Hello everyone, I'm working on implementing a screen sharing feature using RPSystemBroadcastPickerView and a Broadcast Upload Extension to share the entire app screen in an iOS application. The Broadcast Upload Extension is set up following Apple's ReplayKit guidelines. However, I’m encountering an issue during the broadcast startup sequence: ❗ Problem Description The Screen Broadcast UI appears as expected I tap “Start Broadcast” The countdown (3 → 2 → 1) completes Then it immediately reverts to the "Start Broadcast" screen, and screen sharing does not begin No error messages are displayed None of the extension lifecycle methods (broadcastStarted(withSetupInfo:), processSampleBuffer, etc.) are called There are no logs or crash reports, neither in the main app nor in the extension ✅ What Has Been Verified Info.plist of the Broadcast Upload Extension includes: NSExtensionPointIdentifier = com.apple.broadcast-services-upload NSExtensionPrincipalClass set correctly RPBroadcastProcessMode = RPBroadcastProcessModeSampleBuffer preferredExtension is set properly to the extension’s bundle identifier Extension is listed in the main app's build settings under "Frameworks, Libraries, and Embedded Content" ⚠️ Additional Concern We noticed that in Xcode (latest version), the Broadcast Upload Extension is listed under "Embedded Frameworks" with the setting "Embed Without Signing", and there is no option to change it to "Embed & Sign". We're wondering if this could be the reason the extension fails to launch correctly at runtime, despite being detected by the broadcast picker. ❓ Questions Has anyone faced similar issues where the broadcast never starts despite correct setup? Could the "Embed Without Signing" be causing the system to silently cancel or ignore the extension at runtime? Are there any provisioning profile or entitlement requirements specific to Broadcast Upload Extensions that might trigger this behavior silently? Any insights, suggestions, or workarounds would be greatly appreciated. Thank you in advance!

Let's consider the following code. I've created an actor that loads a list of .mp3 files from a Bundle and then makes it available for audio reproduction. Unfortunately, I'm experiencing a memory leak. At the play method. player.play() From Instruments I get _malloc_type_malloc_outlined libsystem_malloc.dylib start_wqthread libsystem_pthread.dylib private actor AudioActor { enum Failure: Error { case soundsNotLoaded([AudioPlayerClient.Sound: Error]) } enum Player { case music(AVAudioPlayer) } var players: [Sound: Player] = [:] let bundles: [Bundle] init(bundles: UncheckedSendable<[Bundle]>) { self.bundles = bundles.wrappedValue } func load(sounds: [Sound]) throws { try AVAudioSession.sharedInstance().setActive(true, options: []) var errors: [Sound: Error] = [:] for sound in sounds { guard let url = bundle.url(forResource: sound.name, withExtension: "mp3") else { continue } do { self.players[sound] = try .music(AVAudioPlayer(contentsOf: url)) } catch { errors[sound] = error } } guard errors.isEmpty else { throw Failure.soundsNotLoaded(errors) } } func play(sound: Sound, loops: Int?) throws { guard let player = self.players[sound] else { return } switch player { case let .music(player): player.numberOfLoops = loops ?? -1 player.play() } } func stop(sound: Sound) throws { guard let player = self.players[sound] else { throw Failure.soundsNotLoaded([:]) } switch player { case let .music(player): player.stop() } } }

