Viz Libero Render Sequence (When Integrated with Viz Engine)

1. Back Layer: Not rendered, even if a Scene is present.

2. Main Layer:

   1. Background image (live video) is rendered.

   2. Containers with the Key plug-in, with Combine with BG chroma key set to On and Render Mode set to Add, are keyed with the background video. The final key of these items results from the key information of the background video, provided by Viz Libero, and the key information of the Container.

   3. Containers with standard keying are rendered with optional lens distortion and linearly keyed over.

3. Front Layer: Only containers with a linear key are rendered on top, without lens distortion.

Viz Arena Render Sequence (Mechanical Based Tracking)

1. Back Layer: The keying information for Containers in the Main Scene is created. Typically objects with the Key plug-in on them are used for this:

   1. Background image without alpha and without lens distortion.

   2. Containers which contain standard key information either as linear key or generated through chroma keying.

2. Main Layer:

   1. Containers with the Key plug-in, with Depth information only set to On, are rendered. Only the depth information is rendered and used to cutout "real" objects from chroma keyed virtual objects.

   2. Containers with the Key plug-in, with Combine with BG chroma key set to On , are rendered. Depth and chroma keying information are used to superimpose these Containers.

   3. Containers with the Key plug-in, with Depth information only set to On, are rendered. The depth information is used for Containers which should be keyed linearly.

   4. Containers with a linear key are rendered with lens distortion and linearly keyed over.

3. Front Layer:

   1. Background image is rendered. This only makes sense if the last visible layer for SDI output is set to Main Layer. If this is so, a different video source can be shown on the Viz Engine preview to prepare the next item(s), which goes On Air.

   2. Containers with a linear key are rendered with lens distortion and keyed over.

Viz Libero/Viz Arena Render Sequence (Image Based Tracking)

1. Back Layer: Nothing is rendered from the Back Layer Scene.

2. Main Layer:

   1. Background images are rendered.

   2. If a foreground image (Stencil Mask) is present, but not active, only the alpha information is used to mask out parts of the containers in the Main Scene. Alpha is rendered without lens distortion.

   3. Containers with the Key plug-in, with Combine with BG chroma key set to On and Render Mode set to Add, are keyed with the background video. Chroma keying information is used to superimpose these Containers.

   4. Containers with a linear key are rendered with lens distortion and linearly keyed over.

3. Front Layer:

   1. Containers with a linear key are rendered.

Rendering Support for Dirty Video Feeds with Viz Arena

A "dirty" video input signal is a video containing graphical items such as logos, banners, clocks, etc., that are overlaid on top of the video signal before being sent to the Viz Engine. To avoid disturbances in the viewing experience, Viz Engine has to render its graphical items "below" the preexisting overlays. If supplied with a "dirty" video input signal, Viz Engine can suppress rendering of graphical items within certain image regions. To do this, alpha image masks need to be sent to Viz Engine, defining in which regions of the image graphical items must be rendered, and in which rendering shall be suppressed.

In other words, areas with existing graphics from the broadcast need to be excluded to avoid rendering new graphics on top of them. To achieve this, an alpha image needs to be loaded as a foreground image in the scene. The area with alpha information is then excluded from rendering, working as an inverse mask. Viz Engine then incorporates the currently activated image mask(s) for rendering of graphical items. In image regions where the value of the image mask is below 255, the alpha of all rendered graphics is adjusted accordingly. Applying the image mask also works when the render scale is different than 1. Independent of the currently selected render scale, the size of the image mask always refers to the original video resolution.