Is there a minimum height width for a collider on a 3d model?

imagineholos

I was working on a mirror project and noticed that when I created a collider sometimes I would not be able to select to the object via air tap. I changed the collider and made it a bit larger and it seemed to work. Just wondering if there is a recommended width / height for a collider.

AlexD

I don't think there is a recommended height value per se...

But, since the air-tap gesture acts on the gaze focused object let's look at how gaze works.
Gaze is essentially a raycast from your head position in the direction you are looking controlled by your head movement. So the farther the object, the more a slight movement of your head will cause a bigger move in your gaze raycast at the distance of the object

So doing a bit of a crude math calculations, if you objects are 2m away, a slight 5 degree up-down movement of your heard (which is very small) will cause your gaze to move 17cm up-down. If you collider is much smaller than that you might be missing it or at the very least move your gaze off it easily.

AlexD

I don't think there is a recommended height value per se...

But, since the air-tap gesture acts on the gaze focused object let's look at how gaze works.
Gaze is essentially a raycast from your head position in the direction you are looking controlled by your head movement. So the farther the object, the more a slight movement of your head will cause a bigger move in your gaze raycast at the distance of the object

So doing a bit of a crude math calculations, if you objects are 2m away, a slight 5 degree up-down movement of your heard (which is very small) will cause your gaze to move 17cm up-down. If you collider is much smaller than that you might be missing it or at the very least move your gaze off it easily.

HoloSheep

@imagineholos you can think of the box collider as an invisible field around your object that is used to sense hit detection by the cursor, gestures like the AirTap and potentially other purposes.

Besides making sure that the box collider is slightly larger than the object that you want to detect such things for, the other important consideration for actions like the TapToPlace action is that the origin of your object should also be inside that box collider (or potentially a secondary box collider at the same level). The origin of the object is used when placing the object and both the position of that origin and the detection of your second AirTap are based on your gaze, so it is therefore necessary to either include the origin in a box collider or to adjust the TapToPlace logic in some way.

Few more details and an image in this post.