Compositing method using Z-channel to apply effects per depth layer — selective focus, volumetric effects, grading. Pixel-accurate spatial control without new renders.
You're working in the edit suite with a multi-channel EXR sequence and realize: the focus isn't quite right, the fog density should vary with depth, the color correction should look different depending on object distance. This is exactly where you turn to the Z-depth channel — a separate information layer that stores the distance to the camera for each pixel. With this depth map, you control effects pixel-by-pixel according to spatial position, without having to stack layers or draw masks.
Practical Workflow: Your 3D department renders a Depth Pass alongside RGB and Alpha — each shade of gray corresponds to a distance. In the compositor (Nuke, After Effects with plugin, Fusion), you use this Z-channel as a control channel for any node operations. Want selective defocus? The Depth channel determines the blur strength per pixel. Volumetric fog? Use Z to introduce dense layers, the further away, the denser. Grading with depth gradation? With a ramp or a keyer based on depth values, you control color and contrast adjustments — foreground cooler, background warmer, all automated instead of manually masked.
The decisive advantage: No masking work — and no hold errors at object edges. If a character stands in front of a door, their Z-values will differ from those of the door. The Depth channel captures this physically correctly. You can make adjustments later without retouching masks. This saves rendering passes and gives you the post-control you need, without back-and-forth with the VFX department.
Common Applications: Selective focus for digital racking, atmospherics (fog, god rays), depth-of-field enhancement, color correction by depth layer, stylized grading with depth gradation (e.g., foreground saturated, background desaturated). Also useful for tracking and reconstruction — if you have the Z-information, you can use it to control parallax effects or correct motion control errors.
Tip: Ensure the Z-pass is linear and within the correct value range — some render engines normalize values to 0-1, others use real camera units. For compositing, you need clarity about the scaling. And: Always activate a preview channel via checkbox to visually control the depth. What looks gray is often closer; white is further — but this can vary depending on the renderer.