The error: When building a RealityKitContent package that contains a USDZ file exported from Blender, Xcode throws the following error: error: [xrsimulator] Exception thrown during compile: Cannot get rkassets content for path .../RealityKitContent.rkassets because 'The file "RealityKitContent.rkassets" couldn't be opened because you don't have permission to view it.' error: Tool exited with code 1 The error message mentions "permission" — but permissions are not the issue. This appears to be a misleading error from realitytool masking a USD validation failure. What I've ruled out File permissions — all files are -rw-r--r--, user has Read & Write on the folder Extended attributes / quarantine flag — other files with the same @ flag work fine Corrupted archive — unzip -t confirms the USDZ is valid (board.usdc + textures) Stale build cache — deleted DerivedData and com.apple.DeveloperTools cache, no change Key observations The same file builds successfully on my colleague's machine running identical Xcode 26.3 - MacOS 26.3 Other USDZ files in the same .rkassets bundle (downloaded from Sketchfab, or created in Reality Composer Pro) compile without any issue. Only USDZ files exported directly from Blender are affected. When the file is placed in Bundle.main and loaded via Entity(named:in:.main), it works perfectly — no errors Reality Converter flags the file with two errors: UsdGeomPointInstancers not allowed, and root layer must be .usdc with no external dependencies The confusing part: the same file compiles fine on an identical Xcode 26.3 setup and importing method. This suggests either a machine-specific difference in Xcode's validation behavior, or a cached .reality bundle on my colleague's machine that isn't being recompiled. Current workaround: Loading from Bundle.main instead of the RealityKitContent package bypasses realitytool entirely and works, but loses Reality Composer Pro integration: if let entity = try? await Entity(named: "test", in: Bundle.main)

Hi Apple Developer Forums, I'm developing a visionOS video streaming app that uses a custom immersive cinema experience with RealityKit. I have a question about enabling video reflections in an immersive environment. My Current Implementation I'm using VideoMaterial with AVPlayer to display video on a ModelEntity plane in an immersive space: // Create screen mesh let screenMesh = MeshResource.generatePlane( width: VideoTheater.screenWidth, height: VideoTheater.screenHeight, cornerRadius: 0.0 ) let screenEntity = ModelEntity(mesh: screenMesh) // Apply VideoMaterial with AVPlayer screenEntity.model?.materials = [VideoMaterial(avPlayer: player)] The video renders correctly in the immersive space, but I don't see any video reflections on surrounding surfaces. Apple's Documentation Approach According to the documentation at https://developer.apple.com/documentation/visionos/enabling-vid eo-reflections-in-an-immersive-environment, the recommended approach uses: AVPlayerViewController for video playback dockingRegion modifier to specify where the video should appear The system automatically handles reflections My Question Is using AVPlayerViewController with dockingRegion the only way to get video reflections in an immersive environment? Or is it possible to enable reflections when using VideoMaterial directly with RealityKit's ModelEntity? My app requires a custom immersive cinema experience with: Custom screen positioning and scaling Danmaku (bullet comments) overlay Custom gesture controls HDR/Dolby Vision support Switching to AVPlayerViewController would require significant architectural changes, so I'd prefer to keep my current VideoMaterial approach if reflections can be enabled somehow. If VideoMaterial cannot produce reflections, are there any alternative approaches to achieve diffuse video reflections with a custom RealityKit setup? Environment visionOS 2.x RealityKit AVPlayer with custom resource loader (for DASH streams) Thank you for any guidance!

Hi, we have been working on an application that attaches a hand model to the users hands. Apple provides an animating hand models in visionOS project that is a useful starting point. https://developer.apple.com/documentation/visionOS/animating-hand-models-in-visionOS We have been trying to create our own hand model to attach but have had some issues with how it is attaching to the hand. For our hand model we want to include the forearm all the way up to the users elbow. I have attached a sample project of what our code currently looks like so you can run it. Just select show immersive space to attach the models. The left hand model is the space glove that we were trying to mirror. The right hand model is our model that we have been using. I have mapped each of the joints to the pertaining joint name on our model. The first issue we are having seems to be based around the placement of the forearm. It attaches itself at the wrist. The second issue seems to be around rotation. Our team is looking for some guidance on what needs to change in order to map this model correctly. Thanks in advance!

