Rendering a 3D Scene

The 3D Viewer displays the scene using an OpenGL hardware render. When you build a scene, Nuke renders high-quality output from the perspective of the camera connected to the render node. The rendered 2D image is then passed along to the next node in the compositing tree, and you can use the result as an input to other nodes in the script.

Choosing a Render Node

Nuke ships with two render nodes, ScanlineRender and RayRender, which are suited to different tasks. ScanlineRender, as the name suggests, renders row-by-row advancing down the picture. RayRender traces a path from the camera, or virtual eye, to the light source pixel-by-pixel.

Note:  NukeX and Nuke Studio also include Pixar’s PhotoRealistic RenderMan®, another ray renderer. See PrmanRender for more information.

The ScanlineRender and RayRender nodes have the same inputs and share some of the same controls in their respective Properties panels, but they both have different strengths and weaknesses:

ScanlineRender generally produces faster results, but is less accurate with reflection and refraction.

ScanlineRender supports Deep workflows downstream by injecting a deep channel into the node tree. See Using ScanlineRender to Generate Deep Data.

RayRender generally produces very accurate reflection, but at the cost of processing time.

RayRender does not currently support Deep workflows or render sprite-based particles from NukeX's Particles system.

As a general rule, if you can afford to wait for a render and don't need Deep data or particles that rely on sprites, use RayRender.

Rendering Out a Scene

To render a 3D scene into 2D space you need three things: your scene, a render camera, and a render node. When set up correctly, the output of the render node is passed down the node tree for compositing as normal.

To render out a scene:

1.   Set up your 3D scene as normal, including geometry, materials, lights, etc.
2.   Add a render node and connect the scene to the obj/scn input.
3.   Make sure the render camera is connected to the render node's cam input. See Cameras for more information.
4.   Add a background to the bg input, if required. You can use the bg input to composite a background image into the scene and to determine the output resolution.

Note:  If the bg input is not used, the render node output defaults to the root.format or root.proxy_format defined in the Project Settings.

5.   Toggle the Viewer back to 2D.
6.   Connect the output of the render node to the appropriate 2D nodes in your script.

See ScanlineRender RayRender for more information on the different render node controls.

To Add Motion Blur to the 3D Scene

You can use the properties on the MultiSample tab to add motion blur to your 3D scene. For more information, see Adding Motion Blur to the 3D Scene.