This is a somewhat complex setup that could be achieved other ways, but there is a reason for the method I am trying to use. I have a video that I am using as a motion background to a graphic. I put the graphic on the video's multi view layer. The graphic is a composite of other inputs with multi-view layers. One of the inputs is another video source that is cropped and zoomed to fit the graphic. This works in this setup. One of the other overlays, however is another multi-view composite. It is another video that has been zoomed and cropped on a separate input then that input is used in the overlay. The video will not display when I do this. If, however, I merely overlay the main graphic on the background video, everything works. Is this just a matter of multi-view layers not being able to look back 3 deep?