Hi everyone, I’m encountering a memory overflow issue in my visionOS app and I’d like to confirm if this is expected behavior or if I’m missing something in cleanup. App Context The app showcases apartments in real scale using AR. Apartments are heavy USDZ models (hundreds of thousands of triangles, high-resolution textures). Users can walk inside the apartments, and performance is good even close to hardware limits. Flow The app starts in a full immersive space (RealityView) for selecting the apartment. When an apartment is selected, a new ImmersiveSpace opens and the apartment scene loads. The scene includes multiple USDZ models, EnvironmentResources, and dynamic textures for skyboxes. When the user dismisses the experience, we attempt cleanup: Nulling out all entity references. Removing ModelComponents. Clearing cached textures and skyboxes. Forcing dictionaries/collections to empty. Despite this cleanup, memory usage remains very high. Problem After dismissing the ImmersiveSpace, memory does not return to baseline. Check the attached screenshot of the profiling made using Instruments: Initial state: ~30MB (main menu). After loading models sequentially: ~3.3GB. Skybox textures bring it near ~4GB. After dismissing the experience (at ~01:00 mark): memory only drops slightly (to ~2.66GB). When loading the second apartment, memory continues to increase until ~5GB, at which point the app crashes due to memory pressure. The issue is consistently visible under VM: IOSurface in Instruments. No leaks are detected. So it looks like RealityKit (or lower-level frameworks) keeps caching meshes and textures, and does not free them when RealityView is ended. But for my use case, these resources should be fully released once the ImmersiveSpace is dismissed, since new apartments will load entirely different models and textures. Cleanup Code Example Here’s a simplified version of the cleanup I’m doing: func clearAllRoomEntities() { for (entityName, entity) in entityFromMarker { entity.removeFromParent() if let modelEntity = entity as? ModelEntity { modelEntity.components.removeAll() modelEntity.children.forEach { $0.removeFromParent() } modelEntity.clearTexturesAndMaterials() } entityFromMarker[entityName] = nil removeSkyboxPortals(from: entityName) } entityFromMarker.removeAll() } extension ModelEntity { func clearTexturesAndMaterials() { guard var modelComponent = self.model else { return } for index in modelComponent.materials.indices { removeTextures(from: &modelComponent.materials[index]) } modelComponent.materials.removeAll() self.model = modelComponent self.model = nil } private func removeTextures(from material: inout any Material) { if var pbr = material as? PhysicallyBasedMaterial { pbr.baseColor.texture = nil pbr.emissiveColor.texture = nil pbr.metallic.texture = nil pbr.roughness.texture = nil pbr.normal.texture = nil pbr.ambientOcclusion.texture = nil pbr.clearcoat.texture = nil material = pbr } else if var simple = material as? SimpleMaterial { simple.color.texture = nil material = simple } } } Questions Is this expected RealityKit behavior (textures/meshes cached internally)? Is there a way to force RealityKit to release GPU resources tied to USDZ models when they’re no longer used? Should dismissing the ImmersiveSpace automatically free those IOSurfaces, or do I need to handle this differently? Any guidance, best practices, or confirmation would be hugely appreciated. Thanks in advance!

The issue reproducible with empty project. When you run it and tap "Open immersive space" it takes a couple of minutes to respond. The issue only reproducible on real device with debugger attached. Reproducible other developers too (not specific to my environment). Issue doesn't exists in Xcode 16. Afer initial long delay subsequent opens works fine. Console logs: nw_socket_copy_info [C1:2] getsockopt TCP_INFO failed [102: Operation not supported on socket] nw_socket_copy_info getsockopt TCP_INFO failed [102: Operation not supported on socket] Failed to set dependencies on asset 9303749952624825765 because NetworkAssetManager does not have an asset entity for that id. void * _Nullable NSMapGet(NSMapTable * _Nonnull, const void * _Nullable): map table argument is NULL PSO compilation completed for driver shader copyFromBufferToTexture so=0 sbpr=256 sbpi=16384 ss=(64, 64, 1) p=70 sc=1 ds=0 dl=0 do=(0, 0, 0) in 1997 XPC connection interrupted <<<< FigAudioSession(AV) >>>> audioSessionAVAudioSession_CopyMXSessionProperty signalled err=-19224 (kFigAudioSessionError_UnsupportedOperation) (getMXSessionProperty unsupported) at FigAudioSession_AVAudioSession.m:606 Failed to load item AXCodeItem<0x14706f250> [Rank:6000] SpringBoardUIServices [AXBundle name:/System/Library/AccessibilityBundles/SpringBoardUIServices.axbundle/SpringBoardUIServices] [Platforms and Targets:{ iOS = SpringBoardUIServices; } Framework] [Excluded: (null)]. error: Error Domain=AXLoading Code=0 "URL does not exist: file:///System/Library/AccessibilityBundles/SpringBoardUIServices.axbundle" UserInfo={NSLocalizedDescription=URL does not exist: file:///System/Library/AccessibilityBundles/SpringBoardUIServices.axbundle} Failed to load item AXCodeItem<0x14706f250> [Rank:6000] SpringBoardUIServices [AXBundle name:/System/Library/AccessibilityBundles/SpringBoardUIServices.axbundle/SpringBoardUIServices] [Platforms and Targets:{ iOS = SpringBoardUIServices; } Framework] [Excluded: (null)]. error: Error Domain=AXLoading Code=0 "URL does not exist: file:///System/Library/AccessibilityBundles/SpringBoardUIServices.axbundle" UserInfo={NSLocalizedDescription=URL does not exist: file:///System/Library/AccessibilityBundles/SpringBoardUIServices.axbundle} [b30780-MRUIFeedbackTypeButtonWithBackgroundTouchDown] Playback timed out before completion (after 3111 ms) Failed to set dependencies on asset 7089614247973236977 because NetworkAssetManager does not have an asset entity for that id.

