⭐ Star Beta Nodes for ComfyUI

Welcome to Star Beta Nodes – a staging ground for experimental custom nodes for ComfyUI. Most beta nodes have now been released to the main ComfyUI_StarNodes repository. This beta repo currently contains the nodes listed below.

🚀 Available Nodes

⭐StarNodes/Image And Latent

• ⭐ Star Qwen Image Ratio (StarQwenImageRatio)

  • Dropdown-driven aspect ratio selector that outputs:
    • An empty SD3 latent sized for the selected ratio (dict with samples tensor)
    • Width and height integers
  • Supported ratios:
    • 1:1 → 1328 × 1328
    • 16:9 → 1664 × 928
    • 9:16 → 928 × 1664
    • 4:3 → 1472 × 1104
    • 3:4 → 1104 × 1472
    • 3:2 → 1584 × 1056
    • 2:3 → 1056 × 1584
    • 5:7 → 1120 × 1568
    • 7:5 → 1568 × 1120
  • Inputs:
    • batch_size (INT, default 1): latent batch dimension
    • custom_width, custom_height (INT): used when selecting "Free Ratio (custom)", both enforced to be divisible by 16

• ⭐ Star Qwen / WAN Ratio (StarQwenWanRatio)

  • Select preset sizes for Qwen, WAN HD, or WAN Full HD, or auto-pick the closest ratio from an input image
  • Outputs an empty latent plus width/height
  • Optional: image to enable "use_nearest_image_ratio"

• ⭐ Star Qwen Image Edit Inputs (StarQwenImageEditInputs)

  • Stitch up to 4 input images and prepare Qwen-sized latent
  • Stitch rules:
    • 1 image: keep size, downscale longest side to 1328 if larger
    • 2 images: 1×2 concat, cap height to 1328
    • 3/4 images: 2×2 grid, letterboxed cells, final 1328×1328 (white background)
  • Resolution dropdown:
    • "Use Best Ratio from Image 1" or choose a Qwen preset / Free custom
  • Outputs: stitched IMAGE, LATENT, width, height

• ⭐ Star Apply Overlay (Depth) (StarApplyOverlayDepth)

  • Blends a filtered image over a source image using either a provided mask or a depth/greyscale image converted to a mask
  • Options include strength, invert mask, preview, and pixel-based Gaussian blur of the mask
  • Outputs: composited IMAGE
  • Inputs (required): source_image, filtered_image, strength
  • Inputs (optional): depth_image or mask (one required), invert_mask, show_preview, blur_mask_px

📦 Installation

Method 1: Git Clone (Recommended)

1. Navigate to your ComfyUI custom nodes directory:

   cd ComfyUI/custom_nodes
   

2. Clone this repository:

   git clone https://github.com/Starnodes2024/ComfyUI_StarBetaNodes
   

3. Restart ComfyUI to load the new nodes.

Method 2: Manual Installation

1. Download the ZIP file from the repository
2. Extract to ComfyUI/custom_nodes/ComfyUI_StarBetaNodes/
3. Restart ComfyUI

🎯 Usage Examples

Star Qwen Image Ratio

• Select a ratio from the dropdown
• Connect the latent output to SD3.5/SD3 workflows expecting a LATENT dict ({"samples": ...})
• Use the width and height outputs to configure downstream nodes
• Increase batch_size to generate batched latents in one step

Star Apply Overlay (Depth)

• Provide source_image and a filtered_image to overlay
• Connect either a mask or a depth_image (greyscale or RGB will be converted)
• Optionally toggle invert_mask, enable show_preview to save a small mask preview, and adjust blur_mask_px for smoother transitions

🔄 Node Categories

All nodes are organized under:

• ⭐StarNodes/Image And Latent – Image/latent utilities

Note: Previous beta nodes (image loaders, video frame tools, panorama saver, etc.) have been migrated to the main repository: Starnodes2024/ComfyUI_StarNodes.

🐛 Beta Testing & Feedback

As this is a beta testing repository, your feedback is crucial! Please:

• Report any bugs or issues
• Suggest new features or improvements
• Share your use cases and workflow examples

📋 Requirements

• ComfyUI (latest version recommended)
• Python 3.8+
• PyTorch (bundled with ComfyUI env)
• PIL/Pillow (for preview saving in overlay node)
• Standard ComfyUI dependencies

🤝 Contributing

Contributions are welcome! Feel free to:

• Submit bug reports
• Propose new node ideas
• Submit pull requests for improvements

📄 License

This project is open source. Please check the repository for specific licensing information.

---

⭐ Happy beta testing! Your feedback helps make these nodes better for everyone.