On visionOS 27 (27.0 Seed 4, 24M5326g), adding a ManipulationComponent to a single entity makes RealityKit remove the RealityView's ENTIRE content hierarchy from the scene and re-add it about 30ms later. Not just the manipulated entity — everything. Same app binary on visionOS 26: never happens. FB24092291. Bisected on the same build and device: two entities, neither with manipulation -> no detach add one entity with a ManipulationComponent -> detach, every time same entity, manipulation swapped for my own drag/rotate/ scale gesture handling -> no detach Everything else is identical between the last two cases — same entity, same collision shape, same InputTargetComponent, same ViewAttachmentComponent, same scene, same view hierarchy. The only variable is whether the component is installed. Setup: an immersive space containing a SwiftUI RealityView. One root entity added to the RealityViewContent in the make closure; all app content built as its descendants. The detach is brief but not harmless: every entity in the hierarchy gets SceneEvents.WillRemoveEntity and then DidAddEntity, so anything keyed to those events runs a full teardown and re-add for content that was never meant to go anywhere. In my app that cascade is what makes the scene visibly empty and rebuild. Nothing in my code removes it. Stack captured from a WillRemoveEntity subscription on the root — frame 0 is my callback, everything above it is framework: SwiftUI (AttributeGraph StatefulRule.withObservation ...) -> RealityFoundation -> CoreRE -> entity removal Ruled out before landing on the component: make runs exactly once, the root re-enters a scene with the same ObjectIdentifier (not a re-host), the hosting controller is never deallocated, there is exactly one RealityViewContent.add(_:) in the whole app, and with Self._printChanges() in the view's body the detach happens in an update pass where the body is not re-evaluated at all. Also ruled out: transient overlay UI (suppressed entirely, still detached) and entity count. Workaround if this is biting you: skip ManipulationComponent on 27 and handle drag/rotate/scale yourself. That is what I am doing for now. If you are debugging something similar: SceneEvents.WillRemoveEntity also fires when an entity merely leaves a scene, so a teardown-looking log is not proof anything was destroyed. Check whether make ran twice and whether the scene identity changed before suspecting your own code.

First of all, great work on the new Reality Composer Pro! I'm really enjoying it. Reality Composer Pro 3.0 (build 80.0.1.500.1/beta 4) currently shows GPT-5.5 as the newest OpenAI model in the Assistant model selector. The bundled configuration appears to contain a static model catalogue and does not include: gpt-5.6-sol gpt-5.6-terra gpt-5.6-luna Could Apple update Reality Composer Pro to support the current GPT-5.6 model family, or document whether users can add supported OpenAI model identifiers through the Custom Model Provider interface? It would also be helpful to clarify whether the Assistant supports GPT-5.6's tool-calling and reasoning behavior through its current Chat Completions integration. Thanks!

I’m working with an asset pack that contains multiple characters with a common skeletal hierarchy. This asset pack also contains animations for these characters as separate files. i.e.: run pick up idle throw In RCP2, I was able to add these animations to an Animation Library component without issue. However, in RCP3, I cannot add the skeletal animations to compatible characters. The only animations available are the transform animations, which do not animate the character. So in the animation library, I will see i.e.: run_transform idle_transform throw_transform But not run idle throw Is there something that I need to do to make the skeletal animations visible? I’ve tried adding the animation library at various levels across the hierarchy and it doesn’t seem to make a difference. Only the _transform animation is visible. I can verify that the animations themselves are compatible by manually editing the tm_entity for a character and adding the animation clip. I have to do this while RCP3 is closed but it works. I did notice that each animation is referencing a skeleton uuid. Is the UI only exposing skeletal animations with a matching skeleton uuid? Lastly, I can import both the character and the animation files as USD in a RealityKit app and successfully assign animations.

