So get a swift file and put this in it import Foundation import AVFoundation let synthesizer = AVSpeechSynthesizer() let utterance = AVSpeechUtterance(string: "Hello, testing speech synthesis on macOS.") if let voice = AVSpeechSynthesisVoice(identifier: "com.apple.voice.compact.en-GB.Daniel") { utterance.voice = voice print("Using voice: \(voice.name), \(voice.language)") } else { print("Daniel voice not found on macOS.") } synthesizer.speak(utterance) I get no speech output and this log output Error reading languages in for local resources. Error reading languages in for local resources. Using voice: Daniel, en-GB Program ended with exit code: 0 Why? and whats with "Error reading languages in for local resources." ?

My app currently captures video using an AVCaptureSession set with the AVCaptureSessionPreset1920x1080 preset. However, I'd like to update this behavior, such that video can be recorded at a range of different resolutions. There isn't a preset aligning to each desired resolution, so I thought I'd instead directly set the AVCaptureDeviceFormat. For any desired resolution, I would find the format that is closest without going under the desired resolution, and then crop it down as a post-processing step. However, what I've observed is that there can be a range of available formats for a device at each resolution, with various differing settings. Presumably there is logic within AVCaptureSession that selects a reasonable default based on all these different settings, but since I am applying the format directly, I think I don't have a way to make use of that default logic? And it is undocumented? Does this mean that the only way to select a format is to implement a comparison function that considers all different values of all different properties on AVCaptureDeviceFormat, and then sort the formats according to this comparator? If so, what if some new property is added to AVCaptureDeviceFormat in the future? The sort would not take this new property into account, and the function might select a format with some new undesired property. Are there any guarantees about what types for formats will be supported on a device? For example, can I take for granted that a '420v' format will exist at each resolution? If so I could filter the formats down only to those with this setting without risking filtering out all of the supported formats. I suspect I may be missing something obvious. Any help would be greatly appreciated!

When trying to edit some Live Photos, calling PHLivePhotoEditingContext.saveLivePhoto results in the following error: Error Domain=AVFoundationErrorDomain Code=-11800 "The operation could not be completed" UserInfo={NSLocalizedFailureReason=An unknown error occurred (-12815), NSLocalizedDescription=The operation could not be completed, NSUnderlyingError=0x300d05380 {Error Domain=NSOSStatusErrorDomain Code=-12815 "(null)"}} I was able to replicate it on my device by taking a new Live Photo. Not sure what's wrong with that one specifically, not all Live Photos replicate the issue. I've submitted FB15880825 with a sysdiagnose and a Photos Diagnostics as well. Any ideas what's going on here? It's impacting multiple customers. Thanks!

I'm trying to implement anti-spoofing in iOS app using iphone true depth front camera. I have checked the following questions still can't find a proper working solution. I trained a coreML model using 22000 depth human face images and 22000 non-human face(objects,food etc) images. The accuracy of the model is very less. When testing out with flat 2d images shown on a smartphone screen I found that I get depth map even for flat 2D images like this. Even though the image is flat how does it give the depth map for the person shown in the flat 2D picture so the model thinks that it is a real face instead of a spoofed one. I implemented depth capture by following this documentation and I made sure that I get depth map instead of disparity map https://developer.apple.com/documentation/avfoundation/additional_data_capture/capturing_photos_with_depth My next approach was to use NCNN framework to implement anti-spoofing by using the model used in the Mini-vision android anti-spoofing sample. I rewrote their library in iOS by using the objective C++ wrapper for C++ as the sample was only available for android app. And I tested by feeding 80x80 UI-Image in a open cv matrix format it's accurracy is less than the android one. How can I solve this problem.

For an upcoming update of one of my apps, I’m facing an issue: The .rate parameter of a AVAudioUnitTimePitch allows me to slow down an audio track without any issues: setting .rate to 0.7 or 0.8 results in an almost perfect playback without changing pitch. However, whenever the .rate parameter is greater than 1 (e.g. 1.1 or 1.15), I’m starting to hear audio artifacts (“flattering”) in the audio output which is not so nice (even at .overlap = 32). Intuitively, I’d’ve thought that speeding up the file should contain less artifacts than slowing it down?? I’ve tried different sample rates (44.1 kHz and 48 kHz), but same result. Grateful for any input on this 🙏

Capturing more than one display is no longer working with macOS Sequoia. We have a product that allows users to capture up to 2 displays/screens. Our application is using gstreamer which in turn is based on AVFoundation. I found a quick way to replicate the issue by just running 2 captures from separate terminals. Assuming display 1 has device index 0, and display 2 has device index 1, here are the steps: install gstreamer with brew install gstreamer Then open 2 terminal windows and launch the following processes: terminal 1 (device-index:0): gst-launch-1.0 avfvideosrc -e device-index=0 capture-screen=true ! queue ! videoscale ! video/x-raw,width=640,height=360 ! videoconvert ! osxvideosink terminal 2 (device-index:1): gst-launch-1.0 avfvideosrc -e device-index=1 capture-screen=true ! queue ! videoscale ! video/x-raw,width=640,height=360 ! videoconvert ! osxvideosink The first process that is launched will show the screen, the second process launched will not. Testing this on macOS Ventura and Sonoma works as expected, showing both screens. I submitted the same issue on Feedback Assistant: FB15900976

