I have a video I am overlaying animations on for illustrations. I have layers which have different instrumentation, and depending on the moment I want to overlay different instruments in different locations via an alpha overlay. The problem I want to solve is that the camera seems tied to the video output rendering.
I want to instead make some very odd cameras with very unusual aspect ratios which will be added or removed at various key frames. Do I need to make every single instrument in a separate blender file, render every single instrument as a custom video resolution, and then overlay them all?? All I really need is to make different cameras which have no dependency on the actual output resolution...e.g., something like a virtual format. I don't know how to pick camera aspect ratio without changing the video rendering itself and I want to keep my video output settings unchanged.
If it helps I am using cycles renderer.
EDIT: I have a camera inside of a transparent sphere pointed outwards from center. I have different objects on the surface which are instrumentation markings...their materials will change depending on conditions. What is seen will depend on the camera orientation relative to the sphere (which uses quaternion rotations matching an aircraft)...and depending on what is seen I will want to reshape the camera width and height to specific angles. I will actually have multiple cameras in that same sphere pointed the same direction to place different views of the same model in different locations.
I have no images, I just want the camera such that I can manually select vertical and horizontal fields of view to a specific angle. This won't render to an external file, it will be an overlay on top of a video. Example: I want the camera to be 110 degree vertical field, and 15 degree horizontal field...this will occupy a location on top of a video which is perhaps 10 percent of the total video size. Imagine an aircraft with an artificial horizon plus a HUD view of the same instrument.