Q100363: Matching camera framing in Maya and Katana
This article explains how to export cameras from Maya and import them into Katana to match the cameras in both applications, and clarifies an important difference in the treatment of camera attributes between Maya and Katana.
Maya to Katana camera workflow
When exporting cameras from Maya in Alembic format and importing them into a Katana project, users may notice that the framing of the render in Katana does not match the framing in Maya.
The main reason for this is the Overscan setting of the camera, which is treated differently in both applications.
To match the camera viewports and render results produced by Maya and Katana, change the camera's Overscan attribute in the Display Options to 1 before exporting an Alembic camera from Maya. Please read on for a more detailed explanation of why this is necessary.
Overscan in Maya vs Katana
Autodesk states in the Online Help for Maya:
Scales the size of the scene in the camera’s view only, not in the rendered image. Adjust the Overscan value to see more or less of the scene than will actually render. If you have view guides displayed, changing the Overscan value changes the amount of space surrounding the view guides, making them easier to see.
This is a different definition of overscan than what is used by other applications like Katana and Nuke, where overscan is a "buffer" of pixels in the render outside of the actual data window, so the rendered image is actually bigger than the render resolution.
Consider the following comparison:
|Maya Viewport with Overscan value of 2.0 set in Display Options||Maya Render - no overscan|
|Katana Viewer - same amount of space around the object in the viewport as in Maya, no resolution gate affecting the render||Katana render with 130 pixels overscan on each side, actual data window (render resolution) indicated by blue line.|
What Katana shows in the Viewer should be basically the same as what Maya shows in the camera Viewport. The difference is that Maya uses the Resolution Gate inside the Viewport to determine what will be rendered and uses the Overscan only to adjust what is shown in the camera Viewport. Katana simply renders everything inside the dotted line indicating the resolution in the Viewer and can add on a number of pixels (specified in RenderSettings) as an overscan around the rendered image.
When exporting cameras as Alembic caches from Maya, the overscan values still get written into the Alembic file and read into Katana, as you can see in the Attributes tab for the camera under the info.abcCamera group attribute.
The values for overscanLeft, overscanRight, overscanTop and overscanBottom will always be the Maya Overscan setting minus 1. The Resolution Gate however is not something that gets exported from Maya when writing out an Alembic cache.
To match the camera viewports and render results produced by Maya and Katana, make sure the Overscan value in the Maya camera attributes > Display Options is set to 1 before exporting an Alembic camera from Maya.
We're sorry to hear that!Please tell us why.