Hello, I am experiencing an issue where lighting is disabled within a Portal Component specifically when using Progressive or Full Immersion styles in visionOS 26.2. [Steps to Reproduce] Create a scene in Reality Composer Pro with custom lighting (Directional Light, IBL, etc.) and load it as an Entity. Add a PortalComponent to this Entity. Observe the Entity in an ImmersiveSpace under different Immersion Styles. [Observed Behavior] Mixed Immersion: Lighting works as expected.Progressive / Full Immersion: Lighting is disabled, causing the content to render incorrectly. [Additional Context] Without PortalComponent: The same Reality Composer Pro scene renders lighting correctly across all Immersion Styles (Mixed, Progressive, and Full).Regression: In applications built with Xcode 16.4 / visionOS 2.5, lighting inside the Portal functioned correctly. This issue appears to have emerged with Xcode 26.2 and visionOS 26.2. [Questions] Is the disabling of lighting inside Portals during Full/Progressive Immersion an intended architectural change or optimization in visionOS 26.2? Are there any workarounds, such as specific PortalComponent configurations or new API flags, to re-enable lighting in these immersion modes? Thank you.

Is there a way to stop a Reality Composer Timeline ? And restart it later on. For me it looks like you can only start a timeline via notification. and related to it. i have the issue that if the notification to start a certain timeline happens twice or more, it looks as if the timeline actually plays multiple times and animations start to glitch and jump. what is best practise here to avoid this.

the timeline editor feels often unfinished. Setting the time cursor to a different time is often not reflected in the preview. You either have to click a clip or wait. sometimes the cursor even disappears, eg. when switching tabs to shadergraph Not being able to select and move multiple clips is missing. There is also no snapping to clips or time cursor as found in other tools. And then there is the timeline compile bug https://developer.apple.com/forums/thread/810868 The timeline as it is, is a good start but it definitely needs some more love to be on par with other commercial tools like Unity or After Effects.

Hi! I have been struggling with this for a little while and most of what I've found has not helped much. I hope to find more success here. Essentially, I have a model I've made in Blender. I rigged it using Auto Rig Pro, and I've also used ARP to add a Mixamo animation to it. That all works fine in Blender. However, when I try to import this model into RCP, I don't get the animation. The "default subtree animation" is completely empty. I attribute this to my lack of experience in this field but here's what I've attempted thus far: Pushing the Mixamo keyframes into an NLA strip. I'm pretty sure this is the correct line of action, but I'm definitely not doing something right. Baking the animation (?) I've made sure that I have animation checked when I export the model! Any ideas or reference projects would be lovely. I haven't really found much that has pushed me in the right direction. This project is, unfortunately, kind of time sensitive, so I would appreciate help ASAP. Thank you and let me know if I can add anymore context!

I am creating an Augmented Reality iOS (Not VisionOS) app using scenes created in Reality Composer Pro. I'd like my code to send a notification to a RCP scene that plays a timeline. The RCP interface has the option to set up a behaviour for this purpose: This Forum thread https://developer.apple.com/forums/thread/756978 suggests the code I need for sending a notification is: name: NSNotification.Name("RealityKit.NotificationTrigger"), object: nil, userInfo: [ "RealityKit.NotificationTrigger.Scene": scene, "RealityKit.NotificationTrigger.Identifier": "HideCharacter" ] ) but the 'scene' var needs to point to the relevant RCP scene, which is loaded within a UIViewRepresentable ARView (because even in iOS26 it seems RealityKit/RealityViews aren't quite ready for AR use) and I can't work out how to correctly access it. Examples in the link above are for working with RealityKit and VisionOS only. Code for loading the scene is as follows. How can I get the notification code above to be situated in a separate SwiftUI View and send the notification to the RCP scene? typealias UIViewType = ARView func makeUIView(context: Context) -> ARView { // Create an ARView let arView = ARView(frame: .zero) // Configure it let arConfiguration = ARWorldTrackingConfiguration() arConfiguration.planeDetection = [.horizontal] arView.session.run(arConfiguration) // Load in Reality Composer Pro scene let scene = try! Entity.load(named:"myScene)", in: realityKitContentBundle) // Create a horizontal plane anchor let anchor = AnchorEntity(.plane(.horizontal, classification: .any, minimumBounds: SIMD2<Float>(0.2, 0.2))) // Append the scene to the anchor anchor.children.append(scene) // Append the anchor to the ARView arView.scene.anchors.append(anchor) return arView } func updateUIView(_ uiView: ARView, context: Context) { } }

randomly, the app does not work after small changes in Reality Composer. Small changes like scaling a object a tiny bit. to fix the error, i have to change another element in reality composer and hope for the best. if this does not help, i change (transform) something else, or deactive/activate something to get the project working again. I can't see a pattern why the Reality Composer Project sometimes gets in a state where it does not compile anymore.