I'm trying to write 16-bit interleaved 2-channel data captured from a LiveSwitch audio source to a AVAudioFile. The buffer and file formats match but I get a bad parameter error from the API. Does this API not support the specified format or is there some other issue? Here is the debugger output. (lldb) po audioFile.url ▿ file:///private/var/mobile/Containers/Data/Application/1EB14379-0CF2-41B6-B742-4C9A80728DB3/tmp/Heart%20Sounds%201 - _url : file:///private/var/mobile/Containers/Data/Application/1EB14379-0CF2-41B6-B742-4C9A80728DB3/tmp/Heart%20Sounds%201 - _parseInfo : nil - _baseParseInfo : nil (lldb) po error Error Domain=com.apple.coreaudio.avfaudio Code=-50 "(null)" UserInfo={failed call=ExtAudioFileWrite(_impl->_extAudioFile, buffer.frameLength, buffer.audioBufferList)} (lldb) po buffer.format <AVAudioFormat 0x302a12b20: 2 ch, 44100 Hz, Int16, interleaved> (lldb) po audioFile.fileFormat <AVAudioFormat 0x302a515e0: 2 ch, 44100 Hz, Int16, interleaved> (lldb) po buffer.frameLength 882 (lldb) po buffer.audioBufferList ▿ 0x0000000300941e60 - pointerValue : 12894608992 This code handles the details of converting the Live Switch frame into an AVAudioPCMBuffer. extension FMLiveSwitchAudioFrame { func convertedToPCMBuffer() -> AVAudioPCMBuffer { Self.convertToAVAudioPCMBuffer(from: self)! } static func convertToAVAudioPCMBuffer(from frame: FMLiveSwitchAudioFrame) -> AVAudioPCMBuffer? { // Retrieve the audio buffer and format details from the FMLiveSwitchAudioFrame guard let buffer = frame.buffer(), let format = buffer.format() as? FMLiveSwitchAudioFormat else { return nil } // Extract PCM format details from FMLiveSwitchAudioFormat let sampleRate = Double(format.clockRate()) let channelCount = AVAudioChannelCount(format.channelCount()) // Determine bytes per sample based on bit depth let bitsPerSample = 16 let bytesPerSample = bitsPerSample / 8 let bytesPerFrame = bytesPerSample * Int(channelCount) let frameLength = AVAudioFrameCount(Int(buffer.dataBuffer().length()) / bytesPerFrame) // Create an AVAudioFormat from the FMLiveSwitchAudioFormat guard let avAudioFormat = AVAudioFormat(commonFormat: .pcmFormatInt16, sampleRate: sampleRate, channels: channelCount, interleaved: true) else { return nil } // Create an AudioBufferList to wrap the existing buffer let audioBufferList = UnsafeMutablePointer<AudioBufferList>.allocate(capacity: 1) audioBufferList.pointee.mNumberBuffers = 1 audioBufferList.pointee.mBuffers.mNumberChannels = channelCount audioBufferList.pointee.mBuffers.mDataByteSize = UInt32(buffer.dataBuffer().length()) audioBufferList.pointee.mBuffers.mData = buffer.dataBuffer().data().mutableBytes // Directly use LiveSwitch buffer // Transfer ownership of the buffer to AVAudioPCMBuffer let pcmBuffer = AVAudioPCMBuffer(pcmFormat: avAudioFormat, bufferListNoCopy: audioBufferList) /* { buffer in // Ensure the buffer is freed when AVAudioPCMBuffer is deallocated buffer.deallocate() // Only call this if LiveSwitch allows manual deallocation } */ pcmBuffer?.frameLength = frameLength return pcmBuffer } } This is the handler that is invoked with every frame in order to convert it for use with AVAudioFile and optionally update a scrolling signal display on the screen. private func onRaisedFrame(obj: Any!) -> Void { // Bail out early if no one is interested in the data. guard isMonitoring else { return } // Convert LS frame to AVAudioPCMBuffer (no-copy) let frame = obj as! FMLiveSwitchAudioFrame let buffer = frame.convertedToPCMBuffer() // Hand subscribers a reference to the buffer for rendering to display. bufferPublisher?.send(buffer) // If we have and output file, store the data there, as well. guard let audioFile = self.audioFile else { return } do { try audioFile.write(from: buffer) // FIXME: This call is throwing error -50 } catch { FMLiveSwitchLog.error(withMessage: "Failed to write buffer to audio file at \(audioFile.url): \(error)") self.audioFile = nil } } This is how the audio file is being setup. static var recordingFormat: AVAudioFormat = { AVAudioFormat(commonFormat: .pcmFormatInt16, sampleRate: 44_100, channels: 2, interleaved: true)! }() let audioFile = try AVAudioFile(forWriting: outputURL, settings: Self.recordingFormat.settings)

Hello there! Is there any list of voices that are always available on iOS/iPadOS devices? It seems that AVSpeechSynthesisVoice(identifier: "com.apple.voice.compact.en-US.Samantha") is always available on all devices. I thought that AVSpeechSynthesisVoice(identifier: "com.apple.ttsbundle.siri_Nicky_en-US_compact") and AVSpeechSynthesisVoice(identifier: "com.apple.ttsbundle.siri_Aaron_en-US_compact") were available by default on certain newer devices. Is this true? I also noticed that on the same iPad where I was using those 2 voices (Nicky and Aaron) - when I updated to the iPadOS 26 beta, those voices were no longer available. Any information you can share about which voices should be reliably available on which devices would be extremely helpful for our development. Thanks so much!

