DazzleNodes Collection

DazzleNodes

A curated collection of image/latent creation and transformation custom nodes for ComfyUI. Each node is independently developed and maintained, but packaged together for convenient installation.

Included Nodes

Smart Resolution Calculator (standalone: Github & Comfy Registry)

Flexible resolution and latent generation with intelligent aspect ratio handling, spectral noise blending, and DazzleCommand seed orchestration.

Features:

• Toggle-based dimension control with automatic missing value calculation
• 10 fill patterns (5 built-in + 5 DazNoise types) with spectral blending
• Image transforms: distort, crop/pad, scale/crop, scale/pad
• Direct latent output for all VAE types (SD1.5/SDXL/Flux/WAN/Qwen)
• Per-node DazzleCommand integration for multi-pipeline seed control

Status: Published standalone in ComfyUI Registry and in DazzleNodes package

Preview Bridge Extended (standalone: Github & Comfy Registry)

Enhanced Preview Bridge with 3-layer mask editing, MaskEditor integration, and DazzleCommand orchestration. IMAGE and LATENT+VAE variants.

Features:

• Optional mask input with 3-layer editing (input/additions/subtractions)
• Cache-compatible deterministic preview filenames
• Smart blocking with DazzleCommand play/pause gate control
• Multi-DazzleCommand support — each PBE reads its own connected DC's state

Status: Published standalone in ComfyUI Registry and in DazzleNodes package

Dazzle Command (Github)

Workflow orchestration node with play/pause toggle. Coordinates seed control (SmartResCalc) and execution gates (Preview Bridge Extended) from a single click — without re-executing expensive generation steps.

Features:

• Play/pause buttons for one-click workflow control
• 5-mode seed control: one run then random, new seed each run, reuse last seed, keep widget value, SmartResCalc decides
• Per-node state registry — multiple DazzleCommands operate independently
• Cache-transparent operation (repeat runs cached in 0.01s)
• Seed bar entry with priority over SmartResCalc widget

Status: Available from DazzleNodes package in the Comfy Registry; download from Github for standalone install

Dazzle Switch (Github)

Smart switch node with dropdown-based input selection and INT override for cascading workflows. Route any data type through a named dropdown instead of moving noodles.

Features:

• Dynamic input expansion (auto-grow/shrink slots as you connect)
• select dropdown populated with connected input names (supports user renames)
• select_override INT input for programmatic cascading across chained switches
• Type detection via graph walking with output type label
• Slot reordering (right-click Move Up / Move Down) for cascading alignment
• Active slot highlight for visual feedback
• Label cache preserves user renames across disconnect/reconnect cycles

Status: Available from DazzleNodes package in the Comfy Registry; download from Github for standalone install

Fit Mask to Image (standalone: Github)

Automatically resizes masks to match image dimensions for inpainting workflows (for when the ksampler output dimensions no longer matches the input mask dimensions)

Features:

• Replaces 10-node workflow with single node
• Automatic dimension matching
• Preview output for verification
• Latent masking support
• Nearest-exact scaling for quality preservation

Status: Available from DazzleNodes package in the Comfy Registry; download from Github to ComfyUI\custom_nodes for standalone install

Plasma Noise Generators (fork: Github)

GPU-accelerated noise generation with 100-200x speedup over CPU implementations. Fork of comfy-plasma by Jordach with PyTorch optimization (PR #7).

Features:

• OmniNoise unified generator - All 5 noise types in one node with dynamic widgets
• 5 noise types: Plasma, Random, Grey, Pink, Brown
• Two random distributions: Uniform (TV Static) and Gaussian (Centered Gray)
• GPU acceleration via PyTorch tensors
• Automatic CPU fallback
• Per-channel RGB clamping
• Backwards compatible with original RandNoise node

Note: This fork may be removed if the upstream project merges the optimization changes.

Installation

Method 1: ComfyUI Manager (Recommended)

Search for "DazzleNodes" in ComfyUI Manager and click Install

Method 2: Git Clone with Submodules

cd ComfyUI/custom_nodes
git clone --recursive https://github.com/DazzleNodes/DazzleNodes.git

⚠️ Important: The --recursive flag is required to download all node submodules!

