3D issues in Motion

Despite them being in the library, I would not call them "Motion 3D objects".

They are USDZ files, and are in fact treated like any other such USDZ file (like ones you can download from the web or build with Reality Composer).

As I noted before, USDZ files live in a separate realm inside Motion, and do not properly interact with the rest.

I wish Motion would be upgraded to a full 3D software, but that is unlikely to happen.

As Luis pointed out. Motion can render USDZ (which are 3D object), but they are rendered in layer order with non-USDZ objects. How rasterization affects 3D text and 3D objects in Motion - Apple Support. FWIW, I highly recommend reviewing the documentation section on Rasterization (Intro to rasterization in Motion - Apple Support )

In a nutshell, things like 3D Text and 3D Objects (USDZ) will render in depth order when next to like objects in the layer list, but will render in layer order with all other objects in the scene. This is why when you rotate your view below the ground plane of the DZ the 3D Objects still render in front. There are compositing techniques you can use to work around this (masking, trimming layer length, pre-composing).

For your example of the light making layers look transparent when put in the middle, this seems right looking at your layer list. You have masks on your images, so these are likely adding transparency already. The light may just be accentuating the transparency.

Finally, for the particles. For performance the particles have a parameter (Render Particles) that defaults to "In Local", this will make them render in their own space ("hello again layer order"). If you change this to "In Global 3D" they will render in depth order to other objects in the scene (unless layer ordering is triggered by rasterization). Also, since Motion does not have a physics system the particles will pass through other layers, don't expect them to collide and bounce off everything in your scene.

Hope this helps

Actually, you can get particles to stay contained in a "box" shape using Behaviors > Simulations > Edge Collision.

The emitter should have the Edge Collision and oriented inside the Edge Collision box; the "box" should be aligned in default orientation (0,0,0) and the angle done by camera. The size of the 3D Text "character" should be sized to fit the box (if the font is prepared well enough, then the actual Size and Depth parameters of the text will closely match the dimensions of the Edge Collision.)

This can work in 3D Text... I don't think you can do this with 3D Objects. (3D Objects have been a rather sore disappointment to me.)

You can, however, cheat a little with 3D Objects using 3D Text for something like this:

How this is pulled off is to create a group with the 3D object ("cube.usdz") - I added a rate behavior to Rotation Y. In another group, create a 3D Text object with a Square shape (just about any shape will do) and increase its size (I used ~780). Set the Opacities of the sides and back to 0. For the Front, set the surface to Generic > Image.

Create a Clone Layer of the 3D object's group (which is set to 2D). The Clone Layer Group should be 2D as well. Turn off both groups.

Add the Clone Layer to the 3D Text > Front > Generic Image and set Wrap Mode to None, open up the Placement options, open the Scale options and turn off Scale With Font Size. Rescale the object to fit into the view.

Basically, you have to build a "projection screen" (texture) for the 3D Object on the 3D Text Surface. Now, other 3D Text objects (and other regular objects in 3D space) will behave as expected (but not anything else).

IMO - it's not generally worth the trouble.

If I understand correctly, you're right...

These look like USDZ models, like those you could build in Reality Composer.

Motion lets us import USDZ files, but they do live separately from other Motion objects, unfortunately.