- Edited
trooowiii The scale can be off in VR and you won't really notice it as much, but in AR, the scaling needs to be on point. Proper distancing and position of the camera as well.
This is exactly why passthrough WONT get any better - at least not anytime soon. You can standardize a camera position, distance, lighting, and scaling - and you still have no idea in what environment the end-user will watch it. The model has to match endless user variables, or it isn't immersive. That's why it is easier to just shoot VR, since it replaces the whole field of vision. In that case, models need only be positioned, lit, and, and scaled so they match the existing room, not the infinite passthrough rooms. That isn't something that's going to get better until the headset can auto-scan your room for dimensions, position, and lighting and feed that information to the player which scales and adjusts the output in real time. That's way in the future, considering we still cant get them to stop cutting models heads off in cowgirl after almost 8 years of practice now. I'm always surprised by how confident people are that AR is going to get SO MUCH BETTER when studios haven't consistently mastered VR yet - which is much easier.