anaglyphTool-Comfyui

A high-performance ComfyUI custom node for creating anaglyph 3D images from color and depth map inputs. Optimized for CUDA GPU acceleration, making it suitable for video to anaglyph conversion.

Features

• CUDA GPU accelerated processing
• Batch processing support for video frames
• Adjustable parallax and depth settings
• Real-time depth map inversion option

Installation

1. Navigate to your ComfyUI custom_nodes folder
2. Clone this repository:

git clone https://github.com/Cryptyox/anaglyphTool-Comfyui

3. Install dependencies:

pip install -r requirements.txt

only torch should be required and can be installed manually.

Usage

The node accepts the following inputs:

• image: Input image or video frame (RGB format)
• depthmap: Corresponding depth map (grayscale or RGB)
• invert_depthmap: Toggle depth map inversion (default: True)
• divergence: Controls the 3D effect strength (-10.0 to 10.0, default: 2.0)
• zero_parallax_depth: Sets the focal plane (0.0 to 1.0, default: 0.5)

Requirements

• ComfyUI
• PyTorch with CUDA support (recommended)
• CPU fallback available

Recommended Workflow

Basic Usage

Basic workflow showing image and depth map inputs

Basic workflow for single image conversion using depth-anything-V2

Results

| Original | Depth Map | Anaglyph Result | |:--------:|:---------:|:---------------:| | | | |

Video Processing Example

Example workflow for batch processing video frames

I highly recommend using Depth Anything Tensorrt for the depth map creation to save time. The batch size is a huge factor when it comes to speed. Increase the batch size until your VRAM is full. On a 3090 with 24gb of VRAM you can batch process 500 480p Frames in 0,5s.

Note: You'll need red-cyan 3D glasses to view the anaglyph effects.

License

This project is licensed under CC BY-NC-SA 4.0 - see the LICENSE file for details.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Acknowledgments

Big shoutout to https://github.com/mikeymcfish who created https://github.com/mikeymcfish/FishTools. This was my original inspiration as I was very fond of his node, but needed something capable of bulk processing.