Conor-Collins-ComfyUI-CoCoTools_IO/README.md

CoCoTools_IO

Advanced image I/O nodes for ComfyUI — professional EXR handling, multilayer support, color science, and scopes.

Installation

ComfyUI Registry (Recommended)

Search for CoCoTools_IO in the Comfy Registry or ComfyUI Manager and install directly.

Manual Installation

1. Clone the repository into your ComfyUI custom_nodes directory:

   cd ComfyUI/custom_nodes
   git clone https://github.com/Conor-Collins/ComfyUI-CoCoTools_IO.git
   

2. Install dependencies from the python_embeded/ folder (portable) or your Python environment:

   python -m pip install -r ComfyUI/custom_nodes/ComfyUI-CoCoTools_IO/requirements.txt
   

3. Restart ComfyUI

Beta Branch

The beta branch contains the latest features and fixes ahead of the stable release. To use it:

cd ComfyUI/custom_nodes/ComfyUI-CoCoTools_IO
git checkout beta

Nodes

Image I/O

Node	Description
Image Loader	Load standard formats (PNG, JPG, TIFF, WebP) with proper bit depth handling (8/16/32-bit)
Load EXR	Load EXR files with full multilayer, multichannel, and cryptomatte support
Load EXR Sequence	Batch-load EXR sequences using #### frame patterns with missing frame recovery
Load EXR Layer by Name	Extract specific layers from EXR layer dictionaries (similar to Nuke's Shuffle node)
Cryptomatte Layer	Interactive click-to-matte selection with zebra stripe preview overlay
Image Saver	Save to EXR, PNG, TIFF, JPG, or WebP with format-specific options (bit depth, compression, quality)

Image Processing

Node	Description
Colorspace Converter	Convert between colorspaces: ACES2065-1, ACEScg, ACEScct, ACEScc, sRGB, Rec.709, Display P3, Rec.2020, Adobe RGB (linear and encoded variants)
Z Normalize	Normalize depth maps and single-channel data to a target range

Scopes

Node	Description
Histogram	Per-channel histogram display (L/RGB/R/G/B/A) with log/linear toggle and bit depth estimation
Vectorscope	Chromaticity scatter display with YCbCr, HSV, and CIE xy color models, skin tone line, and gamut overlays

Features

• Multilayer EXR — full subimage and channel group support for render passes, AOVs, and embedded layers
• Cryptomatte — click-to-matte pixel picking with interactive preview canvas
• Batch sequences — #### frame patterns with configurable start/end/step and white frame placeholders for missing files
• Dynamic file browser — interactive file/folder selection for load and save nodes
• Format-aware saving — per-format widgets for bit depth (8/16/32), EXR compression (zip, zips, dwaa, etc.), and JPEG/WebP quality
• Path tokens — dynamic tokens in saver output paths for dates, filenames, and layer names

Path Tokens (Image Saver)

The Image Saver supports token-based path templating in both the file_path (directory) and filename fields. Tokens are resolved at save time.

Supported Tokens

Token	Example Output	Description
%date%	2026-03-14	Current date (ISO format)
%date:dd-MM-yyyy%	14-03-2026	Current date (custom format)
%date:yyyyMMdd%	20260314	Current date (compact)
%time%	15-30-45	Current time (HH-mm-ss)
%time:HH-mm%	15-30	Current time (custom format)
%filename%	beauty_pass	Base name of connected input image (requires source_path input)
%layer%	diffuse	Layer name from connected EXR data (requires layer_name input)

Format Placeholders

yyyy (year), MM (month), dd (day), HH (hour), mm (minute), ss (second)

Examples

file_path	filename	Result
Maps/%date:dd-MM-yyyy%	depth_0001	Maps/14-03-2026/depth_0001.exr
renders/%date%	%filename%_####	renders/2026-03-14/beauty_pass_0001.exr
%date:yyyyMMdd%_%time:HH-mm%	%layer%	20260314_15-30/diffuse.exr

To use %filename% or %layer%, connect the optional source_path or layer_name string inputs on the Image Saver node from an upstream loader. If not connected, these tokens resolve to FILENAME_MISSING or LAYER_MISSING.

Dependencies

• torch >= 1.10.0
• numpy >= 1.21.0
• colour-science >= 0.4.2
• Pillow >= 9.0.0
• opencv-python >= 4.5.0
• tifffile
• OpenImageIO >= 2.4.13.0
• rich >= 10.0.0

Third-Party Libraries and Licensing

This project uses the following third-party libraries:

• Colour Science for Python: Colorspace transformations. Licensed under the New BSD License.
• OpenColorIO: Color space transformations. Licensed under the BSD 3-Clause License.

For detailed licensing information, see THIRD_PARTY_NOTICES.md.

This project is licensed under the MIT License.