color-assist

Natural language color grading in DaVinci Resolve works by exporting current frames, converting them to sRGB for AI analysis, and applying CDL adjustments like exposure or white balance directly to color nodes. You can simply ask it to "warm up the shadows" or "increase contrast," and the AI will perform the necessary visual analysis and apply the exact color correction values to your selected clip.

Yes, you can adjust white balance and contrast using CDL values through an AI assistant that analyzes your footage frames and automatically modifies the Color Decision List parameters on your nodes. The AI interprets your natural language requests—such as increasing saturation or correcting exposure—and translates them into precise CDL value modifications within DaVinci Resolve's Color page.

AI color correction supports different project color spaces in DaVinci Resolve by automatically converting exported frames to sRGB before analysis. This conversion ensures the multimodal LLM accurately interprets the visual data regardless of your original project's color space, allowing consistent exposure correction, white balance adjustment, and contrast enhancement across various workflows.

The best way to automate exposure correction and saturation control is using an AI-powered assistant that directly analyzes video frames and applies precise CDL adjustments to your grading nodes. By integrating with DaVinci Resolve scripting, it reads the current grade state and modifies CDL values automatically, streamlining professional color correction tasks through simple natural language commands.

Multimodal LLM interpretation handles color space conversion by transforming exported video frames into the sRGB color space before visual analysis. This ensures the AI accurately evaluates the frame's exposure, contrast, and white balance, enabling it to generate precise CDL adjustments that are then applied back into the original project's color space within DaVinci Resolve.

A limitation of using AI for CDL adjustments is that it relies on exporting and analyzing individual frames, meaning real-time playback grading is not supported. Additionally, the AI focuses specifically on CDL parameters—exposure, white balance, contrast, and saturation—so complex secondary color corrections or custom node trees may require manual intervention beyond the automated natural language commands.