After Picture-in-Picture is opened on the camera interface, the custom UI cannot be displayed. Is there any way to make the custom UI visible? If custom UI cannot be displayed, how do teleprompter-type apps in the store manage to show custom teleprompter text within the camera?

Our Final Cut Pro workflow extension built with ProExtensionHost framework uses an advanced NSPasteboardItemDataProvider system with multi-version FCPXML support (1.9, 1.10, 1.13) and proper relative path UIDs for Motion templates. We've implemented clip wrapper approach with placeholder assets and elements containing effects to enable direct timeline drag functionality. However, drag and drop from our Final Cut Pro workflow extension directly to timeline is still not working despite proper element structure in our FCPXML. Our implementation creates valid clip elements with effects applied, but Final Cut Pro timeline doesn't accept them during drag operations from our ProExtensionHost-based workflow extension. Steps to Reproduce: Create Final Cut Pro workflow extension using ProExtensionHost framework with NSPasteboardItemDataProvider implementation Generate FCPXML with proper element structure: Expected Result: Clip should be accepted by timeline and effect applied from workflow extension Actual Result: Timeline rejects drag operation from ProExtensionHost-based workflow extension Question: Are there additional requirements or ProExtensionHost API calls needed beyond standard NSPasteboardItemDataProvider for Final Cut Pro workflow extension timeline drag functionality?

We have a React website build to scan qr codes. The website is properly working for Android devices but for Iphone we see a camera glitch causing delay in scan which is unexpected. Website URL : https://react-qr-code-scanner-app.vercel.app/

I'm developing a Final Cut Pro X workflow extension that transcribes audio and creates a text output. I need to allow users to drag this text directly from my extension into FCPX's timeline as titles. Current Implementation: Using NSFilePromiseProvider as per Apple's guidelines for drag and drop Generating valid FCPXML (v1.10) with proper structure: Complete resources section with format and asset references Event and project hierarchy Asset clip with connected title elements Proper timing and duration calculations Supporting multiple pasteboard types: com.apple.finalcutpro.xml.v1-10 com.apple.finalcutpro.xml.v1-9 com.apple.finalcutpro.xml What's Working: Drag operation initiates correctly File promise provider is set up properly FCPXML generation is successful (verified content) All required pasteboard types are registered Proper logging confirms data is being requested and provided Current Pasteboard Types Offered: com.apple.NSFilePromiseItemMetaData com.apple.pasteboard.promised-file-name com.apple.pasteboard.promised-suggested-file-name com.apple.pasteboard.promised-file-content-type Apple files promise pasteboard type com.apple.pasteboard.NSFilePromiseID com.apple.pasteboard.promised-file-url com.apple.finalcutpro.xml.v1-10 com.apple.finalcutpro.xml.v1-9 com.apple.finalcutpro.xml What additional requirements or considerations are needed to make FCPX accept the dragged FCPXML content? Are there specific requirements for workflow extensions regarding drag and drop operations with titles that aren't documented? Any insights, especially from those who have implemented similar functionality in FCPX workflow extensions, would be greatly appreciated. Technical Details: macOS Version: 15.5 (24F74) FCPX Version: 11.1.1 Extension built with SwiftUI and AppKit integration Using NSFilePromiseProvider and NSPasteboardItemDataProvider Full pasteboard type support for FCPXML versions

