Video Processing Pipeline

Navigation:  Pandoras Box > Devices and Layer Parameters >

Video Processing Pipeline

prev main next

Navigation:  Pandoras Box > Devices and Layer Parameters >

Video Processing Pipeline

prev main next

Please wait. Due to its content this page might take a while to load...

This chapter explains the fundamental video processing pipeline in Pandoras Box. It is separated in two render passes
- the (layer) composition pass, most important for the creation process
- the output pass, most important for the technical setup

pb-video-processing-pipeline_composition-pass
Step 1, the Creation Process:

By default, all Video, Pointer, Light and Notch Layers are part of the Composition Pass. Here, you do your layer composition - setting the media, the video playback, position, rotation and scale parameters as well as the visual effects. The render history is determined by the Z Position respectively, if it is not set, by the layer arrangement whereas the topmost layer in the Device Tree tab is rendered first and thus overlaid by other layers. (For even more options see Camera Inspector > "Determine layer render order")

The end of the Composition Pass is determined by adjusting the camera's perspective (see Camera Layer) onto your 3D composition. Then the "Output to Texture" process generates a so called render target. There is a render target for each camera in the Device Tree.
As said above, per default, all mentioned Layers are seen by all cameras. However, the Layer Inspector > "Render Layer in Pass" settings allow to adjust this. You can for example exclude a Layer from specific Cameras, or send it to an Output directly.

The final texture, the render target, can be forwarded in various ways. In most cases it is simply processed by the next Output pass. This is the case when the Camera Layer is linked to an Output Layer and its render state says "Render".

Further it is forwarded:
- to the graphics card's output, thus your connected display, in case the output's render state is set to bypass
- to the matrix patch
- to be rendered for the video export

pb-video-processing-pipeline_global-cam

Camera setup

pb-video-processing-pipeline_render-target-1

Render Target 1

pb-video-processing-pipeline_render-target-2

Render Target 2

pb-video-processing-pipeline_output-pass
Step 2, the Technical Setup:

Here, you match the texture to your technical setup.
By default, the source media is simply the render target from the assigned Camera Layer.

In the output pass, the controls of the Output Layer allow to set up:
- a warping object if the texture is projected on a bended screen,
- a matrix patch,
- softedge or keystone adjustments,
- a color correction on the whole output (via effects)
- additional perspective settings.

As said above, the Layer Inspector offers Render settings which allow to send a Layer directly to the Output Pass. In the left image, such a Layer is marked with a star (*).  In that case, a layer is not part of the Render Target and not influenced by warp, softedge, keystone and FX settings. Only the perspective settings (named "Output View Controls" in the left image) apply. In short, the feature allows to place Layers outside the warp or softedge area

At the end of the output pass, the "Output to Display" process forwards a texture to the so called back buffer respectively to the graphics card and thus your connected display.
For the above depicted example to possible outputs could be:

pb-video-processing-pipeline_ouput-1

for output 1, a warp object has been applied and softedge settings were used

pb-video-processing-pipeline_ouput-2

for output 2, keystone and softedge settings were used