Method 3: Manual Submodule Setup

cd ComfyUI/custom_nodes
git clone https://github.com/DazzleNodes/DazzleNodes.git
cd DazzleNodes
git submodule update --init --recursive

Method 4: Symlink for Development

If you're developing DazzleNodes locally:

cd ComfyUI/custom_nodes
mklink /D DazzleNodes C:\code\DazzleNodes\local

Usage

After installation, restart ComfyUI. All nodes will appear under the DazzleNodes category in the node menu.

Example Workflows

See examples/ directory for workflow JSON files demonstrating node usage.

Updating

Update All Nodes

cd ComfyUI/custom_nodes/DazzleNodes
git pull
git submodule update --remote

Update Individual Node

cd ComfyUI/custom_nodes/DazzleNodes/nodes/smart-resolution-calc
git pull

Architecture

DazzleNodes uses git submodules for local development to maintain clean, independent histories for each node:

DazzleNodes/local/
├── __init__.py              # Aggregator that imports all nodes
├── nodes/                   # Node submodules
│   ├── smart-resolution-calc/    # Submodule → separate git repo
│   ├── fit-mask-to-image/        # Submodule → separate git repo
│   ├── dazzle-comfy-plasma-fast/ # Submodule → separate git repo
│   ├── dazzle-command/             # Submodule → separate git repo
│   ├── dazzle-switch/              # Submodule → separate git repo
│   └── preview-bridge-extended/    # Submodule → separate git repo
└── examples/                # Collection-level examples

Why Submodules?

• Each node maintains independent git history
• Nodes can be extracted to standalone repos later
• Version locking prevents unexpected breakage
• Supports selective node development

The aggregator uses importlib to dynamically load each submodule at startup, merging all NODE_CLASS_MAPPINGS into a single namespace. This handles Python's inability to import from hyphenated directory names and provides fault isolation so one broken node doesn't take down the collection.

For a detailed walkthrough of the loading system — including how to build your own multi-node collection — see docs/dynamic-node-loading.md.

Development

Working on a Node

Each node is developed in its own repository:

• Smart Resolution Calculator: C:\code\smart-resolution-calc-repo\local
• Fit Mask to Image: C:\code\ComfyUI-ImageMask-Fix\local

# Navigate to node's repository
cd nodes/fit-mask-to-image

# Make changes and commit as usual
git add -A
git commit -m "Add new feature"
git push

# Update collection to use new version
cd ../..
git add nodes/fit-mask-to-image
git commit -m "Update fit-mask-to-image to latest"
git push

Dev Mode and Local Paths

The dev_mode.py script toggles nodes between dev mode (symlinks to local source repos) and publish mode (submodules). Since .gitmodules frequently contains GitHub URLs rather than local paths, you can configure local repo locations via dev_mode_local.yaml:

python scripts/dev_mode.py dev all --local    # Switch to dev mode using local paths
python scripts/dev_mode.py publish all         # Switch back to publish mode

For setup instructions, configuration options, and all available developer scripts, see docs/developer-tools.md.

Adding a New Node

1. Create node repository
2. Add as submodule:

   git submodule add <repo-url> nodes/<node-name>
   

   Note: git submodule add can silently fail on some Git versions (observed on Git 2.52.0 for Windows). If the command exits successfully but the submodule directory is empty or missing, see docs/git-submodule-workaround.md for diagnostic steps and workarounds.

3. Update __init__.py to import the new node (follow the existing load_node_module() pattern)
4. Test in ComfyUI
5. Update version (version.py, pyproject.toml) and CHANGELOG.md
6. Commit changes

Troubleshooting

Nodes Not Loading

Symptom: ComfyUI says "No nodes loaded!" or specific nodes are missing.

Solution: Initialize submodules:

cd ComfyUI/custom_nodes/DazzleNodes
git submodule update --init --recursive

Empty Node Directories

Symptom: nodes/ subdirectories exist but are empty.

Solution: You cloned without --recursive flag. Run:

cd ComfyUI/custom_nodes/DazzleNodes
git submodule update --init --recursive

Import Errors

Symptom: Python import errors in console.

Solution:

1. Verify submodules are initialized
2. Check each node's __init__.py exists
3. Restart ComfyUI completely
4. Check ComfyUI console for detailed error messages

Contributing

Contributions welcome! Each node has its own repository and contribution guidelines. See individual node READMEs for details.

For collection-level issues (installation, documentation, architecture), open an issue in this repository.

Like the project?

License

DazzleNodes, Copyright (C) 2025 Dustin Darcy

Collection-level code is licensed under the MIT license - see the LICENSE file for details.

See individual node directories for specific licenses.