We encounter issue with avplayer in case of EXT-X-DISCONTINUITY misalignment between audio and video produced after insertion of gaps. The initial objective is to introduce an EXT-X-DISCONTINUITY in audio playlist after some missing segments (EXT-X-GAP) which durations are aligned to video segments durations, to handle irregular audio durations. Please find below an example of corresponding video and audio playlists: video: #EXTM3U #EXT-X-VERSION:7 #EXT-X-MEDIA-SEQUENCE:872524632 #EXT-X-INDEPENDENT-SEGMENTS #EXT-X-TARGETDURATION:2 #USP-X-TIMESTAMP-MAP:MPEGTS=7096045027,LOCAL=2025-05-09T12:38:32.369100Z #EXT-X-MAP:URI="hls/StreamingBasic-video=979200.m4s" #EXT-X-PROGRAM-DATE-TIME:2025-05-09T12:38:32.369111Z #EXTINF:2.002, no desc hls/StreamingBasic-video=979200-872524632.m4s #EXTINF:2.002, no desc hls/StreamingBasic-video=979200-872524633.m4s #EXTINF:2.002, no desc hls/StreamingBasic-video=979200-872524634.m4s #EXTINF:2.002, no desc hls/StreamingBasic-video=979200-872524635.m4s #EXTINF:2.002, no desc hls/StreamingBasic-video=979200-872524636.m4s ## Media sequence discontinuity #EXT-X-GAP #EXTINF:2.002, no desc hls/StreamingBasic-video=979200-872524637.m4s ## Media sequence discontinuity #EXT-X-GAP #EXTINF:2.002, no desc hls/StreamingBasic-video=979200-872524638.m4s #EXT-X-PROGRAM-DATE-TIME:2025-05-09T12:38:46.383111Z #EXTINF:2.002, no desc hls/StreamingBasic-video=979200-872524639.m4s #EXTINF:2.002, no desc hls/StreamingBasic-video=979200-872524640.m4s audio: EXTM3U #EXT-X-VERSION:7 #EXT-X-MEDIA-SEQUENCE:872524632 #EXT-X-INDEPENDENT-SEGMENTS #EXT-X-TARGETDURATION:2 #USP-X-TIMESTAMP-MAP:MPEGTS=7096045867,LOCAL=2025-05-09T12:38:32.378400Z #EXT-X-MAP:URI="hls/StreamingBasic-audio_99500_eng=98800.m4s" #EXT-X-PROGRAM-DATE-TIME:2025-05-09T12:38:32.378444Z #EXTINF:2.0053, no desc hls/StreamingBasic-audio_99500_eng=98800-872524632.m4s #EXTINF:2.0053, no desc hls/StreamingBasic-audio_99500_eng=98800-872524633.m4s #EXTINF:2.0053, no desc hls/StreamingBasic-audio_99500_eng=98800-872524634.m4s #EXTINF:1.984, no desc hls/StreamingBasic-audio_99500_eng=98800-872524635.m4s #EXTINF:2.0053, no desc hls/StreamingBasic-audio_99500_eng=98800-872524636.m4s ## Media sequence discontinuity #EXT-X-GAP #EXTINF:2.002, no desc hls/StreamingBasic-audio_99500_eng=98800-872524637.m4s ## Media sequence discontinuity #EXT-X-GAP #EXTINF:2.002, no desc hls/StreamingBasic-audio_99500_eng=98800-872524638.m4s #EXT-X-DISCONTINUITY #EXT-X-PROGRAM-DATE-TIME:2025-05-09T12:38:46.778444Z #EXTINF:1.6213, no desc hls/StreamingBasic-audio_99500_eng=98800-872524639.m4s #EXTINF:2.0053, no desc hls/StreamingBasic-audio_99500_eng=98800-872524640.m4s In this case playback is broken with avplayer. Is it conformed to Http Live Streaming? Is it an avplayer bug? What are the guidelines to handle such gaps?

Before you post —Camera doesn't work on the Simulator— that's no longer true. I've made a solution that makes the Simulator believe there's an actual hardware device connected, allowing users to stream the macOS camera to the iOS Simulator (see for more info RocketSim's documentation: https://docs.rocketsim.app/features/hzQMSrSga7BGWvxdNVdwYs/simulator-camera-support/58tQ5jvevLNSnyUEA7VgAv) Now, it works for VNDocumentCameraViewController, but when I try opening DataScannerViewController, I directly run into: Failed to start scanning: The operation couldn’t be completed. (VisionKit.DataScannerViewController.ScanningUnavailable error 0.) My question: How does this view controller determine whether scanning is available? Is there a certain capability the available AVCaptureDevice's need to support maybe? Any direction would be helpful for me to make this work for developers, making them build apps faster!

