Hi all, I've encountered a potential issue with how the winding order of geometry is handled when their transformations involve negative scaling. I created a simple test asset, a single triangle, to demonstrate this. The triangle's vertices are defined in a counter-clockwise ("right-handed") winding order, and its transform has a negative scale on the X-axis. According to the OpenUSD specification, this negative determinant in the transformation matrix should effectively reverse the winding order of the geometry: However, any given gprim's local-to-world transformation can flip its effective orientation, when it contains an odd number of negative scales. This condition can be reliably detected using the (Jacobian) determinant of the local-to-world transform: if the determinant is less than zero, then the gprim's orientation has been flipped, and therefore one must apply the opposite handedness rule when computing its surface normals (or just flip the computed normals) for the purposes of hidden surface detection and lighting calculations. When I view the asset in tools like Blender or Preview on macOS, it behaves as expected. The triangle's effective orientation is flipped to CW. However, when the same asset is viewed in Reality Composer Pro or with QuickLook on iOS, its effective orientation remains CCW. In other words, the triangle faces the opposite direction. My questions for the community and Apple are: Is this behavior in RealityKit a known issue? If this is a known issue, is there official guidance for DCC tools on how to export USDZ assets to ensure they appear correctly in the Apple ecosystem? Any insights or recommendations would be greatly appreciated.

During regular use, RealityKit generates an excessive amount of internal logging that is not actionable by third party developers. When developing an iOS RealityKit/ARKit app, this makes the Xcode console challenging to use for regular work. (FB19173812) See screenshots below. Xcode does have an option for filtering out logging from specific SDKs, but enabling this feature to suppress the logging of RealityKit and related SDKs like PHASE is something developers have to do dozens of times each day. After a year of developing a RealityKit app, this process becomes frustrating. If SDKs like Foundation, UIKit, and SwiftUI generated as much logging as RealityKit and related SDKs, Xcode's console would be unusable. Is there any way to disable the logging of RealityKit and PHASE permanently? Thank you for any help you provide.

On an iPad running iPadOS 26 beta 4, when tapping the Game Center Access Point, the overlay doesn’t show the configured achievements, leaderboards or challenges. I should specify this is an in-development app and the achievements and leaderboards are in the “Not Live” state, however they show on other devices running iOS 18 in the Access Point UI. Anyone else having this issue? If so, how should I test achievements and leaderboards while iOS 26 beta is out? The UI looks like this on iPadOS 26:

Hello! I'm developing a GPU (shader) language, where I aim to target multiple backends with a common frontend. I wanted to avoid having to round trip through Metal, and go straight to IR just like I have with SPIRV, in order to have a fast and efficient compilation process. I've been looking for a reference page where I can read about Metals IR, and as far as I'm aware, it exists, but I can't seem to find it anywhere. Furthermore, if such a reference is available, is there also a toolkit where I can run validation on the output IR, and perhaps even run optimizations, much like spv-tools for SPIRV? Any help would be appreciated! Thanks, Gustav

Hi, developers, I maintain a shipped app that uses string concatenation to construct Metal shader and compile on-device. Beta 4 seems disabled __asm keyword, resulting the compilation failure. The error is: v2/GEMMKernel.cpp:229: error: program_source:23:9: error: illegal string literal in 'asm' __asm("air.simdgroup_async_copy_1d.p3i8.p1i8"); The relevant code is available at https://github.com/liuliu/ccv/blob/unstable/lib/nnc/mfa/v2/GEMMHeaders.cpp#L30 although any __asm will trip this. Please give us guidance on whether this is a regression or this will be something enforced in 26 release. Personally, I would consider this as a bug given it won't impact anything "compiled" shaders. Thanks for your patience reading this!

