If you’re using “Add Movement Input,” then that’s a function on pawn movement component that expects a vector in world space.
If you use the current camera, and get its actor rotation, that will give you the rotation in world space.
If you get the “right” and “forward” vectors of that camera rotation, they will be expressed in world space, too.
So, you can do something like:
On input right/left, get the right vector of the camera, set Z to 0 (to make it in the plane,) normalize it, and then multiply by input axis value. Use this for AddMovementInput.
On input forward/back, get the forward vector of the camera, set Z to 0, normalize it, and multiply by the input axis value. Use this for AddMovementInput.
1 Like