I'm using AVFoundation to make a multi-track editor app, which can insert multiple track and clip, including scale some clip to change the speed of the clip, (also I'm not sure whether AVFoundation the best choice for me) but after making the scale with scaleTimeRange API, there is some short noise sound in play back. Also, sometimes it's fine when play AVMutableCompostion using AVPlayer with AVPlayerItem, but after exporting with AVAssetReader, will catch some short noise sounds in result file.... Not sure why. Here is the example project, which can build and run directly. https://github.com/luckysmg/daily_images/raw/refs/heads/main/TestDemo.zip

On Apple TV 4K 3rd generation, with tvOS 26 beta 2, when two HomePod 2 are paired to the device, music and movie sources with Dolby Atmos can only be listened to in stereo. dolby atmos not supported

Hi everyone! Here's what I observed so far: On device it's reproducible on iOS/iPadOS18.5, but works on iPadOS17.7. On iPhone16 iOS 18.5 simulator that I was extensively using for development it was reproducible until I reset content and settings. On iPhone 16 iOS18.4 simulator, which was also used a lot during development it still works always, so I tend to think it's 18.5 issue. Setting config.websiteDataStore = .nonPersistent() doesn't help. Cleaning WKWebsiteDataStore doesn't help. It works fine using direct URL from the embedded code (see the code below). Can someone provide some insight on how this could be fixed? Here's the code: import SwiftUI import WebKit @main struct IGVideoApp: App { var body: some Scene { WindowGroup { WebView() } } } private struct WebView: UIViewRepresentable { func makeUIView(context: Context) -> WKWebView { let config = WKWebViewConfiguration() config.allowsInlineMediaPlayback = true return .init(frame: .zero, configuration: config) } func updateUIView(_ uiView: WKWebView, context: Context) { let urlString = "https://www.instagram.com/reel/DKHFOGct3z7/?utm_source=ig_embed&amp;utm_campaign=loading" /// It works when loading from the data-instgrm-permalink URL directly // uiView.load(.init(url: .init(string: "\(urlString)")!)) /// It doesn't work whith embedding /// Note: the code part for embedding (<blockquote>...</blockquote>) is taken from my /// Instagram post (https://www.instagram.com/p/DKHFOGct3z7/) /// and stripped down. The urlString was also extracted for demonstration of direct loading. let string = """ <!doctype html> <meta charset="utf-8" /> <meta name="viewport" content="width=device-width, initial-scale=1"> <html> <head /> <body style="background-color:black; margin:0px"> <blockquote class="instagram-media" data-instgrm-captioned data-instgrm-version="14" data-instgrm-permalink="\(urlString)"> </blockquote> <script async src="https://www.instagram.com/embed.js"></script> </body> </html> """ uiView.loadHTMLString(string, baseURL: .init(string: "https://www.instagram.com")) } }

I'm capturing video stream from GoPro camera (via UDP MPEG-TS packets) but unable to play in iOS app. can someone provide a source code to decode MPEG-TS data to CMSampleBuffer.

