Consider a very wide angle head mounted display like the StarVR. Imagine it has two screens per eye, so total of four (like StarVR).
From the Panda3D side and from the experience of a software developer, which is more computationally intensive? Rendering to 4 separate screens and feeding to the head mounted display via DisplayPort MST (which supports connecting 4 screens via one cable), or feeding one wide screen and letting the ICs inside the HMD do the cropping before sending to the 4 screens?
If there is not much difference in rendering time, first option is more attractive as saves the head mounted display developer from using expensive video cropping ICs.
Asking because I’m working on a DIY headset and need t build a prototype board for it.