Installation Guide

Quick start setup for FXTD Studio Radiance

Quick Install

1

Clone the Repository

Navigate to your ComfyUI custom nodes directory:

cd ComfyUI/custom_nodes
git clone https://github.com/fxtdstudios/radiance.git

2

Install Dependencies

Install the required packages using pip:

cd radiance
pip install -r requirements.txt

Note: This will install core requirements. Optional features (like EXR or Depth Maps) may require extra steps below.

3

Restart ComfyUI

Restart your server. Check the console for:

[FXTDStudio Radiance]  Loaded 50+ nodes (v2.1.0)

System Requirements

Core Framework Required

PyTorch (CUDA) GPU-accelerated processing used for tone mapping and gradients.

Pillow & NumPy Standard image handling libraries.

Feature-Specific Recommended

transformers Required for Depth Anything V2 integration.

OpenEXR Required for loading/saving 16/32-bit .exr files.

colour-science Required for advanced ACES/spectral calculations.

Manual Configuration

Installing Optional Modules

To enable the full feature set, install these libraries if they aren't picked up automatically:

# Enable Depth Map Generation
pip install transformers

# Enable Professional EXR workflows
pip install OpenEXR

# Enable Advanced Color Science
pip install colour-science

Troubleshooting OpenEXR on Windows

If pip install OpenEXR fails on Windows due to missing C++ compilers, try:

pip install openexr-python

Or download a pre-built wheel for your Python version.

Nuke Bridge Setup

Connect ComfyUI to Foundry Nuke for real-time image and depth map transfer.

1

Copy the Server Script

Copy start_nuke_server.py from the Radiance scripts/ folder to a location Nuke can access (e.g. your .nuke directory):

radiance/scripts/start_nuke_server.py  →  ~/.nuke/start_nuke_server.py

2

Source the Script in Nuke

Open Nuke's Script Editor and run:

exec(open("/path/to/start_nuke_server.py").read())

Or add this line to your menu.py / init.py so it starts automatically.

3

Configure IP & Port

By default, the bridge listens on 127.0.0.1:8765. To change this, edit the variables at the top of start_nuke_server.py:

HOST = "127.0.0.1"
PORT = 8765

Make sure the same IP/port is set in the RadianceNukeBridge node inside ComfyUI.

Installation Complete?

Head over to the Tutorials to build your first professional grading pipeline, or check Node Reference to verify all nodes are loaded.