(Note: this is part 1 of a 3 part posting. See Part 2 or Part 3) At WWDC25 we launched a new type of Lab event for the developer community - Group Labs. A Group Lab is a panel Q&A designed for a large audience of developers. Group Labs are a unique opportunity for the community to submit questions directly to a panel of Apple engineers and designers. Here are the highlights from the WWDC25 Group Lab for Camera & Photos. WWDC25 Camera & Photos group lab ran for one hour at 6 PM PST on Tuesday June 10th, 2025 Introductory kick-off questions Question 1 Tell us a little about the new AVFoundation Capture APIs we've made available in the new iOS 26 developer preview? Cinematic Capture API (strong/weak focus, tracking focus)(scene monitoring)(simulated aperture)(dog/cat heads/groupIDs) Camera Controls and AirPod Stem Clicks Spatial Audio and Studio Quality AirPod Mics in Camera Lens Smudge Detection Exposure and Focus Rect of Interest Question 2 I built QR code scanning into my app, but on newer iPhones I have to hold the phone very far away from the QR code, otherwise the image is blurry and it doesn't scan. Why is this happening and how can I fix it? Every year, the cameras get better and better as we push the state of the art on iPhone photography and videography. This sometimes results in changes to the characteristics of the lenses. min focus distance newer phones have multiple lenses automatic switching behavior Use virtual device like the builtInDualWide or built in Triple, rather than just the builtInWide Set the videoZoomFactor to 2. You're done. Question 3 Last year, we saw some exciting new APIs introduced in AVFoundation in the health space. With Constant Color photography, developers can take pictures that have constant color regardless of ambient lighting. There are some further advancements this year. Davide, could you tell us about them? constant color photography is mean to remove the "tone mapping" applied to photograph captured with camera app, usually incldsuing artistic intent, and instead try to be a close as possible to the real color of the scene, regardless of the illumination constant color images could be captured in HEIF anf JPEG laste year. this year we are adding Support for the DICOM medical imaging photo format. It is a fomrat used by the health industry to store images related to medical subjects like MRI, skin problems, xray and so on. It's writable and also readable format on all OS26, supported through AVCapturePhotoOutput APIs and through the coregraphics api. for coregrapphics there is a new DICOM entry in the property dictionary which includes all the dicom availbale and defined propertie in a file. finder will also display all those in the info panel (Address why a developer would want to use it) - not for regualr picture taking apps. for those HEIF and JPEG are the preferred delivery format. use dicom if your app produces output that are health related, that you can also share with health providers or your doctors Main session developer questions Question 1 LiDAR vs. Dual Camera depth generation: Which resolution does the LiDAR sensor natively have (iPhone 16 Pro) and when to prefer LiDAR over Dual Camera? Both report formats with output resolutions (we don't advertise sensor resolution) Lidar vs Dual, etc: Lidar: Best for absolute depth, real world scale and computer vision Dual, etc: relative, disparity-based, less power, photo effects Also see: 2022 WWDC session "Discovery advancements in iOS camera capture: Depth, focus and multitasking" Question 2 Can true depth and lidar camera run at 60fps? Lidar can do 30fps (edited) Front true depth can do 60fps. Question 3 What’s the first class way to use PhotoKit to reimplement a high performance photo grid? We’ve been using a LazyVGrid and the photos caching manager, but are never able to hit the holy trinity (60hz, efficient memory footprint, minimal flashes of placeholder/empty cells) use the PHCachingImageManager to get media content delivered before you need to display it specify the size you need for grid sized display set the options PHVideoRequestOptionsDeliveryModeFastFormat, PHImageRequestOptionsDeliveryModeFastFormat and PHImageRequestOptionsResizeModeFast Question 4 For rending live preview of video stream, Is there performance overhead from using async and Swift UI for image updates vs UIViewRepresentable + AVCaptureVideoPreviewLayer.self? AVCaptureVideoPreviewLayer is the most efficient display path Use VDO + AVSampleBufferDisplayLayer if you need to modify the image data Swift UI image is optimized for static image content Question 5 Is there a way to configure the AVFoundation BuiltInLiDarDepthCamera mode to provide a depth map as accurate as ARKit at close range? The AVCaptureDepthDataOutput supports filtering that reduces noise and fills in invalid values. Consider using this for smoother depth maps Question 6 Pyramid-based photo editing in core image (such as adobe camera raw highlights and shadows)? First off you may want to look a the builtin filter called CIHighlightShadowAdjust Also the noise reduction in the CIRawFilter uses a pyramid-based algorithm. You can also write your own pyramid-based algorithms by taking an input image: down sample it by two multiply times using imageByApplyingAffineTransform apply additional CIKernels to each downsampled image as needed. use a custom CIKernel to combine the results. Question 7 Is the best way to integrate an in-app camera for a “non-camera” app UIImagePickerController? Yes, UIImagePickerController provides system-provided UI for capturing photos and movies. Question 8 Hello, my question is on Deferred Photo Processing? Say I have a photo capture app that adds a CIFilter to the capture. How can I take advantage of Deferred Photo Processing? Since I don’t know how to detect when the deferred captured photo is ready CIFilter can be called on the final at that point Photo will have to be re-inserted into the Photo library as adjustment Question 9 For shipping photo style assets in the app that need transparency what is the best format to use? JPEG2000? will moving to this save a lot of space comapred to PNG or other options? If you want lossless compression PNG is good and supports unpremutiplied alpha If you want lossy compression HEIF supports premutiplied or unpremutiplied alpha (Note: this is part 1 of a 3 part posting. See Part 2 or Part 3)

Please consider adding the ability to programatically download Premium and Enhanced voices. At the moment it is extremely inconvenient for our users, as they have to navigate to settings themselves to download voices. Our app relies heavily on SpeechSynthesis integration, and it would greatly benefit from this feature. FB16307193

