Remade-AI/Crash-zoom-out

Crash zoom out LoRA for Wan2.1 14B I2V 480p

Overview

Abruptly zooms out from the subject to reveal the surrounding scene, creating a sudden sense of scale, surprise, or disorientation. Ideal for dramatic or comedic reveals.This LoRA is trained on the Wan2.1 14B I2V 480p model.

Features

• Trained on the Wan2.1 14B 480p I2V base model
• Consistent results across different object types
• Simple prompt structure that's easy to adapt

Community

• Generate videos with 100+ Camera Control and VFX LoRAs on the Remade Canvas.
• Discord: Join our community to generate videos with this LoRA for free

Prompt

The video begins with a close-up on a man's face, his hands tied with rope, and an anxious expression. Then, a cr34sh crash zoom out effect reveals a dark and obscure room, the man is still tied up and two men wearing balaklavas and holding guns appear to be standing behind him.

Prompt

The video begins with a close-up on the man's face, with ice covering his beard and eyelashes. He has a concerned or startled expression, his eyes are a vivid blue. A cr34sh crash zoom out effect rapidly pulls the camera back, revealing the man in a yellow jacket set in a icy landscape. The cr34sh crash zoom out effect shows his position: standing on the edge of the sea with icebergs in the background.

Prompt

The video begins with a close-up shot of a woman's face with intricate black and white tribal markings on her face, neck, and chest. Her eyes are closed and she is wearing dark red eyeshadow and lipstick. The cr34sh crash zoom out effect then begins, quickly pulling back to reveal that the woman is in a dimly lit room, with candles all around her.

Model File and Inference Workflow

📥 Download Links:

• crash_zoom_out.safetensors - LoRA Model File
• wan_img2vid_lora_workflow.json - Wan I2V with LoRA Workflow for ComfyUI

---

Recommended Settings

• LoRA Strength: 1.0
• Embedded Guidance Scale: 6.0
• Flow Shift: 5.0

Trigger Words

The key trigger phrase is: cr34sh crash zoom out effect

Prompt Template

For prompting, check out the example prompts; this way of prompting seems to work very well.

ComfyUI Workflow

This LoRA works with a modified version of Kijai's Wan Video Wrapper workflow. The main modification is adding a Wan LoRA node connected to the base model.

See the Downloads section above for the modified workflow.

Model Information

The model weights are available in Safetensors format. See the Downloads section above.

Training Details

• Base Model: Wan2.1 14B I2V 480p
• Training Data: Trained on 50 seconds of video comprised of 10 short clips (each clip captioned separately) of scenes that used the crash zoom out camera motion.
• Epochs: 25

Additional Information

Training was done using Diffusion Pipe for Training

Acknowledgments

Special thanks to Kijai for the ComfyUI Wan Video Wrapper and tdrussell for the training scripts!