I have an arguably massive project and am not sure if the issue is with the assets or my approach in the code. the error says : Tool terminated due to error "SIGNAL 6:Abort trap:6" Basically I have around 15-20 assets (usda files built out of usdz files). In the code i am loading a scene with all the usda files and then have the functions to enable and disable a particular asset when needed. This was working as intended when i am using dummy assets(with less polygons, lesser textures) But when i placed the actual assets the error appears and persists. Do I have a bad approach of loading all the scenes at once? Previously i have used an approach which loads the scenes when needed and that involved some lag before rendering the assets. But my current approach(when using dummies) works like a dime rendering and hiding the assets in realtime with no lag. Kindly suggest any workarounds.

I exported some usd assets from IsaacSim but they are not showing up correctly on my Apple Vision Pro. Even though the mesh looks to be the correct color in Finder and I can see the Diffuse Color looks correct, the object is still just gray. It should be green!

I'm developing a custom gesture-based visionOS project that uses hand tracking with collision detection spheres on fingers to register user interactions through collision components. I'm experiencing a critical occlusion issue where collision detection spheres are intermittently occluded by the background/depth buffer, causing fingers to pass through the 3D model entities without registering interactions. Detailed Description: I have added 3D entities in an immersive scene with collision spheres attached to fingers for detecting user interactions. Each sphere has: CollisionComponent with sphere shape Proper collision masks and groups configured Real-time position updates from hand joint transforms Each entity has: InputTarget components to register collisions The Issue: When users move their fingers to the entity to interact, some collision spheres (particularly on the pinkie and ring fingers) become occluded and pass directly through the 3D model without triggering collision events. Meanwhile, other fingers (like the index finger) continue to work correctly. This appears to be a depth perception/z-buffer issue between the model entity and the hand tracking collision spheres Questions: Is there a recommended approach for maintaining consistent depth ordering between hand-tracking entities and 3D models in immersive spaces to prevent occlusion issues? Should I be using AnchorEntities to anchor the entity to a plane or world position to establish a more stable depth reference? Are there specific RenderingComponent or material settings that could help ensure collision entities maintain their depth priority and don't get occluded? Could this be related to z-fighting when collision spheres and entity geometry occupy similar depth ranges? If so, what's the recommended depth bias approach? Is there a better architectural approach for implementing interactions with custom hand gesture tracking that avoids these depth perception issues? What Would Help: Implementation guidance for ensuring reliable collision detection between hand-tracked entities through custom gestures and 3D models. Best practices for depth management in immersive spaces with custom hand gesture tracking. Sample code demonstrating stable hand-to-object interaction patterns. Information about whether this is a known limitation or if there are specific APIs I should be leveraging This issue is significantly impacting the reliability of our app experience, as users cannot consistently interact with all model components. Any guidance from Apple engineers or developers who have solved similar depth/occlusion challenges would be greatly appreciated. Additional Context: This is for a productivity-focused application where accuracy and reliability are critical. Thank you for any assistance!

How can I create 180-degree apple immersive videos using game engine

Hello! Back from last week's amazing visit to Cupertino for the Game Dev session and diving back into Vision Pro experimentation. I've exported a simple geometry nodes with animation test from Blender for use in RCP, with intended output to Vision Pro. I've attached a few screenshots showing the node setup and how it animates over time. I select the Cube mesh and export as .usdc with animation. In the finder via quick look, I can actually see it working! If I try exporting as .usdz, however, i'm not seeing any animation in the finder preview. Next, I import the .usdc file to RCP and add an Animation Library component to the cube mesh, but am not seeing any animation selectable, even though I see animation playing back in preview. Next, I import the .usdc into Maya (via proper USD Stage pipeline - i'm learning to be USD compliant for authoring!) to verify if the animation is working, and it does. What step(s) am I missing to get this working in Reality Composer Pro? My goal is to experiment with animating these geometry node instances - along with color animation if possible - over to Vision Pro for full scale, immersive presentation. Of particular note, I am not a programmer, so I am trying my best to brute force this the only way I currently know possible, by keyframe animation and importing through Reality Composer Pro. I realize that, ideally, I should be learning how to leverage the code portion so I can start programatically controlling my 3d entities (with animation), but need more hand holding and real-world examples to help me get there. Thx!