Hi all, I’m running into an issue when trying to reconstruct a 3D model using PhotogrammetrySession on macOS from a set of images captured via the iOS Object Capture sample app, specifically in Area mode. When I attempt to create the model from these images (using the raw Images/ folder exported directly from the capture session), I get the following errors: ERROR cv3dapi.pg: Internal error codes (2): 4011 4012 WARN cv3dapi.pg: Internal warning codes (1): 4507 Output error with code = -15 requestError: CoreOC.PhotogrammetrySession.Error.processError I use the "Images" directory directly exported from Object Capture with my iphone 12 pro max (has lidar) set to "area mode" in the object capture app here is an example heic image metadata from the sequence. heif-info Images/00044.869568833.HEIC MIME type: image/heic main brand: heic compatible brands: mif1, MiHE, MiPr, miaf, MiHB, heic image: 3024x4032 (id=49), primary tiles: 6x8, tile size: 512x512 colorspace: YCbCr, 4:2:0 bit depth: 8 thumbnail: 240x320 color profile: nclx alpha channel: no depth channel: yes size: 192x256 bits per pixel: 8 z-near: 1.173828 z-far: 2.552734 d-min: undefined d-max: undefined representation: uniform Z metadata: Exif: 960 bytes uri /tag:apple.com,2023:ObjectCapture#CameraTrackingState: 4 bytes uri /tag:apple.com,2023:ObjectCapture#CameraCalibrationData: 1015 bytes uri /tag:apple.com,2023:ObjectCapture#ObjectTransform: 48 bytes uri /tag:apple.com,2023:ObjectCapture#ObjectBoundingBox: 48 bytes uri /tag:apple.com,2023:ObjectCapture#RawFeaturePoints: 832 bytes uri /tag:apple.com,2023:ObjectCapture#PointCloudData: 23984 bytes uri /tag:apple.com,2023:ObjectCapture#BundleVersion: 5 bytes uri /tag:apple.com,2023:ObjectCapture#SegmentID: 4 bytes uri /tag:apple.com,2024:ObjectCapture#SessionUUID: 36 bytes uri /tag:apple.com,2024:ObjectCapture#CaptureMode: 4 bytes uri /tag:apple.com,2023:ObjectCapture#Feedback: 4 bytes uri /tag:apple.com,2023:ObjectCapture#WideToDepthCameraTransform: 48 bytes uri /tag:apple.com,2023:ObjectCapture#TemporalDepthPointClouds: 864026 bytes transformations: angle (ccw): 270 region annotations: none properties: camera intrinsic matrix: focal length: 2813.695557; 2813.695557 principal point: 1522.338502; 2002.843018 skew: 0.000000 camera extrinsic matrix: rotation matrix: -0.695 0.344 -0.632 0.007 -0.875 -0.483 -0.719 -0.340 0.606 Questions: • What do internal error codes 4011 and 4012 refer to? • Is there something specific about Area mode captures that require preprocessing before they’re compatible with PhotogrammetrySession? • Has anyone successfully reconstructed a model from an Area mode session using the stock Apple tools? NOTE: I can provide the folder of images for debugging if that would help!

I would love to use Background GPU Access to do some video processing in the background. However the documentation of BGContinuedProcessingTaskRequest.Resources.gpu clearly states: Not all devices support background GPU use. For more information, see Performing long-running tasks on iOS and iPadOS. Is there a list available of currently released devices that do (or don't) support GPU background usage? That would help to understand what part of our user base can use this feature. (And what hardware we need to test this on as developers.) For example it seems that it isn't supported on an iPad Pro M1 with the current iOS 26 beta. The simulators also seem to not support the background GPU resource. So would be great to understand what hardware is capable of using this feature!

Hello, we are working on a iOS game project, as progress, the project grows larger and larger. Because we are using other game dependencies and libraries, here larger and larger refers to the whole project, and our source files integrated and compiled by Xcode are not many. Now, it seems we hit a bottleneck, when I add new files or functions to the previous files to implement a new feature, Xcode compile stucks(stops), it's Indexing | Initializing datastore forever, cannot produce a final build. macOS 15.1, Xcode 16.2 Can you provide any solutions to solve this problem? Also submitted Feedback ID #FB18432749

Hi team, I'm looking for the RealityKit debugger in Xcode 26 beta 3. I'm running a RealityKit app on my iPad running iPadOS 26 b3, but the debugger option is not there in Xcode.

Hello everyone, I must have missed something but why isn't there a depthAttachmentPixelFormat to the new Metal 4 MTL4RenderPipelineDescriptor, unlike the old MTLRenderPipelineDescriptor? So how do you set the depth pixel format? Thanks in advance!