When trying to preview a new scene in Reality Composer Pro 3 on my Vision Pro, the mac app is launching the preview app in the headset but not actually establishing the connection. The headset app shows a red Disconnected icon and says "Connect to this device form Reality Composer Pro on Mac" and then the mac side says "Connecting" for a few seconds before saying "Connection Failed". Neither app provides any other details of the connection issue and I could not find any helpful details in Console messages. I got this to work somehow on two scenes in beta 3 but have had no luck in Beta 4. It seems that once a scene does connect to the device once, it will happily reconnect in the future. But establishing that first connection seems to be the key hurdle and I cannot get over it tonight. Anyone else run into similar issues?

Hi there, I'd appreciate it if someone could test this on their AVPro, and give feedback etc. https://face-flight.com I'm hoping to add WebXR support for as many headsets as possible.

Hello, I'm a VTuber, or a content creator who uses a virtual avatar instead of my real life face for presentation. I've been trying to use the iOS implementation of ARKit to get face tracking working, but noticed a major issue, especially for my purposes. ARKit appears to move both eyebrows the same, regardless of whether or not I am physically moving my eyebrows the same. Every ARKit based application I've used simply mirrors eyebrow movement. Is there a fix to this, or is this just a bug currently in the software?

Hello, I am not a developer, and I am programming a macOS application in Xcode with the help of Claude (Anthropic). Claude writes the code for me, but I need your help regarding the technical specifications. I am not trying to understand every technical detail, nor am I looking for someone to write the code for me. My goal is to obtain the most accurate technical information possible so that I can pass it directly to Claude, who will then generate the corresponding code in Xcode. This application is a personal project. It is not intended to be published on the App Store. Its sole purpose is to automatically process my own videos recorded with two DJI Action 6 cameras. It requires no user settings : the optimal parameters are defined by default within the application, and it directly outputs the final video file. My objective is to create an application that takes as input the two videos from the DJI Action 6 cameras mounted on a rig, and then produces a file that delivers a 180° immersive video with the highest possible quality. Claude has already conducted research in Apple's documentation as well as various online resources to determine the correct output format, encoding parameters, and necessary constraints. Despite this, the current result is not satisfactory. Currently, my application does not directly generate a file in the Apple Immersive format. Instead, it produces a stereoscopic video file containing both the left and right views side-by-side. I then use Moonlink on my Mac Studio and on my Apple Vision Pro, to play this video (the video is on the Mac and sent to the vision pro, wich works perfectly fine on some videos I have found on internet). The file is correctly recognized and can be opened, but the result presents two major issues: Playback is extremely choppy (about one second of playback followed by several seconds of freezing); The image is pinched at the top and bottom, which likely indicates an issue with the format, projection, or how the player interprets the file. I have no preference regarding the final format. If the best solution is to directly generate a file in the Apple Immersive format compatible with the Apple Vision Pro, that works for me. If, on the other hand, the best approach is to produce a high-quality stereoscopic video with the left and right views side-by-side, which can then be interpreted by a player application on the Apple Vision Pro or potentially other headsets, that works for me as well. My sole objective is to obtain: The best possible image quality (even though I am aware it will not match the quality produced by a Blackmagic URSA Cine Immersive camera); Perfectly smooth playback, without any stuttering or freezing; A correctly displayed 180° immersive video. Here are the specifications of my source videos: Two DJI Action 6 cameras. Interpupillary/Inter-axial distance (between optical centers): 6.5 cm. FOV Boost lens mounted on each camera (182° coverage). Video mode: 4K Custom 3840 \times 3840 at 60 fps. Color profile: D-Log M. Color conversion planned in the application via a DJI D-Log M to Rec.709 LUT in .cube format (.cube provided to the app). In FOV Boost mode, DJI enforces Electronic Image Stabilization (EIS) and removes gyroscope data from the MP4 files. Stereoscopic alignment is therefore performed via image analysis within the application (using ORB feature detection). With no FOV Boost I do not reach 180° with is the target. Fixed aperture. Both cameras are triggered simultaneously using the DJI GPS remote control. Fine synchronization is subsequently performed in the application via audio cross-correlation. My question is therefore the following : What exactly are the output specifications that my application should produce (format, codec, resolution, projection, metadata, encoding parameters, and any other technical constraints) in order to achieve the best possible result for a 180° immersive video on the Apple Vision Pro? I will pass all this information directly to Claude so that he can generate the corresponding code in Xcode. Therefore, I am not looking for a code example, but rather the technical specifications to follow. I would also welcome any advice or recommendations that a specialist in Vision Pro, immersive video, or Apple technologies might deem useful to improve the quality of the result or simplify my approach. Thank you in advance for your help. Best regards, Georges attached is a screen capture of the pinched image in the Vision Pro, this happens both at top and bottom, only displayed bottom.

