpg |
|
|
text: >- |
|
|
A woman is holding a glass of wine in a pub, cinematic shot, colorful lights |
|
|
in the background |
|
|
- output: |
|
|
url: images/1764464522440__000005000_2.jpg |
|
|
text: >- |
|
|
A cinematic close-up shot of a dog, he is looking at the camera, indoor |
|
|
scene and wooden floor. |
|
|
- output: |
|
|
url: images/1764464527618__000005000_3.jpg |
|
|
text: A photo of a car explosion on a street, fire and smoke. |
|
|
- output: |
|
|
url: images/1764464532760__000005000_4.jpg |
|
|
text: A photo of a mossy rock in a rain forest, god ray coming from the sun. |
|
|
- output: |
|
|
url: images/1764464537917__000005000_5.jpg |
|
|
text: A photo of a underwater ocean, beautiful corals are spread on the surface. |
|
|
base_model: Tongyi-MAI/Z-Image-Turbo |
|
|
instance_prompt: null |
|
|
license: apache-2.0 |
|
|
--- |
|
|
# Z-image-Turbo_LogC4_lora |
|
|
|
|
|
<Gallery /> |
|
|
|
|
|
## Model description |
|
|
|
|
|
LogC4 Cinematic Foundation (LoRA) |
|
|
This LoRA is designed for Z-image-turbo to generate images using the ARRI LogC4 logarithmic transfer function. Unlike standard AI generations that output a "finished" Rec.709 look, this model produces a flat, low-contrast image that preserves maximum shadow and highlight detail. |
|
|
|
|
|
Why use LogC4 Style? |
|
|
Standard AI-generated images often suffer from "clipped" highlights or crushed blacks, making them difficult to adjust in post-production. By generating in a LogC4-style space, the resulting images act as a high-latitude base for professional workflows. |
|
|
|
|
|
Maximum Dynamic Range: Retains detail in high-brightness areas (skies, light sources) and deep shadows. |
|
|
|
|
|
Color Grading Ready: Perfect for use in DaVinci Resolve, Adobe Premiere, or Nuke. |
|
|
|
|
|
Linear Conversion: When converted to linear light space, these images provide a high dynamic range (HDR) foundation suitable for professional VFX compositing and lighting. |
|
|
|
|
|
Technical Specifications |
|
|
Base Model: Z-image-turbo |
|
|
|
|
|
Format: LoRA |
|
|
|
|
|
Transfer Function: Modeled after ARRI LogC4 |
|
|
|
|
|
Target Use: Cinematic Concept Art, Professional Photography Base, VFX Plates |
|
|
|
|
|
How to Use |
|
|
1. Generation |
|
|
Don't need any trigger word |
|
|
|
|
|
Recommended Strength: 0.6 - 1.0 (Adjust based on how "flat" you want the initial image to be). |
|
|
|
|
|
2. Post-Processing (The "Development" Phase) |
|
|
To get the best results, do not use the raw output as your final image. Instead: |
|
|
|
|
|
Import the image into a color grading suite (e.g., DaVinci Resolve, Nuke). |
|
|
|
|
|
Apply an ARRI LogC4 to Linear LUT or use a Color Space Transform (CST) node. |
|
|
|
|
|
Adjust exposure and contrast. You will notice significantly more "room" to move the colors without breaking the pixels. |
|
|
|
|
|
Use Cases |
|
|
Concept Artists: Generate plates that match the look of professional film dailies. |
|
|
|
|
|
VFX Artists: Create environment maps or textures that require a linear workflow. |
|
|
|
|
|
Colorists: Practice grading on AI-generated subjects using industry-standard color science. |
|
|
|
|
|
|
|
|
## Download model |
|
|
|
|
|
|
|
|
[Download](/Sumitc13/Z-image-Turbo_LogC4_lora/tree/main) them in the Files & versions tab. |
|
|
|
