I noticed in the latest macOS beta 3 that there was this update: A new algorithm significantly improves PhotogrammetrySession reconstruction quality of low-texture objects not captured with the ObjectCaptureSession front end. It will be downloaded and cached once in the background when the PhotogrammetrySession is used at runtime. If network isn’t available at that time, the old low quality model will be used until the new one can be downloaded. There is no code change needed to get this improved model. (145220451) I am not noticing any difference to before with the reconstructions I tested so I am assuming it's reverting to the old model but in the logs there is no way to see if it succeeds or fails to download that new model. do you have any more information on what was improved here with some examples and what we should be looking for? also how can confirm the download of that new model has not failed?

Hi Apple Team, We noticed the following exciting changelog in the latest macOS 26 beta: A new algorithm significantly improves PhotogrammetrySession reconstruction quality of low-texture objects not captured with the ObjectCaptureSession front end. It will be downloaded and cached once in the background when the PhotogrammetrySession is used at runtime. If network isn’t available at that time, the old low quality model will be used until the new one can be downloaded. There is no code change needed to get this improved model. (145220451) However after trying this on the latest beta and running some tests we do not see any differences on objects with low textures such as single coloured surfaces. Is there anything we are missing? the machine is definitely connected to the internet but we have no way of knowing from the logs if the new model is being used? thanks

TL;DR: RealityKit and Reality Composer Pro aren't forward or backward compatible with each other, and the resulting error message is terse and unhelpful. (FB14828873) So far, I've been sticking with Xcode 16.4 for development and only using Xcode 26.0 beta experimentally. Yesterday, I used xcode-select to switch to Xcode 26.0 beta 3 to test it, but I forgot to switch back. Consequently, this morning I unintentionally used the future Reality Composer Pro (the version included with Xcode 26) to make a small change to a USD file. Now I realize that if I'm unlucky, it's possible Reality Composer Pro may have silently introduced a small change into the USD file that may make RealityKit fail to read the file on iOS 18 and visionOS 2, which in the past has resulted in hours of debugging to track down the source of the failure, often a single line in the USD file that RealityKit can't communicate to me other than with the error "the operation couldn't be completed". As an analogy, this situation is as if, during regular development (not involving Reality Composer Pro), Xcode didn't warn you about specific API version conflicts, but instead failed with a generic error message, without highlighting the line in your Swift file that was the source of the error.

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)

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?

I have discovered that RemoteImmersiveSpace is limited to utilizing the structure of the CompositorContent protocol, precluding direct invocation of RealityView. Consequently, I am interested in understanding the appropriate method for integrating CompositorContent within RemoteImmersiveSpace. Thanks.

Spatial widget is a new feature of visionos 26. I notice The system’s Photo app can add a Spatial Image in the widget. I wonder if third apps can use spatial image or any 3D content in it's widget? I try to use RealityView in widget and it run with a crash. So does spatial Image in widget only supported by the system Photo app, and not available to developers now?

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) }

Hello, Thank you for your time. I have a question regarding visionOS app development. When placing a SwiftUI TextField inside RealityView.attachments, we found that focusing on the field does not bring up the virtual keyboard in front of the user. Instead, the keyboard appears around the user’s lower abdomen area. However, when placing the same TextField in a regular SwiftUI layer outside of RealityView, the keyboard appears in the correct position as expected. This suggests that the issue is specific to RealityView.attachments. We are currently exploring ways to have the virtual keyboard appear directly in front of the user when using TextField inside RealityViewAttachments. If there is any method to explicitly control the keyboard position or any known workarounds—including alternative UI approaches—we would greatly appreciate your guidance. Best regards, Sadao Tokuyama

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.

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.

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 working with the new PortalComponent introduced in visionOS 2.0, and I've encountered some issues when transitioning entities between virtual and real-world spaces using crossingMode. Specifically: Lighting inconsistency: When CG content (ModelEntities with PhysicallyBasedMaterial) crosses the portal from virtual space into the real environment, the way light reflects on the objects changes noticeably. This causes a jarring visual effect, as the same material appears differently depending on the space it's in. Unnatural transition visuals: During the transition, the CG models often appear to "emerge from the wall," especially when crossing from virtual to real. This ruins the immersive illusion and feels visually unnatural. IBL adjustment attempts: I’ve tried adding an ImageBasedLightComponent to the world entity, and while it slightly improves the lighting consistency, the issue still remains to a noticeable degree. My goal is to create a seamless visual experience when CG entities cross between spaces, without sudden lighting shifts or immersion-breaking geometry reveals. Has anyone else experienced similar issues? Is there a recommended setup or workaround to better control lighting and visual fidelity when using crossingMode with portals in visionOS 2.0? Any guidance would be greatly appreciated. Thank you!