It appears that Reality Composer Pro 3 does not export a full USDZ file with textures, animations, etc. The best I can tell is that it exports only the mesh (as suggested by the menu name). I have been using Reality Composer Pro 2 to create USDZ files to use as website environments. Will we no longer be able to create website environments with RCP 3? Will there be any way to download and use the old RCP 2 after upgrading to macOS 27 and Xcode 27?

Hello Apple Developer Team, We are currently developing an enterprise medical navigation application for Apple Vision Pro and would like to request clarification regarding the currently available visionOS Enterprise APIs related to camera access. Our application scenario involves real-time medical/surgical navigation and instrument tracking in a professional enterprise environment. We would like to better understand the following: How many cameras on Apple Vision Pro are currently accessible through the Enterprise APIs? Which specific cameras are accessible? For example: Main RGB cameras Passthrough cameras Tracking cameras Front-facing cameras Depth sensors LiDAR or structured-light related sensors Are simultaneous multi-camera streams supported? Does the Enterprise API provide: Real-time image frames Camera intrinsic/extrinsic parameters Stereo camera data Depth information Low-latency tracking-related data Are there any restrictions regarding the use of Vision Pro cameras for: Medical navigation Surgical guidance Instrument tracking Enterprise healthcare software Is Apple Vision Pro currently permitted or recommended for medical enterprise spatial-navigation workflows under the Enterprise APIs? We would greatly appreciate any official clarification regarding the current capabilities and limitations of camera access on Apple Vision Pro for enterprise medical applications.

Hello I'm trying to create a custom component but I can't see to make It work on the RCP3. The video 'Extend Reality Composer Pro 3 functionality with Xcode' makes it seem too simple but I can't seem to work following the steps.

Hello, I have an app preset meant for none vr/ar game. Th RCP3 was linked to the xcode successfully. I want to do everything in code as much as possible. not inside RCP3. So i would like to load my assets into variables in xcode, and then initialize them the way i like. (Get entities, get each of their ModelComponents, and place those components in my own array) But for some reason i cant get xcode to load my linked content using the API commands. Maybe it have something to do with the fact im using the APP present not an visionOS preset? xcode cant find the RCP3 project. if i drag and drop the realitycomposerpro into my xcode outliner, it will find the project but will refuse to find any entity or root Is it possible to have please a baby step tutorial from start to finish how to import content from RCP3 beta 3 into xcode Version 26.5 using the app preset ? Thank you

Hello, everyone. How to paste a mov transparent video of a generation channel on a model in RCP? In addition, how to paste a video with a black background on the model and remove the black part to make it transparent.

Hello guys, Reality composer pro really packed with a lot of math tools. i can easily create modular textures just as good as substance painter's smart materials and then reuse them across my games with minor node seed changes. The problem is, they will calculate in real time. Considering every game needs baked textures, Is there any future plans to develop a baking system inside reality composer pro ?

The particle system produced in RCP, when the VisionPro knob is long pressed to reset, the motion direction of the Motion parameter is misaligned. Is there a way to solve it?

