Z-Depth Mapping

The Controls menu of the Z-Depth map offers options to control the effect of the map on the surface it is attached to.

• To create occlusion in the scene as if the content of the image was a real 3D object, enable Displace Vertices. You can further refine the effect:

  • The default mode without tessellation. It offers the best performance while still allowing for occlusion.

  • A 3D mode where the Surface is extruded in 3D space, with all the benefits of having a real 3D object in the scene in regard to lighting, but at the cost of having to increase the Surface resolution to get the best rendering results. An option to also bump the normals.

    Tip: Using Shift drag, you can now Z-Snap an Axis to a Z position defined by a Z-Depth texture map.

• To support both Z-Depth coming from another Action node or external renderer, you can specify if the input is normalized or absolute. Action renders its depth in a normalized range bringing back the near and far of the camera values between 0 and 1. Most other software renderers express the depth in units from the camera point of view, creating huge value range requiring the use of the Absolute setting.

• By enabling Z-Depth Pass option the Z-Depth texture map provides the Action Output Z-Depth render pass the Surface Z-Depth information: you are no longer limited to having only a single depth for the whole Surface.

You can provide your Z-Depth map, or use the Face or Global Analysis options to create a Z-Depth map.

To add a z-depth map:

1. In the schematic, select the surface or 3D geometry to which you want to apply the z-depth.

2. Click Media.

3. In the Media menu, select the media you want to use for the z-depth.

4. Do one of the following:

   • Drag the Z-Depth Map node from the node bin and place it in the schematic.
   • Drag the Z-Depth Map node from the node bin and place it where you want it in Result view.
   • Double-click the Z-Depth Map node. You do not need to be in Schematic view to add a node in this manner.

   The z-depth object is added to the schematic with its own parent axis. The new axis is the child of the selected surface or geometry. In Schematic view, the number in brackets next to the name of the Z-Depth node indicates the media used for the z-depth.

   To specify different media as the source, select the media in the Media menu, then click Apply, or use the Read File tab to manage the texture media (see Working With Textures in Map Nodes).

   Note: If you accessed Action as a Timeline FX, you are limited to one front/matte media, and therefore may not get the desired result. In this case, you can access Action from Batch or Batch FX, or from the Tools tab.

5. Double-click the Z-Depth node in the schematic, or follow the tab population rules for the Object menu (see Populating Menu Tabs of Selected Objects).

   The Z-Depth menu appears.

Using Face or Global Analysis to Generate Z-Depth Data

The Z-Depth map analysis leverages machine learning to generate Z-Depth data.

Tip: If you're working with a Selective node, use Selective object menu 3D AOV Create Map. Create Map automatically adds to the surface the map required by the Type of the AOV.

To create Z-Depth data for a human face:

1. In the Media menu, select the media with the human face to analyse.

2. In the schematic, right-click the surface with that same media, and select Add Map Z-Depth.

3. Double-click the Z-Depth map to display its menu.

4. In the Z-Depth map menu, set the Analysis Type box to Face Analysis.

   The analysis tries to detect a face and to extract Z-Depth data.

To create Z-Depth data for a whole image:

1. In the Media menu, select the media to analyse.

2. In the schematic, right-click the surface with that same media, and select Add Map Z-Depth.

3. Double-click the Z-Depth map to display its menu.

4. In the Z-Depth map menu, set the Analysis Type box to Global Analysis.

   The analysis the image and extracts Z-Depth data.

Tip: Use Analysis Region of Interest to refine the region to analyse and exclude areas from the analysis.

Actual results will vary according to the image content. As a rule of thumb,

Face analysis works best if (in order of importance, from most to least):

• You can see the eyes, nose, and mouth of the talent .
• The talent is facing the camera.
• The shot is a close-up shot.

Global analysis works best if you have in the shot:

• Clear foreground objects
• A well-defined background

Things to avoid:

• Motion blur
• Depth of field
• Strong specular highlights such as headlights in a night scene, or a bright sun in an outdoor scene