Hello again! So, I am trying my best to figure out how to get a simple, fully Immersive scene that is navigable via the PS5 controller to work on Vision Pro. I am even using ChatGPT (both externally AND in Xcode when it makes fix recommendations) however, I seem to be moving around in circles, having to constantly send error messages back to ChatGPT, then try the new "confirmed to work" fixed code, only to have more errors. I also have explicitly let the AI know that I'm using all the current betas, only to learn there has been lots of stuff deprecated (of course!). This is my initial query to ChatGPT: "How can I create a fully immersive VR experience for Apple Vision Pro, using a PS5 game controller to navigate the user via first person view through a 3d scene?" And then down the rabbit hole the past few days. Seriously, I thought AI was supposed to be helpful! LOL Per Xcode logs, I'm seeing thumbstick movement, but simply cannot get the camera to move. Is this even possible?! Is it a camera defined in Reality Composer Pro? Or, does it need to be defined outside. I'm trying to follow along with the AI, but it seems to go back and forth with different solutions. Here is the latest it's spitting out: __ Thanks for this super clear log — and you’re absolutely right: the gamepad input is working, CameraRoot is moving as expected… but your actual camera view is not changing. This strongly suggests the issue is that RealityKit is not using your custom camera (PerspectiveCameraComponent) for rendering the view. ⸻ 🔍 What’s happening Even though CameraRoot is moving (with correct logs), RealityKit will default to its own internal camera unless: 1. You explicitly assign your camera entity to be the one used, or 2. Your camera is the only one and RealityKit picks it up automatically. But in visionOS 2.0 / Xcode 26, there seems to be a bug or behavior change where RealityKit no longer auto-uses your custom PerspectiveCameraComponent in RealityView unless it’s: • properly set with content.camera, or • part of a .usda/Reality Composer camera hierarchy that’s already active. Unfortunately: 🟥 RealityView.camera = .entity(cameraRoot) ❌ Is no longer allowed (we confirmed this breaks with a type error) ⸻ ✅ The Reliable Workaround: Attach your camera to the Immersive scene itself Since we’re loading a .usda scene (Reality Composer Pro), here’s the most reliable path forward: ✔️ Option A: Embed your camera into the Reality Composer Pro scene 1. Open your Reality Composer Pro scene 2. Add a Camera entity directly in your Immersive.usda scene 3. Name it something like "PlayerCamera" 4. Position it at [0, 1.6, 0] 5. Leave it selected as the default camera Then in your Swift code: • Find the camera entity with .findEntity(named:) • Move that entity instead of creating a new one. ⸻ ✔️ Option B: Force RealityKit to use your custom camera (via transform parenting) Since direct camera assignment is unavailable, we can use a trick: Move the whole RealityView scene via a parent entity with a camera. But let’s go with Option A, since you’ve already got a Reality Composer Pro scene. __ Can any developer over there can make a suggestion on how to move forward? I JUST want to create a sandbox template so I can start exploring 3d scenes created in Maya and sent over to headset. I'll deal with animation next, but come on, why is this so difficult to get working? I'm not a programmer, but have been trying to wrap my head around Xcode and SwiftUI. This needs to be much simpler. Or, you need to provide us creatives with better sample templates and non programmer speak on how to set this up properly. Ideally, you HIRE us 3d professionals to work side by side with the programmers to help make these tools useable - especially Reality Composer Pro. Seriously, I am making a concerted effort to use the native tools, even though I would love to be porting Unreal Engine scenes over. If anyone can help point me in the right direction, coming from a 3D Creator/Animator/Modeler perspective, I, and my fellow peers in the XR/AR/VR community would greatly appreciate it. Thank you.

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.

I'm running into an issue in Xcode when working on a visionOS app. Whenever I try to drag a 3D model entity in my scene, the drag gesture doesn't work if there's a UITextView (or SwiftUI TextEditor) in background of the 3D entity. It seems like the UITextView is intercepting the gesture or preventing the drag interaction from reaching the 3D content. Interestingly, when the 3D entity is placed infront of the ScrollView, the drag works as expected. Has anyone else experienced this behavior? Is this a known limitation or a bug in the current tooling? Any workarounds or fixes would be appreciated. Thanks!

In Beta 1,2, and 3, we could pick up and inspect entities, bringing them closer while moving them outside of the bounds of a volume. As of Beta 4, these entities are now clipped by the bounds of the volume. I'm not sure if this is a bug or an intended change, but I files a Feedback report (FB19005083). The release notes don't mention a change in behavior–at least not that I can find. Is this an intentional change or a bug? Here is a video that shows the issue. https://youtu.be/ajBAaSxLL2Y In the previous versions of visionOS 26, I could move these entities out of the volume and inspect them close up. Releasing would return them to the volume. Now they are clipped as soon as they reach the end of the volume. I haven't had a chance to test with windows or with the SwiftUI modifier version of manipulation.

Hello, I'm adding a CollisionComponent to an entity in RealityView. CollisionComponent requires that a Mesh must be provided as a reference for collision detection. However, in order to achieve more accurate detection, I hope that this Mesh resource is a geometric shape of a USDZ model. Is there any way to make it happen? Thank you!

Hello everyone, I am currently developing an experience for visionOS using RealityKit and I would like to achieve volumetric light effects, such as visible light rays or shafts through fog or dust. I found this GitHub project: https://github.com/robcupisz/LightShafts, which demonstrates the kind of visual style I am aiming for. I would like to know if there is a way to create similar effects using RealityKit on visionOS. So far, I have experimented with DirectionalLight, SpotLight, ImageBasedLight, and custom materials (e.g., additive blending on translucent meshes), but none of these approaches can replicate the volumetric light shaft look shown in the repository above. Questions: Is there a recommended technique or workaround in RealityKit to simulate light shafts or volumetric lighting? Is creating a custom mesh (e.g., cone or volume geometry with gradient alpha and additive blending) the only feasible method? Are there any examples, best practices, or sample projects from Apple or other developers that showcase a similar visual style? Any advice or hints would be greatly appreciated. Thank you in advance!

Hi all, Is there a standard, good intuition about RealityKit battery drain vs how the Xcode profiler tends to display things? I have a realitykit ARview that i have permanently in my view hierarchy and the profiler is saying constantly 40% CPU spin, ~160mb increase of memory usage compared to when not using this view, and battery increased from Low to middle of the yellow bar for "high". Yet the actual battery drain attributed to my app is basically an order of magnitude lower than the battery drain of something like Tiktok for the same amount of foreground time (15% vs 2%) So I'm a little confused whether or not to trust the Xcode battery profiler when working with RealityKit and ARViews (unless Tiktok is in very high usage all the time!). Is this a largely ignorable signal, and more generally, how useful is this profiler? Doesn't seem immediately intuitive to me at first use. Thanks!