Depth Map Manipulation: Relighting 2D Footage via AI

Creating Information From Nothing

The miracle of AI is its ability to infer 3D structure from 2D data. By understanding the Z-axis of a video, we can apply lighting effects that were previously reserved for high-end 3D renders. This allows you to "re-shoot" the lighting of a scene long after the camera has stopped rolling.

Key Features

• Z-Space Estimation: The AI analyzes your flat 2D video and creates a grayscale depth map, understanding what is far and what is near.
• Virtual 3D Lights: Place a virtual "point light" or "spotlight" into your 2D scene, and watch it illuminate the subjects while casting accurate shadows.
• Volumetric Atmosphere: Use the depth data to add fog or mist that only exists *behind* your characters.

How It Works

1. Extract Depth: Run the "Depth Engine" on your imported clip.
2. Stage the Illumination: Drag virtual light sources around the subjects on the canvas.
3. Bake the Lighting: Render the final relit clip with perfect photographic integration.