Trying to figure this out - imported the dice model into Reality Composer Pro, and of the three options, neither will work with this particular model. The next obvious question is, how do we assign a custom mesh (ideally, the original mesh, or a low-res equivalent) with collision component to the mesh with dynamics?

Here's a quick Maya dynamics test showing what I'm trying to achieve with either Reality Composer, Reality Composer Pro, or both. :) https://youtu.be/l-T3DkCMRD8

I've been wanting to learn the same thing. I was hoping that we'd have access to the original Reality Composer file so we could study and learn from it. Unfortunately, there's no way to open a .reality file other than Quick Look to try and reverse engineer it. SIGH It would be super useful and highly educational if the original author of those three animation samples would publish a detailed tutorial on how they were created. 🤞

Four months later and no actual response to this simple query? This should be an easy tutorial for someone to crank out, no? Specifically, I'm looking to learn how and where to author the button, line and callouts, then trigger the animations. Is this done in SceneKit, ARKit, Reality Composer Pro, or a combination? I like how this is all embedded in a portable USDZ file, which looks great in Vision OS. Surely, there is someone who knows how, or actually worked on these demo files that could provide the proper guidance? Thx!

Thx for the feedback! I should mention that I actually talked with one of my contacts at Substance 3d specifically about this. I am able to use multiple UDIMs, but keep as a single texture set. This way, I can paint over the seams of 14 hand-defined UV chunks, each with 4k baseColor maps with baked lighting. This enables me to re-project higher, 8K images from the drone and handheld Nikon footage, to make the ground texture that more detailed. I added to my previous RCP navigation->industry standard feedback. :) Fingers crossed!

Thank you for the quick reply! Regarding the UDIM workflow, do you have any suggestions for an alternative approach? I was told that the developers mentioned it would be no problem with multiple 4K maps, which these large environments need to ensure crisp textures (at least up close and within a 50-100 feet radius). I will look into the first 2 options presented immediately. Option 3 is good to be aware of, but I'd need to work with a programmer familiar enough to execute it. Side note while I have your attention - is there any way in RCP to combine camera Orbit, Dolly and Pan into a single command assessable via Option+Right/Middle/Left mouse click, exactly similar to how Maya works? This would be a huge improvement for quick navigation. :)