I’m currently developing a visionOS app that includes an RCP scene with a large USDZ file (around 2GB). Each time I make adjustments to the CG model in Blender, I export it as USDZ again, place it in the RCP scene, and then build the app using Xcode. However, because the USDZ file is quite large, the build process takes a long time, significantly slowing down my development speed. For example, I’d like to know if there are any effective ways to: Improve overall build performance Reduce the time between updating the USDZ file and completing the build Any advice or best practices for optimizing this workflow would be greatly appreciated. Best regards, Sadao

In Reality Composer Pro, why is the Sky Sphere so much larger than the Sky Dome? By my estimate, the Sky Sphere has a radius of 100m, while the Sky only has a radius of only 12m.

I have a question about Apple’s preinstalled visionOS app “Encounter Dinosaurs.” In this app, the dinosaurs are displayed over the real-world background, but the PhysicallyBasedMaterial (PBM) in RealityKit doesn’t appear to respond to the actual brightness of the environment. Even when I change the lighting in the room, the dinosaurs’ brightness and shading remain almost the same. If this behavior is intentional — for example, if the app disables real-world lighting influence or uses a fixed lighting setup — could someone explain how and why it’s implemented that way?

After writing the code, when debugging on VisionPro, the program will encounter a blocking situation when running from Xcode to VisionPro. It will take a long time for the execution information to appear on the Xcode console