This post documents an issue I reported in feedback FB19610114 and see if anyone knows of a workaround. Here is a copy of the feedback. Short version Manipulation (SwiftUI OR RealityKit) fails to translate entities after changing rooms. By changing rooms, I mean a human wearing an Apple Vision Pro leaving one room and entering another room. Once this issue occurs, it impacts all apps that use these features. A device restart is the only solution I have to fix it. Feedback FB19610114 This is an odd one. I'm using the new Manipulation Component in visionOS 26. Most of the time this works well. Sometime it stops working and when it does the only way to get it working again is to reboot the headset. When this happens, I can continue to rotate and scale items, but translation no longer works. It is as if the item is stuck to a fixed point in the parent scene (window, volume, etc). When this bug occurs, it affects every app across the entire operating system that is using manipulation, including the RealityKit component AND the SwiftUI version. This is not limited to one app and is not limited to apps that I am working on. Once this error occurs, it affects literally any application across the operating system that is using this API, including apps from Apple. I won't speculate on the cause of this, but I do know of one way where I can always get it to happen. Here is how to reproduce it: Make an Xcode project with a single entity that uses the Manipulation Component. There is no need to customize the configuration of this component. The default implementation will work. Build and run this app on device. You can keep running from device or quit and launch the app like normal on device. Open the app and manipulate the entity - it should work as expected. Physically walk into another room. It is vital that you leave the current room that you are in and enter a different room entirely. Use the digital crown to recenter your view and bring your window or volume to you. Test the manipulation on the entity again - it should still be working as expected at this point. Physically, move yourself and your headset into the original room where you started. Use the digital crown to recenter your view and bring your window or volume to you. Test the manipulation on the entity again - you should now see the issue. When I follow the steps above, then 100% of the time manipulation translation stops working at this point. It will impact any application using this API. The only way to fix it is to restart my headset. A few points to keep in mind It does not matter if an app is actively being run from Xcode. When this occurs, it impacts every single app, not just one. When this occurs, rotation and scaling continue to work, but the entity/view cannot be translated. This impacts BOTH the SwiftUI version and the RealityKit version. When this occurs, the only way to "fix" it is to reboot the device.

Has Roomplan been abandoned? Two years have gone by without comments from Apple on improvements. Are the improvements behind the scenes? Is there going to be any major updates?

Apple's new Spatial Personas use Gaussian Splatting, but I have not found any APIs for visionOS to display a Gaussian Splat like a PLY file. Am I just missing the Apple documentation? If not, are there common practices developers are using for displaying Gaussian Splats in visionOS?

I'm getting the following error message when compiling the Apple provided sample, Spaceship game for the Apple Visio Pro. I've already tried deleting the derived data resetting the package cache and restarting Xcode but still getting the following error: [xrsimulator] Exception thrown during compile: Cannot get rkassets content for path /Users/myoungkang/Downloads/CreatingASpaceshipGame/Packages/Studio/Sources/Studio/Studio.rkassets because 'The file “Studio.rkassets” couldn’t be opened because you don’t have permission to view it.' error: Tool exited with code 1

For visionOS 2.0+, it has been announced the object tracking feature. Is there any support for PolySpatial in Unity or is it only available in Swift and Xcode?

