Today with something that may be a bug, maybe not. But either way, a normal line trace will properly return the impact normal as expected and demonstrated in the photo below:
But when I run a sphere trace. Mind that it is not a multitrace (which also returns the same results). As you can see, it almost uses a look-at rotation to determine the impact normal. Rather than an actual surface normal as not expected but demonstrated in the photo below.
Alternative solutions to the issue are appreciated, in case of a needed context, this is for a custom camera system. All blueprints used are in fact consistent between the use of line trace and sphere trace. Which means that the error is not to do with the code.
It does not appear to return not the surface normal but returns a normal perpendicular to the inner sphere that is closest to the impact point.
You seem to be correct. This is odd as I am using the default primitive shapes and the normals appear correct. I’m not going to mark this as a solution quite yet as I can’t be certain if this is luck of the draw or if it’s a fix all solution. I’ll keep you posted.
Greetings @silnarm!
So I am yet to find an issue with this after extensive testing. For now I will mark you post as a solution. Are you alright though if I personally message you for assistance if this issue returns as though you are evidentially knowledgeable on the subject?