Goal: To render in an apple vision pro app, the solid-mechanics 3D simulation results coming form an FEA code. Starting point: I have surface vtks with deformations on each node. Each time step has a a mesh with the nodal coordinates. This is straighforward translatable to a usd MeshSequence. Unfortunately, the results cannot be simplified to a scaling o linear transformation as you would do with other game-oriented animations. Tools: Right now, I am using Xcode and reality composer pro (RCP) to build the scenes. Technical limitations: I am aware that RCP can do animations with BlendMesh and skeletons and that MeshSequence is not a problem. Progress: Coverting to the sequence of vtk meshes to a usd MeshSequence is straighforward. This animates correctly in Preview and Blender (see screenshot). I managed to convert from MeshSequence to multiple keys and BlendMesh. This also animates correctly in Blender and preview. Unfortunately, the BlendMesh of multiple blended meshes shows a zero animation time in RCP (see screenshot below) Also, see below usda file scheme for the animation. Of course I am not showing full vectors such as faceVertexCounts, faceVertexIndex, normals. Question: what is the right set up to create a BlendMesh animation that RCP will correctly import and animate, form a set of Meshes or multiple key shapes? Blender animation Time zero RCP "animations" #usda 1.0 ( defaultPrim = "BlendMeshRoot" doc = "Blender v4.5.3 LTS" endTimeCode = 48 framesPerSecond = 24 metersPerUnit = 1 startTimeCode = 0 timeCodesPerSecond = 24 upAxis = "Z" ) def Xform "BlendMeshRoot" ( customData = { dictionary Blender = { bool generated = 1 } } ) { def SkelRoot "Mesh" { custom string userProperties:blender:object_name = "Mesh" float3 xformOp:rotateXYZ = (89.99999, -0, 0) float3 xformOp:scale = (0.009999999, 0.01, 0.01) double3 xformOp:translate = (0, 0, 0) uniform token[] xformOpOrder = ["xformOp:translate", "xformOp:rotateXYZ", "xformOp:scale"] def Mesh "Mesh" ( active = true prepend apiSchemas = ["MaterialBindingAPI", "SkelBindingAPI"] ) { uniform bool doubleSided = 1 float3[] extent = [(25.091871, -34.121277, -13.298501), (299.94482, 245.10088, 202.35126)] int[] faceVertexCounts = [3, 3, ... int[] faceVertexIndices = [0, 10293, ... rel material:binding = </BlendMeshRoot/_materials/MeshSequence_Default> normal3f[] normals = [(-0.3632836, -0.9102419, -0.19870725), .... point3f[] points = [(244.41148, 155.42062, 70.454926),..... float3[] primvars:node_displacement = [(93.54703, 110.9341, 48.37992).... float3[] primvars:Normals = [(-0.0050530406, -0.9910114, -0.13368203),... int[] primvars:skel:jointIndices = [0, 0, 0, 0, 0 ... float[] primvars:skel:jointWeights = [1, 1, 1, 1, 1... uniform token[] skel:blendShapes = ["frame_0000", "frame_0001", "frame_0002", "frame_0003", "frame_0004", "frame_0005"] rel skel:blendShapeTargets = [ </BlendMeshRoot/Mesh/Mesh/frame_0000>, ....... </BlendMeshRoot/Mesh/Mesh/frame_0005>, ] prepend rel skel:skeleton = </BlendMeshRoot/Mesh/Skel> uniform token subdivisionScheme = "none" custom string userProperties:blender:data_name = "Mesh" custom float userProperties:originalTime float userProperties:originalTime.timeSamples = { 0: 0, } def BlendShape "frame_0000" { uniform vector3f[] offsets = [(0, 0, 0), (0, 0, 0),..... uniform int[] pointIndices = [0, 1, 2, ..... } ..... ..... #### BlendShape frame to 0005 ..... def Skeleton "Skel" ( prepend apiSchemas = ["SkelBindingAPI"] ) { uniform matrix4d[] bindTransforms = [( (1, 0, 0, 0), (0, 1, 0, 0), (0, 0, 1, 0), (0, 0, 0, 1) )] uniform token[] joints = ["joint1"] uniform matrix4d[] restTransforms = [( (1, 0, 0, 0), (0, 1, 0, 0), (0, 0, 1, 0), (0, 0, 0, 1) )] prepend rel skel:animationSource = </BlendMeshRoot/Mesh/Skel/Anim> def SkelAnimation "Anim" { uniform token[] blendShapes = ["frame_0000", "frame_0001", "frame_0002", "frame_0003", "frame_0004", "frame_0005"] float[] blendShapeWeights.timeSamples = { 0: [1, 0, 0, 0, 0, 0], 1: [0.9697085, 0.03029152, 0, 0, 0, 0], 2: [0.88787615, 0.11212383, 0, 0, 0, 0], ..... 46: [0, 0, 0, 0, 0.11212379, 0.8878762], 47: [0, 0, 0, 0, 0.030291557, 0.96970844], 48: [0, 0, 0, 0, 0, 1], } } } } def Scope "_materials" { def Material "MeshSequence_Default" { token outputs:surface.connect = </BlendMeshRoot/_materials/MeshSequence_Default/Principled_BSDF.outputs:surface> custom string userProperties:blender:data_name = "MeshSequence_Default" def Shader "Principled_BSDF" { uniform token info:id = "UsdPreviewSurface" float inputs:clearcoat = 0 float inputs:clearcoatRoughness = 0.03 color3f inputs:diffuseColor = (0.8, 0.4, 0.3) float inputs:ior = 1.5 float inputs:metallic = 0 float inputs:opacity = 1 float inputs:roughness = 0.5 float inputs:specular = 0.2 token outputs:surface } } } def Scope "AnimationClips" { custom rel animations = </BlendMeshRoot/Mesh/Skel/Anim> } def RealityKitComponent "AnimationLibrary" { custom rel animations = </BlendMeshRoot/Mesh/Skel/Anim> custom token info:id = "RealityKit.AnimationLibrary" custom double realitykit:approximateDuration = 2 custom double[] realitykit:clipDurations = [2] custom string[] realitykit:clipNames = ["Anim"] custom rel realitykit:clipTargets = </BlendMeshRoot/Mesh/Skel/Anim> custom double realitykit:frameRate = 24 custom bool realitykit:isAnimationLibrary = 1 } }

In courses like Compose interactive 3D content in Reality Composer Pro Realitykit Engineers recommended working with Reality Composer Pro to create RealityKit packages to embed in our Realitykit Xcode projects. And, comparing the workflow to Unity/Unreal, I can see the reasoning since it is nice to prepare scenes/materials/assets visually. Now when we also want to run a Xcode Cloud CI/CD pipeline this seems to come into conflict: When adding a basic *.usdz to the RealityKitContent.rkassets folder, every build we run on Xcode cloud fails with: Compile Reality Asset RealityKitContent.rkassets ❌realitytool requires Metal for this operation and it is not available in this build environment I have also found this related forum post here but it was specifically about compiling a *.skybox.