Orion4D_FXMax

🌌 Orion4D FXMax for ComfyUI

A professional post-production, color grading, and sharpness suite for ComfyUI, featuring a custom built-in web application for real-time adjustments.

Bring the Lightroom and DaVinci Resolve experience directly into your ComfyUI workflow. Stop guessing values and queuing prompts: adjust your settings visually with real-time split-screen previews.

<img width="1521" height="952" alt="image" src="https://github.com/user-attachments/assets/6c30076a-621a-4ab1-a5c4-fb265911e736" /> <img width="2546" height="1418" alt="image" src="https://github.com/user-attachments/assets/dd642988-a145-42cf-b6a2-329ffcae5b53" /> <img width="1341" height="913" alt="image" src="https://github.com/user-attachments/assets/62943d2d-7eb3-424e-937c-777ed43564b8" />

✨ Key Features

• 🖥️ The "APP" Interface: Click the Open APP button on any node to launch a full-screen or windowed editor. Features include Pan, Zoom, Before/After toggle, and an interactive split-screen comparison slider.
• ⚡ Real-Time Client-Side Preview: Adjust sliders and see the result instantly without triggering a ComfyUI generation queue.
• 📦 100% Portable Workflows: When you tweak a preset in the APP, the mathematical configuration is injected into a hidden raw_settings widget. When you share your workflow (PNG/JSON), your friends will get the exact same result, even if they don't have your local JSON presets.
• 🔬 Advanced Sharpness Algorithms: Edge-aware masks, guided filters, and midtone protections to sharpen images without introducing artifacts.
• 🎨 Cinema-Grade Color Processing: DCTL-inspired Tone Mapping, Catmull-Rom Curves, proper Skin Tone protection in Vibrance, and mathematical Film Grain.

---

🛠️ Installation

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

   git clone https://github.com/orion4d/Orion4D_FXMax.git
   

3. Restart ComfyUI.

Note: This package requires Pillow and numpy, which are already included in standard ComfyUI installations.

---

📦 Available Nodes

The suite is divided into two distinct categories: Sharpness and Color FX.

1. Sharpness Suite

• SharpnessMax (All-in-one pipeline)
• Clarity
• Texture
• Smart Sharpen
• Unsharp Mask
• High Pass

2. Color & Filters Suite

• Film Grain (Poisson, Perlin, Gaussian, Halation, Lens Distortion)
• DCTL Tone Mapper (ACES, Filmic, Reinhard, Cineon)
• Curves (Catmull-Rom interpolation for RGB channels)
• HSL & Vibrance
• Color Balance & Channel Mixer
• CSS Filters & Photo Filter
• Matrix 3x3

---

⚠️ Performance & Optimization (Important)

To achieve maximum visual fidelity, this suite uses two different processing engines. Please read this if you are generating videos (AnimateDiff) or processing large batches of 4K images.

🟢 Lightning Fast (100% GPU)

All nodes in the Sharpness Suite are written natively in PyTorch. The image data never leaves your VRAM. They benefit from full CUDA/Tensor acceleration and will run near-instantly, making them perfect for Video and Batch processing.

🟡 Moderate Processing (CPU / NumPy)

Nodes like HSL, Vibrance, DCTL Tone Mapper, Color Balance, Channel Mixer, Matrix 3x3, and Photo Filter transfer the image to the CPU to process complex mathematical color transformations via NumPy.

• Impact: Barely noticeable on single images. May add a slight delay on video batches.

🔴 Heavy Processing (CPU Bottleneck)

Nodes that require complex spatial rendering or format conversions use NumPy and PIL on the CPU.

• Film Grain: Generates procedural noise (Poisson/Perlin) pixel-by-pixel and calculates bilinear interpolations for Lens Distortion.
• CSS Filters & Curves: Require continuous conversions between arrays and PIL objects for blur algorithms and look-up tables (LUTs).
• Impact: Expect significant slowdowns on high-resolution images or large video batches.
• Advice: For video workflows, use these specific nodes sparingly or rely on native PyTorch alternatives. They are entirely fine for standard image generation.

---

💾 Presets System

Orion4D FXMax uses a JSON-based preset system stored locally in presets/.

However, you do not need to share your JSON files when sharing your ComfyUI workflows. The custom UI automatically binds your current settings into the node's metadata. Anyone loading your workflow will see the exact same result, and the node will label it as an "Embedded workflow setting".