Happy new year to all! I have created an iOS app that also runs on Apple Vision Pro. On iOS, when you activate the fileImporter modal, you can swipe down the modal in iOS to dismiss. However, in visionOS, this same modal CANNOT be swiped down to cancel/dismiss. If you are drilled deep into a file hierarchy, you have to navigate back to the top level to tap X to dismiss. Is there a way to add swipe down to the visionOS implementation of fileImporter, or any other workaround so the user doesn't have to navigate back to the top to dismiss? Again, this is not a visionOS app but an iOS app compatible for use in Vision Pro. Thanks!

Sorry for the cross-post but it's now two days in and this isn't fixed. If you try to use Xcode 16.3b3 with visionOS, it won't download the visionOS SDK, gives a 'network error' so you can't use the latest beta for Apple Vision Pro. FB16927025 FB16917874 FB16910449

Hi I know it's possible to play equirectangular VR180 video either SBS or MV-HEVC. And for fisheye video, the only way I know is to convert it into an AIVU for playback. Is there any way to directly play fisheye video using AVPlayer? Thanks a lot!

I got more than 1 TB Immersive videos, and I want to play from them. Is there a way I can connect a ssd to Vision Pro via developer strap? Or is it possible to connect to a 10G Ethernet ad, and then using Ethernet to connect to a disk/NAS and attach the drive via ip?

Apple, please provide access to face tracking blend shapes on vision os, just like you do on iOS. You have the best eye and face tracking implementation on the market, please let us use it. There is a sizable audience who will buy the headset just for it. I personally know multiple people who are not buying the headset simply because you locked those features out. No raw camera access is needed, just abstracted blendshapes values. You will make the headset so much more useful if you do this simple thing.

After adding TextComponents to my Entities on visionOS, I have observed that visualBounds will ignore the TextComponents. Documentation states that it should render a rounded rectangle mesh. These mashes are visible on the device, but not visible in the debugger ("Capture Entity Hierarchy") and ignored by visualBounds. Am I missing something? static func makeDirection(_ direction: Direction) -> Entity { let text = Entity() text.name = direction.rawValue text.setScale(SIMD3(repeating: 5), relativeTo: nil) text.transform.rotation = direction.rotation text.components.set(direction.textComponent) return text } My workaround is to add a disabled ModelEntity and take its bounds 😬

I'm trying to run a PhotogrammetrySession based on photos taken in an AVCaptureSession and stored as .heic files. When I load the files I'm always seeing the error "Sample 0 missing LiDAR point cloud!" showing up for each individual sample. Debugging shows that sample.depthDataMap is populated, also the .heic contains depth data which can be extracted using e.g. heif-convert on my Mac. Comparing the .heic I created to one of the ObjectCaptureSession which doesn't show the LiDAR warning, I noticed the only difference being the HEIC information here: So my questions are: Are these the missing information in my manual capture causing this warning? Can I somehow add these information in an AVCaptureSession? Do these information allow better photogrammetry results?

I've got an Immersive scene that I want to be able to bring additional users into via SharePlay where each user would be able to see (and hopefully interact) with the Immersive scene. How does one implement that?

Hello, I am building a kiosk-style app for VisionOS which will be used in Guided Access mode, to be given to various visitors. So each of them will do hands + eyes setup, standard Guided Access thing. I want my experience to auto-start playing content when setup is done. I looked everywhere, but found no way do detect whether setup is complete? Also adding any kind of interface to start the app manually is risky, since buttons etc remain visible an interactable WHILE setup takes place. Delay-based approach also wont work, since setup can be skipped, or failed, or be done quickly, slowly... So it takes between 10 seconds and a few minutes. So the question is - is there any way to get notification, or check some bool or something that will tell me that Hands + Eyes setup in Guided mode is complete (or skipped)? Thanks in advance!

