ComfyUI-ProgressiveBlend

ComfyUI-ProgressiveBlend

A collection of custom nodes for ComfyUI that enable progressive blending and color matching effects across image batches/video frames.

Features

This node pack includes two powerful nodes for creating smooth transitions and color grading effects:

1. Progressive Image Batch Blend

Blends two image batches with a progressively changing blend factor from 0.0 to 1.0 across the sequence.

Use Cases:

• Creating smooth transitions between two videos
• Morphing effects between different image sequences
• Artistic blend effects with customizable curves

Inputs:

• images1: First image batch
• images2: Second image batch
• blend_curve: Blending curve type (linear, ease_in, ease_out, ease_in_out, ease_out_in)
• reverse: Reverse the blend direction

Output:

• images: Blended image batch

2. Progressive Color Match Blend

Applies progressive color matching from reference images across a target batch, creating smooth color transitions.

Modes:

• Both references: Transitions from start to end reference colors
• Start reference only: Fades from color matched to original colors
• End reference only: Fades from original to color matched colors
• No references: Returns original images unchanged

Use Cases:

• Color grading transitions in videos
• Matching footage shot under different lighting conditions
• Creating artistic color progression effects
• Automatic color correction with smooth transitions
• Creative pulse effects with single reference and ease_in_out curves

Required Inputs:

• target_images: Target image batch to process
• method: Color matching algorithm (mkl, hm, reinhard, mvgd, hm-mvgd-hm, hm-mkl-hm)
• strength: Strength of color matching effect (0.0-10.0)

Optional Inputs:

• start_reference: Reference image for start color palette
• end_reference: Reference image for end color palette
• multithread: Enable multithreading for faster processing
• blend_curve: Blending curve type
• reverse: Reverse the blend direction

Output:

• images: Processed image batch with progressive color matching

Installation

Method 1: ComfyUI Manager (Recommended when available)

Once this node pack is registered with ComfyUI Manager, you'll be able to install it directly from the manager interface.

Method 2: Manual Installation

1. Navigate to your ComfyUI custom nodes directory:

cd ~/ComfyUI/custom_nodes/

2. Clone this repository:

git clone https://github.com/[your-username]/ComfyUI-ProgressiveBlend.git

3. Install required dependencies:

cd ComfyUI-ProgressiveBlend
pip install -r requirements.txt

4. Restart ComfyUI

How It Works

Progressive Blending

Both nodes use a progressive blending system where each frame in a batch gets a unique blend factor based on its position:

• Frame 0: blend_factor = 0.0 (100% first input)
• Frame N: blend_factor = 1.0 (100% second input)
• Frames 1 to N-1: Linear or curved interpolation

Blend Curves

• linear: Constant rate of change
• ease_in: Starts slow, accelerates (quadratic)
• ease_out: Starts fast, decelerates (quadratic)
• ease_in_out: Slow at both ends, fast in middle (cubic)
• ease_out_in: Fast at both ends, slow in middle (inverse S-curve)

Color Matching Methods

The Progressive Color Match Blend node uses the color-matcher library with these methods:

• mkl: Monge-Kantorovich Linearization - Good general purpose
• hm: Histogram Matching - Fast and simple
• reinhard: Reinhard et al. approach - Classic method
• mvgd: Multi-Variate Gaussian Distribution - Statistical approach
• hm-mvgd-hm: Compound method, often best results
• hm-mkl-hm: Alternative compound method

Examples

Example 1: Video Transition

Create a smooth transition between two video clips:

1. Load two video batches
2. Connect to Progressive Image Batch Blend
3. Set blend_curve to "ease_in_out" for smooth transition

Example 2: Color Grading Transition

Apply a day-to-night color grade transition:

1. Load a day reference image as start_reference
2. Load a night reference image as end_reference
3. Load your video as target_images
4. Use Progressive Color Match Blend with method "hm-mvgd-hm"

Technical Details

• Thread-safe: Supports multithreading for batch processing
• Memory efficient: Processes frames individually to minimize memory usage
• GPU compatible: Works with CUDA tensors when available
• Error handling: Gracefully handles edge cases and dimension mismatches

Credits & Attribution

Color Matching Implementation

The color matching functionality in the Progressive Color Match Blend node is based on the ColorMatch node from:

• ComfyUI-KJNodes by Kijai
• Repository: https://github.com/kijai/ComfyUI-KJNodes
• License: GPL-3.0

Color-Matcher Library

This project uses the color-matcher library:

• Repository: https://github.com/hahnec/color-matcher
• Provides the underlying color transfer algorithms

License

This project is licensed under GPL-3.0 License, in compliance with the inherited license from ComfyUI-KJNodes.

Contributing

Contributions are welcome! Please feel free to submit issues or pull requests.

Support

For issues or questions:

1. Check the ComfyUI documentation
2. Review the examples in this README
3. Submit an issue on the repository

Version History

• v1.1.0: Added Progressive Color Match Blend with multi-threading support
• v1.0.0: Initial release with Progressive Image Batch Blend

---

<p align="center"> Made with ❤️ for the ComfyUI Community </p>