I also feel it’s easier to have misalignment 3d refinement in relion5 than relion4 when dealing with particles restrained by other densities like membrane attached proteins. Although I have not tried 3d particles yet