We're developing an iOS application that integrates RoomCaptureSession with ARSCNView for room scanning. Our implementation differs from the standard RoomCaptureView because we need custom UI guidance with 3D dots placed in the scanning environment to guide users through the capture process. Bug Description: The application crashes when users attempt to scan multiple rooms or apartments in sequence. The crash specifically occurs with the following pattern: User successfully scans first room with multiple hotspots (working correctly) User stops scanning, moves to a new room In the new room, first 1-2 hotspots work correctly Application crashes when attempting to scan additional hotspots Technical Details: Error: SLAM Anchor assertion failure in SlamAnchor.cpp:37 : HasValidPose() Crash occurs in Thread 27 with CAPIDetectionOutputFwdNode Error suggests invalid positioning when placing AR anchors Steps to Reproduce: Start room scan Complete multiple hotspot captures in first room Stop scanning Start new room scan Capture 1-2 hotspots successfully Attempt additional hotspot captures -> crashes Attempted Solutions: Implemented anchor cleanup between sessions Added position validation before anchor placement Implemented ARSession error handling Added proper thread management for AR operations Environment: Device: iPhone 14 Pro (LiDAR equipped) iOS Version: 18.1.1 (22B91) Testing through TestFlight Crash Log Details: Exception Type: EXC_CRASH (SIGABRT) Exception Codes: 0x0000000000000000, 0x0000000000000000 Exception Note: EXC_CORPSE_NOTIFY Triggered by Thread: 27 Thread 27 Crashed: 0 libsystem_kernel.dylib 0x00000001f0cc91d4 __pthread_kill + 8 1 libsystem_pthread.dylib 0x0000000228e12ef8 pthread_kill + 268 2 libsystem_c.dylib 0x00000001a86bbad8 abort + 128 3 AppleCV3D 0x0000000234d71a28 cv3d::vio::capi::SlamAnchor::SlamAnchor Question: Is there a recommended approach for handling multiple room captures with custom ARSCNView integration? The standard RoomCaptureView implementation doesn't show this behavior, but we need the custom guidance functionality that ARSCNView provides. Crash Log Code and full crash logs can be provided if needed.

I'm developing a VisionOS app with bouncing ball physics and struggling to achieve natural bouncing behavior using RealityKit's physics system. Despite following Apple's recommended parameters, the ball loses significant energy on each bounce and doesn't behave like a real basketball, tennis ball, or football would. With identical physics parameters (restitution = 1.0), RealityKit shows significant energy loss. I've had to implement a custom physics system to compensate, but I want to use native RealityKit physics. It's impossible to make it work by applying custom impulses. Ball Physics Setup (Following Apple Forum Recommendations) // From PhysicsManager.swift private func createBallEntityRealityKit() -> Entity { let ballRadius: Float = 0.05 let ballEntity = Entity() ballEntity.name = "bouncingBall" // Mesh and material let mesh = MeshResource.generateSphere(radius: ballRadius) var material = PhysicallyBasedMaterial() material.baseColor = .init(tint: .cyan) material.roughness = .float(0.3) material.metallic = .float(0.8) ballEntity.components.set(ModelComponent(mesh: mesh, materials: [material])) // Physics setup from Apple Developer Forums let physics = PhysicsBodyComponent( massProperties: .init(mass: 0.624), // Seems too heavy for 5cm ball material: PhysicsMaterialResource.generate( staticFriction: 0.8, dynamicFriction: 0.6, restitution: 1.0 // Perfect elasticity, yet still loses energy ), mode: .dynamic ) ballEntity.components.set(physics) ballEntity.components.set(PhysicsMotionComponent()) // Collision setup let collisionShape = ShapeResource.generateSphere(radius: ballRadius) ballEntity.components.set(CollisionComponent(shapes: [collisionShape])) return ballEntity } Ground Plane Physics // From GroundPlaneView.swift let groundPhysics = PhysicsBodyComponent( massProperties: .init(mass: 1000), material: PhysicsMaterialResource.generate( staticFriction: 0.7, dynamicFriction: 0.6, restitution: 1.0 // Perfect bounce ), mode: .static ) entity.components.set(groundPhysics) Wall Physics // From WalledBoxManager.swift let wallPhysics = PhysicsBodyComponent( massProperties: .init(mass: 1000), material: PhysicsMaterialResource.generate( staticFriction: 0.7, dynamicFriction: 0.6, restitution: 0.85 // Slightly less than ground ), mode: .static ) wall.components.set(wallPhysics) Collision Detection // From GroundPlaneView.swift content.subscribe(to: CollisionEvents.Began.self) { event in guard physicsMode == .realityKit else { return } let currentTime = Date().timeIntervalSince1970 guard currentTime - lastCollisionTime > 0.1 else { return } if event.entityA.name == "bouncingBall" || event.entityB.name == "bouncingBall" { let normal = event.collision.normal // Distinguish between wall and ground collisions if abs(normal.y) < 0.3 { // Wall bounce print("Wall collision detected") } else if normal.y > 0.7 { // Ground bounce print("Ground collision detected") } lastCollisionTime = currentTime } } Issues Observed Energy Loss: Despite restitution = 1.0 (perfect elasticity), the ball loses ~20-30% energy per bounce Wall Sliding: Ball tends to slide down walls instead of bouncing naturally No Damping Control: Comments mention damping values but they don't seem to affect the physics Change in mass also doesn't do much. Custom Physics System (Workaround) I've implemented a custom physics system that manually calculates velocities and applies more realistic restitution values: // From BouncingBallComponent.swift struct BouncingBallComponent: Component { var velocity: SIMD3<Float> = .zero var angularVelocity: SIMD3<Float> = .zero var bounceState: BounceState = .idle var lastBounceTime: TimeInterval = 0 var bounceCount: Int = 0 var peakHeight: Float = 0 var totalFallDistance: Float = 0 enum BounceState { case idle case falling case justBounced case bouncing case settled } } Is this energy loss expected behavior in RealityKit, even with perfect restitution (1.0)? Are there additional physics parameters (damping, solver iterations, etc.) that could improve bounce behavior? Would switching to Unity be necessary for more realistic ball physics, or am I missing something in RealityKit? Even in the last video here: https://stepinto.vision/example-code/collisions-physics-physics-material/ bounce of the ball is very unnatural - stops after 3-4 bounces. I apply custom impulses, but then if I have walls around the ball, it's almost impossible to make it look natural. I also saw this post https://developer.apple.com/forums/thread/759422 and ball is still not bouncing naturally.