Ok trying to play Spatial Audio on my VisionPro. OS26.4, using Xcode 26.4.1. Every attempt gives me the following error. <<<< FigAudioSession(AV) >>>> signalled err=-19224 at <>:612 I have tried the sample code at https://developer.apple.com/documentation/visionos/playing-spatial-audio-in-visionos and it gives the same error.

I have been analyzing Apple's Gaussian-Splats-on-visionOS demo project, and it appears that the GaussianSplatComponent is treated differently than other Components. In the function makeSplatEntity(), a GaussianSplatComponent is created and added to an Entity's components: let comp = try assembleSplatComponent(from: buffers, isLinear: isLinear) let entity = Entity() entity.components.set(comp) But when I check the entity's components collection before and after the add/set, nothing has changed. The entity still has the same children. Is the GaussianSplatComponent stored in a different location in the Entity? If so, is it a public property I can inspect? (Or is my interpretation of the components system wrong?) Here are the results of print statements before and after the set() BEFORE ADDING COMPONENT Entity number of components: 2 Component: Transform(scale: SIMD3<Float>(1.0, 1.0, 1.0), rotation: simd_quatf(real: 1.0, imag: SIMD3<Float>(0.0, 0.0, 0.0)), translation: SIMD3<Float>(0.0, 0.0, 0.0)) Component: RealityKit.SynchronizationComponent AFTER ADDING COMPONENT Entity number of components: 2 Component: Transform(scale: SIMD3<Float>(1.0, 1.0, 1.0), rotation: simd_quatf(real: 1.0, imag: SIMD3<Float>(0.0, 0.0, 0.0)), translation: SIMD3<Float>(0.0, 0.0, 0.0)) Component: RealityKit.SynchronizationComponent

I am trying to view immersive environments that are Gaussian Splats created by marble [dot] worldlabs [dot] ai. For example, here is one named "Steampunk City Night View": (Note: the low-res version is 500,000 splats, so I am just testing the first 200,000 splats). But when I try to view it at scale, nothing appears. I tried shrinking it down (scale factor of 0.00595458) and positioning it in front of me, and I could see the content, so the parsing of the PLY file is working. Example: But when I move into the model, it goes clear, also hiding virtual content behind it that is also in the splats' space (like the Mac virtual screen). Example: It is as if when I move inside the Gaussian Splats' bounding box, it sets an opacity material on the splats. This prevents me from using these Gaussian Splats as immersive environments. Is there any way to disable this behavior?

Hello everyone, ​I’m an R&D software engineer researching the Vision Pro for on-site CAD visualization and quality assurance. This requires using the headset outside in the open air in KSA, where ambient daytime temperatures easily hit 45°C (113°F). ​Before I try deploying this for field work, I have a few specific questions: ​Operating Limits: Will the headset function in 45°C ambient heat, or will it immediately trigger a thermal shutdown? ​Throttling: How aggressively does visionOS throttle performance or dim the displays when running demanding spatial apps in this heat? ​Battery: Will the external battery stop supplying power or refuse to charge under these conditions? ​Workarounds: Are there any known best practices or shading accessories for using the Vision Pro outdoors in extreme climates? ​I know the official operating limit is 30°C, so I'm looking for real-world insights from anyone who has pushed these environmental limits. ​Thank you!

May I ask if there is a complete source code project for this instructional video that needs to be learned. Work with Reality Composer Pro content in Xcode

In Reality Composer Pro 3 Beta 2 so version 3.0 (75.0.21.500.2), I’m unable to connect any model provider or complete setup for the Assistant feature. I tried the following API endpoints: OpenAI: https://api.openai.com/v1 https://api.openai.com/v1/responses https://api.openai.com/v1/chat/completions Anthropic / Claude: https://api.anthropic.com/v1 https://api.anthropic.com/v1/messages None of these allow the Assistant setup to complete. It is unclear what endpoint format Reality Composer Pro expects, or whether provider setup is currently broken in Beta 2. Ticket with a video showing the issue can be found here: FB23493842