(Note: this is part 2 of a 3 part posting. See Part 1 or Part 3) At WWDC25 we launched a new type of Lab event for the developer community - Group Labs. A Group Lab is a panel Q&A designed for a large audience of developers. Group Labs are a unique opportunity for the community to submit questions directly to a panel of Apple engineers and designers. Here are the highlights from the WWDC25 Group Lab for Camera & Photos. WWDC25 Camera & Photos group lab ran for one hour at 6 PM PST on Tuesday June 10th, 2025 Question 10 Can we directly integrate auto-capture triggers (e.g., when image is steady or text is detected) using Vision and AVFoundation? Yes apps can use AVCaptureSession's VDO + AVCapturePhotoOutput, run vision on VDO buffers and capture photo when certain scene or text is detected. Just to be careful to run Vision on VDO buffers async so it doesn't cause frame drops. Question 11 What Camera or Photos framework features support working with images from external media, like connected cameras or SD cards? Any best practices? The ImageCaptureCore framework supports camera devices, memory cards, scanners read and write, where supported check out the docs to see how to browse connected devices, folders, files, etc. Question 12 Hi Brad, to follow up on your SwiftUI cautionary note: using AVCaptureVideoPreview inside a UIViewRepresentable, is okay, right? Thanks all for the great info! Yes, this is totally fine. AppKit or UIKit views inside appropriate SwiftUI representables should be equivalent performance Question 13 What’s the “right” way to transition media in my photos app between HDR modes? When I’m in a one-up view, we use HDR, but in other contexts (like thumbnail) we don’t want HDR. Is there a nice way to tone map? There’s a suite of new System Tone Mapper APIs in this years’ OSes CoreImage ImageKit CoreAnimation, CoreGraphics For example: CoreImage: new CISystemToneMap filter. CoreAnimation: layer.preferredDynamicRange = CADynamicRangeConstrainedHigh Using image views (NSImageView/UIImageView/SwiftUI Image/CALayer) support animations on preferredDynamicRange Can go from high to constrained to standard Tone mapping is provided by the system (CISystemToneMap for controllable example) Question 14 What is your recommendation to preprocess and upscale your depth map in order to render a realistic portrait mode image? One way to do this: the CIEdgePreserveUpsample CIFilter can be use to upsample a lower resolution depth map by using a higher resolution RGB image as a guide. Question 15 For buffering frames for later processing from real-time camera output should we prefer a AVSampleBufferDisplayLayer centered approach or AVCaptureVideoDataOutputSampleBufferDelegate centered approach? When would we use each? AVSampleBufferDisplayLayer and AVCaptureVideoDataOutputSampleBufferDelegate are used hand in hand for custom camera preview. For buffering for later processing, ensure you make copies of VDO buffers to not drop frames from the output Question 16 Hello, my question is on Deferred Photo Processing? Say I have a photo capture app that adds a CIFilter to the capture. How can I take advantage of Deferred Photo Processing? Since I don’t know how to detect when the deferred captured photo is ready CIFilter can be called on the final at that point Photo will have to be re-inserted into the Photo library as adjustment Question 17 Is digital zoom (e.g., 1.5x) before taking a photo the same as cropping the photo afterward? digital zoom upscales the image to output dimensions and cropping will yield a smaller output image while digital zoom will crop, it also upscales Question 18 How do you design camera interfaces that work for both casual users and photography enthusiasts? Progressive disclosure: Put the most common controls up front, and make it easy for pros to drill down. Sensible Defaults: Choose defaults that work well for casual users, but allow those defaults to be modified for photography enthusiasts A good philosophy is: Keep the simple things easy, make the hard things possible Question 19 Recent iPhone models introduced macro mode which automatically switch between lenses to take into account of the focal distance difference. Is there official API to implement this, or should I implement them myself using LiDAR values. Using builtInTripleCamera and builtInDualWideCamera will automatically switch to macro when available Question 20 a couple of years ago at WWDC, the option of replacing a camera with a virtual camera was mentioned. How does one do that - make the “physical” camera effectively disappear, so only the virtual camera is accessible to the user? You can't prevent the built-in camera from being available to the user Question 21 Can developers now integrate custom Core ML models with Vision for on-device photo analysis more seamlessly? Yes they can, use CoreMLRequest , provide their model container Been supported for a while (iOS 18/macOS 15) For more details go to Machine Learning & AI group lab Thursday use smaller images for better performance Question 22 What would you recommend for capture of the new immersive and spatial formats? To capture Spatial Video use AVCaptureMovieFileOutput’s spatialVideoCaptureEnabled property Not all device formats support spatial capture, check AVCaptureDevice.activeFormat.spatialVideoCaptureSupported See WWDC 2024 talk “Build compelling spatial photo and video experiences” for more details Question 23 You mentioned JPEG-XL. What is the current status of support on iOS and macOS for encoding and decoding? For decoding, we support JPEG-XL files in all our OSes, regular SDR files, as well as ISO HDR files. For encoding, we only support JPEG-XL for ProRAW DNG capture in the Camera app or via third-party AVFoundation APIs. If you have any requests for improvement or new features related to JPEG-XL, please file a Feedback request using the Feedback Assistant. (Note: this is part 2 of a 3 part posting. See Part 1 or Part 3)