I'm developing a visionOS panorama viewer app where I need to implement an auto-hiding floating menu in immersive space. The menu should: Show for 3 seconds when entering immersive mode Auto-hide after 3 seconds, Reappear when user taps anywhere (using SpatialTapGesture). Buttons should respond to gaze + pinch interaction The Problem: When I add .windowStyle(.plain) to achieve transparent window background for the auto-hide effect, all buttons in the menu become completely unresponsive to gaze + pinch interaction. The buttons only respond to direct finger touch (poking). Without .windowStyle(.plain): Buttons work correctly with gaze + pinch, but I cannot achieve transparent window background for hiding. With .windowStyle(.plain): Window can be transparent, but buttons lose gaze + pinch interaction. Code: App.swift: @main struct MyApp: App { @StateObject private var model = AppModel() var body: some Scene { WindowGroup(id: "MainWindow") { ContentView() .environmentObject(model) } .defaultSize(width: 900, height: 700) .windowResizability(.contentSize) .windowStyle(.plain) // <-- This causes the interaction issue ImmersiveSpace(id: "ImmersiveSpace") { ImmersiveView() .environmentObject(model) } } } ContentView.swift (simplified): struct ContentView: View { @EnvironmentObject var model: AppModel @State private var isMenuVisible: Bool = true var body: some View { VStack { if model.isImmersiveViewActive { if isMenuVisible { // This menu's buttons don't respond to gaze+pinch immersiveControlMenu } } else { mainMenuButtons } } .glassBackgroundEffect() } private var immersiveControlMenu: some View { HStack { Button("Exit") { exitImmersiveSpace() } .buttonStyle(.bordered) // Also tried .plain, same issue } .padding() .glassBackgroundEffect() } } ImmersiveView.swift: struct ImmersiveView: View { @EnvironmentObject var model: AppModel var body: some View { RealityView { content in // Panorama sphere let sphere = ModelEntity(mesh: .generateSphere(radius: 1000), materials: [material]) content.add(sphere) // Tap detector for menu toggle let tapDetector = Entity() tapDetector.components.set(CollisionComponent(shapes: [.generateSphere(radius: 900)])) tapDetector.components.set(InputTargetComponent()) content.add(tapDetector) } .gesture( SpatialTapGesture() .targetedToAnyEntity() .onEnded { _ in model.shouldShowMenu = true } ) } } Environment: Xcode 26.2 visionOS 26.3 Vision Pro device Questions: Is .windowStyle(.plain) expected to affect button interaction behavior? What is the recommended approach to achieve a transparent/hidden window in immersive mode while maintaining button interactivity? Is there an alternative to .windowStyle(.plain) for hiding window chrome in visionOS? Thank you for any guidance!

Hi everyone, I’m building a visualization app for VisionPro that uses SharePlay and GroupActivities to explore datasets collaboratively. I’ve successfully implemented the new SharedWorldAnchor feature, and everything works well with nearby, local participants. However, I’m stuck on one point: How can I share a world anchor with remote participants who join via FaceTime as spatial personas? Apple’s demo app (where multiple users move a plane model around) seems to suggest that this is possible. For context, I’m building an immersive app with Metal rendering. Any guidance or examples would be greatly appreciated! Thanks, Jens