Hi everyone, I'm building a native iOS app using Unreal Engine 5.6 with Firebase for authentication and Firestore. The app uses a MetaHuman avatar and is meant to run as a standalone UE app on iPhone. I'm using this Firebase wrapper: 👉 https://pandoa.github.io/FirebaseFeatures/ I've followed all the steps, including: Adding GoogleService-Info.plist to the Xcode project and ensuring it’s in the correct target Calling FIRApp.configure() in AppDelegate Verifying the plist is bundled correctly However, the app crashes on launch, and Firebase does not initialize properly. Crash log shows: [FirebaseCore][I-COR000005] No app has been configured yet. Setup details: Unreal Engine: 5.6 (source build, macOS) iOS Deployment: 17.5 MetaHuman character packaged correctly and app launches fine without Firebase Has anyone here managed to get Firebase working inside a native Unreal Engine iOS app with this setup? I'd love to hear if there’s something I’m missing — maybe something with initialization timing or module loading? Thanks so much in advance 🙏

I've had no issue calling image files in my .swift files, but they are causing crashes when used in my .SKS files. When I set a sprite texture to an image in the inspector/ editor bar, at runtime when that sprite is being called I get the error: "Cannot get value with size 16. The type encoded as {CGRect={CGPoint=dd}{CGSize=dd}} is expected to be 32 bytes." From my research it has something to do with Apple switching from 32 to 64 bite machines. From chatGPT “SpriteKit under the hood uses NSKeyedUnarchiver to load your .sks file. That unarchiver decodes each archived property by reading a fixed‑size blob of bytes and mapping it into a C struct. In your case it ran into a mismatch”. I am using a 64-bite machine to write my code and 64-bite simulators and physical devices, so there isn't a clear cause of the mismatch. My scenes play fine in Xcode 16's preview window and my code builds, it just crashes at runtime. When I don’t use image textured assets in the SKS file it works fine. It loads animated labels, and plain color squares. I’ve been able to work around this for static things like a sprite with a background texture by. in a normal non-game swift file, writing code like: if let scene = SKScene(fileNamed: "GameScene2") { let bg = SKSpriteNode(imageNamed: "YourBackgroundImage") bg.position = CGPoint(x: scene.frame.midX, y: scene.frame.midY) bg.zPosition = -1 scene.addChild(bg) } The issue now is I want to make a particle emitter and other non static sprites, but my understanding of their properities isn’t deep enough to create them without the editor. Also when I set SKTexture in a swift file that causes the same runtime crash with the 16/32 error. Could you help me figure out how to fix the bug so I can use the editor again? Otherwise could you help me figure out how to write a workaround like I do for background images? I have a feeling the answer is in writing my own NSKeyedUnarchiver but I don’t know how to make sure it’s called instead of the default one. I've already tried cleaning my code multiple times and deleting and reading sprite nodes. Thank you.

Hi everyone, I’m running into an issue with RealityKit when trying to animate BlendShapes (ShapeKeys) while a skeletal animation is playing. The model is a rigged character in .usdz format with both predefined skeletal animations and BlendShapes (exported from Blender). The problem: when I play any animation using entity.playAnimation(...), the BlendShapes stop responding. Calling setBlendShapes(...) still logs that weights are being updated, but the visual changes are not visible. The exact same blend shape animation works perfectly when no animation is playing. In SceneKit the same model works as expected: shape keys get animated during animation playback. But not in realitykit Still, as soon as an animation starts, the shape keys don’t animate anymore. Here’s the test project on GitHub that demonstrates the issue clearly: https://github.com/IAMTHEBURT/RealityKitWitnBlendShapesSample The goal is to play facial expressions (like blinking or talking) while a body animation (like waving) is playing. Is this a known limitation in RealityKit? Or is there a recommended way to combine skeletal animations with real-time BlendShape updates? Thanks in advance for any insights.

Hello, I'm tracking down a bug where useResource doesn't seem to apply proper synchronization when a resource is produced by the render pass then consumed by the compute pass, but when I use MTLFence between the to signal and wait between the render/compute encoders, the artifact goes away. The resource is created with MTLHazardTrackingModeTracked and useResource is called on the compute encoder after the render pass. Metal API Validation doesn't report any warnings/errors. Am I misunderstanding the difference between the two APIs? I dug through the Metal documentation and it looks like useResource should handle synchronization given the resource has MTLHazardTrackingModeTracked but on the other hand, MTLFence should be used to ensure proper synchronization between command encoders. Can someone can clarify the difference between the two APIs and when to use them.

