what is comfyui-ocio fr?

ComfyUI-OCIO adds professional color management tools to ComfyUI, a visual programming app used for building AI image and video workflows. The tools are modeled on the color nodes found in Nuke, a video editing and visual effects program used in the film industry, and they are powered by OpenColorIO, an open source color management engine, using the industry standard ACES color system. Normally, ComfyUI treats images as plain sRGB color, the format most computer screens and web images use, with no real color management behind it. This project adds nine nodes that convert between different color spaces, including the ones used by cameras from ARRI, RED, Sony, and others, apply color grading adjustments, apply display transforms or LUTs, lookup tables used to shift color, and let a user preview the result directly inside a node. Two of the nodes matter most: one reads still images, image sequences, or video files off disk and converts them into a consistent working color space, and the other writes the result back out in a properly color managed format, including formats like EXR, TIFF, PNG, JPEG, ProRes, DNxHR, h264, and hevc. The nodes also connect to ComfyUI's newer native video pipeline, so they can sit directly between video related nodes like Load Video and Save Video, working the same way whether the input is a single image or a full video clip. Installation is manual for now: the user clones the project into ComfyUI's custom_nodes folder and installs its Python requirements. Reading and writing EXR files needs an extra environment variable set before ComfyUI starts, and working with video formats like ProRes needs ffmpeg installed separately on the system. The project also ships a Docker based test setup that runs the color math checks without needing a GPU. The project is released under the MIT license, one of the most permissive open source licenses, meaning it can be freely used, modified, and redistributed, including for commercial purposes.