Hello. My team and I think we have an issue where our app is asked to gracefully shutdown with a following SIGTERM. As we’ve learned, this is normally not an issue. However, it seems to also be happening while our app (an audio streamer) is actively playing in the background. From our perspective, starting playback is indicating strong user intent. We understand that there can be extreme circumstances where the background audio needs to be killed, but should it be considered part of normal operation? We hope that’s not the case. All we see in the logs is the graceful shutdown request. We can say with high certainty that it’s happening though, as we know that playback is running within 0.5 seconds of the crash, without any other tracked user interaction. Can you verify if this is intended behavior, and if there’s something we can do about it from our end. From our logs it doesn’t look to be related to either memory usage within the app, or the system as a whole. Best, John

Use case: When SharePlay -ing a fully immersive 3D scene (e.g. a virtual stage), I would like to shine lights on specific Personas, so they show up brighter when someone in the scene is recording the feed (think a camera person in the scene wearing Vision Pro). Note: This spotlight effect only needs to render in the camera person's headset and does NOT need to be journaled or shared. Before I dive into this, my technical question: Can environmental and/or scene lighting affect Persona brightness in a SharePlay? If not, is there a way to programmatically make Personas "brighter" when recording? My screen recordings always seem to turn out darker than what's rendered in environment, and manually adjusting the contrast tends to blow out the details in a Persona's face (especially in visionOS 26).

Hello, I'm observing an intermittent memory leak being reported in the iOS Simulator when initializing and starting an AVAudioEngine. Even with minimal setup—just attaching a single AVAudioPlayerNode and connecting it to the mainMixerNode—Xcode's memory diagnostics and Instruments sometimes flag a leak. Here is a simplified version of the code I'm using: // This function is called when the user taps a button in the view controller: #import "ViewController.h" @interface ViewController () @end @implementation ViewController - (void)viewDidLoad { [super viewDidLoad]; } - (IBAction)myButtonAction:(id)sender { NSLog(@"Test"); soundCreate(); } @end // media.m static AVAudioEngine *audioEngine = nil; void soundCreate(void) { if (audioEngine != nil) return; [[AVAudioSession sharedInstance] setCategory:AVAudioSessionCategoryAmbient error:nil]; [[AVAudioSession sharedInstance] setActive:YES error:nil]; audioEngine = [[AVAudioEngine alloc] init]; AVAudioPlayerNode* playerNode = [[AVAudioPlayerNode alloc] init]; [audioEngine attachNode:playerNode]; [audioEngine connect:playerNode to:(AVAudioNode *)[audioEngine mainMixerNode] format:nil]; [audioEngine startAndReturnError:nil]; } In the memory leak report, the following call stack is repeated, seemingly in a loop: ListenerMap::InsertEvent(XAudioUnitEvent const&, ListenerBinding*) AudioToolboxCore ListenerMap::AddParameter(AUListener*, void*, XAudioUnitEvent const&) AudioToolboxCore AUListenerAddParameter AudioToolboxCore addOrRemoveParameterListeners(OpaqueAudioComponentInstance*, AUListenerBase*, AUParameterTree*, bool) AudioToolboxCore 0x180178ddf