I'm having issues getting a well built package from the Apple Unity Plug-in project. When building the my game project in Unity the following error is printed to the console: Apple.Core.AppleNativeLibraryUtility] Cannot locate a Debug or Release Apple.Core native library for iOS. Please ensure that the build invocation (build.py, xcodebuild, or Xcode) compiled cleanly and that the build was configured to support Debug on iOS. As far as I can tell the build did compile cleanly, but I might be missing something. If anyone can see what I'm doing wrong or has any insight it would be greatly appreciated. Setup is the following: macOS Tahoe 26 Beta Xcode-beta Version 26.0 beta 3 (17A5276g) Unity Plug-in branch: 2025-beta1 Unity game project version: 2022.3.60f M1 Macbook Pro The built packages have been imported into the game project through the Unity Package Manager using the tarball option pointing to the built packages from the Unity Plug-in project. The Unity Plug-in project has been built using the build.py file with the following: python3 build.py -m iOS iPhoneSimulator -p Core GameKit CoreHaptics GameController -k all The output is available in the attached file. build-output.txt Here's an image of the NativeLibraries~ folder inside the built Apple.Core package.

Hello RealityKit developers, I'm currently working on physics simulations in my visionOS app and am trying to adapt the concepts from the official sample Simulating physics joints in your RealityKit app. In the sample, a sphere is connected to the ceiling using a PhysicsRevoluteJoint to create a hinge-like simulation. I've successfully modified this setup to use a PhysicsSphericalJoint instead. The basic replacement works as expected: pin1 (attached to the sphere) rotates freely around pin0 (attached to the ceiling), much like a ball-and-socket joint should, removing all translational degrees of freedom. My challenge lies with the PhysicsSphericalJoint's angularLimitInYZ property. The documentation mentions that this property allows limiting the rotation around the Y and Z axes, defining an "elliptical cone shape around the x-axis of pin0." However, I'm struggling to understand how to specify these values to achieve a desired rotational limit. If I have a sphere that is currently capable of rotating 360 degrees around pin0 (like a free-spinning ball on a string), how would I use angularLimitInYZ to restrict its rotation to a certain height or angular range, preventing it from completing a full circle? Specifically, I'm trying to achieve a "swing" like behavior where the sphere oscillates back and forth but cannot rotate completely overhead or underfoot. What values or approach should I use for the angularLimitInYZ tuple to define such a restricted pendulum-like motion? Any insights, code examples, or explanations on how to properly configure angularLimitInYZ for this kind of behavior would be incredibly helpful! The following code is modified from the sample. extension MainView { func addPinsTo(ballEntity: Entity, attachmentEntity: Entity) throws { let hingeOrientation = simd_quatf(from: [1, 0, 0], to: [0, 0, 1]) let attachmentPin = attachmentEntity.pins.set( named: "attachment_hinge", position: .zero, orientation: hingeOrientation ) let relativeJointLocation = attachmentEntity.position( relativeTo: ballEntity ) let ballPin = ballEntity.pins.set( named: "ball_hinge", position: relativeJointLocation, orientation: hingeOrientation ) // Create a PhysicsSphericalJoint between the two pins. let revoluteJoint = PhysicsSphericalJoint(pin0: attachmentPin, pin1: ballPin) try revoluteJoint.addToSimulation() } } The following image is a screenshot of the operation when changing to PhysicsSphericalJoint. Thank you in advance for your assistance.

Hello everyone, I'm working on a visionOS application using RealityKit and am encountering a common coordinate system challenge when integrating 3D models created in Blender. My goal is to display and dynamically update the Transform (position, rotation, scale) of models created in Blender within RealityKit. The issue arises because Blender's default coordinate system is Z-up, and while exporting to USD/USDZ, I don't have a reliable "Y-up" export option that correctly reorients the model and its transform data for RealityKit's Y-up convention. This means I'm essentially exporting models with their "up" direction along the Z-axis. When I load these Z-up exported models into RealityKit, they are often oriented incorrectly. To then programmatically update their Transform (e.g., move them, rotate them based on game logic, or apply physics), I need to ensure that the Transform values I set align with RealityKit's Y-up system, even though the original model data was authored in a Z-up context. My questions are: What is the recommended transformation process (e.g., using simd_quatf or simd_float4x4) to convert a Transform that was conceptually defined in a Z-up coordinate system to RealityKit's Y-up coordinate system? Specifically, when I have a Transform (or its translation, rotation, scale components) from a Z-up context, how should I apply this to a RealityKit Entity so it appears and behaves correctly in a Y-up world? Are there any existing convenience APIs or helper functions within RealityKit, simd, or other Apple frameworks that simplify this Z-up to Y-up Transform conversion process? Or is a manual application of a transformation quaternion (e.g., simd_quatf(angle: -.pi / 2, axis: [1, 0, 0])) the standard approach? Any guidance, code examples, or best practices from those who have faced similar challenges would be greatly appreciated! Thank you.

