The Camera

A camera describes a view of the scene, and all the properties assigned to a camera, like film or image size and resolution. In mental ray this also includes options for post-processing in output shaders and storing images in files on disk, as well as lens effects performed by lens shaders, global atmosphere or fog simulation in volume shaders, global environment shaders to control what happens to rays that leave the scene, and other parameters. A camera will be positioned and oriented with an instance element, in the same way that objects are placed anywhere in world space of the scene.

By default, a pin-hole perspective camera is used for which the focal length, aperture and aspect ratio may be specified on either the camera element of the scene description, or overridden on the command line of mental ray. Optionally, visual effects such as depth of field can be achieved by specifying additional shaders.

A scene may specify any number of cameras. However, only one camera is rendered at a time, called the render camera. Cameras may be used for purposes other than primary rendering, like to define a view projection on a light for shadowmap creation, or for references in custom shaders.

Stereoscopic

mental ray supports to render stereoscopic views in a single run and with optimized performance. If enabled then the two images for the left and right eye are computed simultaneously, and will be stored automatically in separate frame buffers and written to separate output image files. In this mode, it is possible to compute certain expensive view-dependent effects like displacement or final gathering just once, and share the results in the image rendering for the eyes. This allows to take performance advantage of the view coherence, and avoids diverging results in the two images like popping of displacement details due to slightly different angle of view.

The stereoscopic rendering is activated by adding new parameters to a perspective render camera, or on the command line of a standalone mental ray. Existing shaders can be used without changes to render stereo. See also Known Limitations.