Motion Blur

mental ray can compute motion blur of any animated shading effect, any movement of scene elements as well as shape deformation. This includes motion blur of highlights and textured materials as primary effects, and secondary ray tracing features like shadows, reflections, refractions, and transparency. mental ray provides different motion blur algorithms to chose from, with different quality and performance impact on the final rendering.

There are two methods of defining motion blur:

Both instance motion and motion vectors can be combined.

A shutter speed and delay may be given for the render camera in the scene options or on the command line. A shutter speed of zero (the default) will turn motion blur rendering off, regardless of the presence of motion transformations or motion vectors in the scene. The shutter opens at time delay and closes after the shutter speed time has elapsed. An object, light, or camera moves by the distance defined by its motion transformation plus motion vector if the shutter interval is set to the full motion interval of 0.0…1.0. The motion blur trail will be rendered shorter for shutter ranges that are smaller than the motion interval, for example 0.2…0.8, and longer for shutter ranges exceeding that interval, for example 0.0…2.0. Setting the shutter open and close times to the same value allows to perform bidirectional motion blur in output shaders.

mental ray supports motion blur rendering with both the rasterizer/scanline and raytracing modes. Depending on the effects used in a rendering and the desired amount and quality of blurring this gives the choice of fast motion blur rendering without ray tracing, or very accurate but more expensive motion blur of ray tracing results like a moving reflection. Both modes can be used together to achieve a compromise between quality and performance. Motion blur does introduce rendering overhead, but normally only where blur must be computed. For example, if the camera moves then motion blur would need to be computed for the entire image and many if not all objects in the scene, compared to the movement of a single object where the overhead may be much smaller.

Rendering motion blur will soften the details of objects. mental ray can take advantage of this characteristic by automatically reducing the quality of view-dependent displacement for moving objects according to the length of the motion path in screen space within the current shutter interval. This currently affects any length approximation criteria for displacement of polygonal or triangle geometry. The decrease of displacement detail has a big impact on rendering performance and memory consumption by cutting down the typically huge count of tessellation data noticeably. A motion blur rendering of a heavy displacement scene may even be tuned to render faster compared to the static case. The amount of quality reduction in relation to the motion blur length is controlled by the geom displace motion factor scene option. mental ray enables this feature by default with a reasonable compromise between quality and speed, but it can be turned off for backwards compatibility.

Deprecated mental ray supports an alternative motion blur algorithm. While the regular motion blur algorithm is based on temporal oversampling, and is enabled if the shutter interval is nonzero (i.e. the shutter time is greater than the shutter delay time), the fast motion blur mode is based on coordinated spatial and temporal oversampling. This requires much less samples (typically by a factor of five) for motion blurred pixels. Fast motion blur mode is enabled if the shutter period is nonzero and the time contrast is set to 0. In fast motion blur mode, min samples is forced to be equal to max samples.
Note, that the rasterizer ignores these settings.

Copyright © 1986, 2013 NVIDIA Corporation