The following code using the new GestureComponent demonstrates inconsistency. The tap gesture prints output, but the drag gesture does not. I already checked this post, which points to this seemingly outdated sample code I assume that example is deprecated in favour of the now built in version of GestureComponent. Nonetheless, there are no compiler warnings or errors, it just fails silently. TapGesture, LongPressGesture, MagnifyGesture, RotateGesture all work, so this feels like an oversight. RealityView { content in let testEntity = ModelEntity(mesh: .generateBox(size: .init(x: 1, y: 1, z: 1))) testEntity.position = SIMD3<Float>(0,0,-1) testEntity.components.set(InputTargetComponent()) testEntity.components.set(CollisionComponent( shapes: [.generateBox(size: .init(x: 1, y: 1, z: 1))] )) let testGesture = TapGesture() .onEnded { value in print("Tapped") } testEntity.components.set(GestureComponent(testGesture)) let dragGesture = DragGesture() .onEnded { value in print("Dragged") } testEntity.components.set(GestureComponent(dragGesture)) content.add(testEntity) }

In SceneKit, when creating an .scn file from a rigged model, the framework created an SCNNode for each bone/joint, so you could add and remove child nodes directly to and from joints, and like any other SCNNode, you could access world position and world orientation for each joint. The analog would be for joints to be accessible as child entities of a ModelEntity in RealityKit. I am unable to proceed with migrating my project from SceneKit because of this, as there does not seem to be a way to even access the true world position of a joint with the current jointNames/jointTransforms paradigm. The translation information from the given transforms is insufficient to determine the location of a joint at any given time, and other approaches like creating a GeometricPin for the given joint name and attaching it to another entity do not seem to work. So conveniently being able to attach an item to the hand of a rigged model was trivial in SceneKit and now feels impossible in RealityKit. I am not the first person to notice this, and am feeling demoralized about proceeding with RealityKit with such a critical piece of functionality blocked https://stackoverflow.com/questions/76726241/how-do-i-attach-an-entity-to-a-skeletons-joint-in-realitykit Will this be addressed in some way?

Breaking Through PolySpatial's ~8k Object Limit – Seeking Alternative Approaches for Large-Scale Digital Twins Confirmed: PolySpatial make Doubles MeshFilter Count – Hard Limit at ~8k Active Objects (15.9k Total) Project Context & Research Goals I’m developing an industrial digital twin application for Apple Vision Pro using Unity’s PolySpatial framework (RealityKit rendering in Unbounded_Volume mode). The scene contains complex factory environments with: Production line equipment Many fragmented grid objects need to be merged.) Dynamic product racks (state-switchable assets) Animated worker avatars To optimize performance, I’m systematically testing visionOS’s rendering capacity limits. Through controlled stress tests, I’ve identified a critical threshold: Key Finding When the total MeshFilter count reaches 15,970 (system baseline + 7,985 user-created objects × 2 due to PolySpatial cloning), the application crashes consistently. This suggests: PolySpatial’s mirroring mechanism effectively doubles GameObject overhead An apparent hard limit exists around ~8k active mesh objects in practice Objectives for This Discussion Verify if others have encountered similar limits with PolySpatial/RealityKit Understand whether this is a: Memory constraint (per-app allocation) Render pipeline limit (Metal draw calls) Unity-specific PolySpatial behavior Explore optimization strategies beyond brute-force object reduction Why This Matters Industrial metaverse applications require rendering thousands of interactive objects . Confirming these limits will help our team: Design safer content guidelines Prioritize GPU instancing/LOD investments Potentially contribute back to PolySpatial’s optimization I’d appreciate insights from engineers who’ve: Pushed similar large-scale scenes in visionOS Worked around PolySpatial’s cloning overhead Discovered alternative capacity limits (vertices/draw calls)