Issue Description When playing certain MIDI files using AVMIDIPlayer, the initial volume settings for individual tracks are being ignored during the first playback. This results in all tracks playing at the same volume level, regardless of their specified volume settings in the MIDI file. Steps to Reproduce Load a MIDI file that contains different volume settings for multiple tracks Start playback using AVMIDIPlayer Observe that all tracks play at the same volume level, ignoring their individual volume settings Current Behavior All tracks play at the same volume level during initial playback Track volume settings specified in the MIDI file are not being respected This behavior consistently occurs on first playback of affected MIDI files Expected Behavior Each track should play at its specified volume level from the beginning Volume settings in the MIDI file should be respected from the first playback Workaround I discovered that the correct volume settings can be restored by: Starting playback of the MIDI file Setting the currentPosition property to (current time - 1 second) After this operation, all tracks play at their intended volume levels However, this is not an ideal solution as it requires manual intervention and may affect the playback experience. Questions Is there a way to ensure the track volume settings are respected during the initial playback? Is this a known issue with AVMIDIPlayer? Are there any configuration settings or alternative approaches that could resolve this issue? Technical Details iOS Version: 18.1.1 (22B91) Xcode Version: 16.1 (16B40)

In my iOS App I present a QLPreviewController where I want to display a locally stored Video from the iPhone's document directory. let previewController = QLPreviewController() previewController.dataSource = self self.present(previewController, animated: true, completion: nil) func previewController(_ controller: QLPreviewController, previewItemAt index: Int) -> QLPreviewItem { let url = urlForPreview return url! as QLPreviewItem } This seems to work fine for all but one of my testflight users. He is using an iPhone 12 with iOS18.0.1. The screen becomes unresponsive. He cannot pause the video, share it or close the QLPreviewController. In his logfile I see the following error... [AVAssetTrack loadValuesAsynchronouslyForKeys:completionHandler:] invoked with unrecognized keys ( "currentVideoTrack.preferredTransform") Any ideas?.

We are experiencing thousands of crashes in our application when attempting to present the camera through a Web View. The app crashes during this process, and the crash logs point to WebCore::AVVideoCaptureSource::create WebCore::RealtimeMediaSourceCenter::getUserMediaDevices. This issue has only been observed in iOS 18.2 beta versions (beta 1 - 22C5109p, beta 2 - 22C5125e, beta 3 - 22C5131e). In iOS versions below 18.2, the functionality works and we haven't identified any correlation with specific device models. The problem seems to stem from a WebCore framework introduced in these beta releases 18.2. We kindly request a review and fix for this issue in upcoming beta releases to restore functionality. Let us know if there are any workarounds or adjustments we can implement in the interim. Thank you for your attention to this matter.

Hi. I encounter some random crashes of my camera app. After some investigations, I found that it's terminated by the system and the crash log did be generated but the information is not quite useful, and here is the log found via the Console app. Termination & Crash log "Camera not actively used; AVCaptureEventInteraction not installed": Received termination request from [osservice<com.apple.SpringBoard>:10931] on <RBSProcessPredicate <RBSProcessInstancePredicate| [app<com.juniperphoton.PhotonCam]>> with context <RBSTerminateContext| explanation:Capture Application Requirements Unmet: "Camera not actively used; AVCaptureEventInteraction not installed" reportType:CrashLog maxTerminationResistance:Interactive> The crash log exported from the device will have some common information like: It's a EXC_CRASH (SIGKILL) type with no termination reason. Exception Type: EXC_CRASH (SIGKILL) Exception Codes: 0x0000000000000000, 0x0000000000000000 Termination Reason: RUNNINGBOARD 0 It's triggered by the main thread, but it seems to be waiting for an event to process. Triggered by Thread: 0 Thread 0 name: Dispatch queue: com.apple.main-thread Thread 0 Crashed: 0 libsystem_kernel.dylib 0x1ee165788 mach_msg2_trap + 8 1 libsystem_kernel.dylib 0x1ee168e98 mach_msg2_internal + 80 2 libsystem_kernel.dylib 0x1ee168db0 mach_msg_overwrite + 424 3 libsystem_kernel.dylib 0x1ee168bfc mach_msg + 24 4 CoreFoundation 0x19cbe47f4 __CFRunLoopServiceMachPort + 160 5 CoreFoundation 0x19cbe3ea0 __CFRunLoopRun + 1212 6 CoreFoundation 0x19cc36274 CFRunLoopRunSpecific + 588 7 GraphicsServices 0x1e9d6d4c0 GSEventRunModal + 164 8 UIKitCore 0x19f783480 -[UIApplication _run] + 816 9 UIKitCore 0x19f3a9410 UIApplicationMain + 340 10 UIKitCore 0x19fae4bb0 0x19f394000 + 7670704 11 PhotonCam 0x1002e7e3c 0x1002cc000 + 114236 12 dyld 0x1c2d5ade8 start + 2724 Address size fault on the main thread Thread 0 crashed with ARM Thread State (64-bit): ... far: 0x0000000000000000 esr: 0x56000080 Address size fault I have once tried to reproduce this issue when the app is attached with debugger, and it says: Terminated due to signal 9 When the crash or termination happened, the app: No AVCaptureSession is running. The app is in the foreground and users are interacting with some functions like viewing photos or editing photos in the app. When users exit the camera view, like entering the gallery or settings, the camera session will be stopped. Both TestFlight and Debug build will have the same issue. No 3rd party crash reporter is installed(I deliberately disable it in Debug build and TestFlight build) It has adopted the LockedCameraCapture, but current it's running on the main app target(if not, my app will have a button of unlock, so I can confirm about this). Also, when it comes to the memory consumption, there is no JetsamEvent around the crash time. Device and app information Additionally, some information about the tech stack and the current state of my device and my app: iPhone 16 Pro with iOS 18.2 Beta 3. The app is a camera based app(it's PhotonCam and you can find it on the App Store), its main functionality is the camera feature using AVFoundation + Core Image + Metal to deliver camera functionality. It has adopted the Camera Control, AVCaptureEventInteraction and LockedCameraCapture features. If I remember it right, it occurs in iOS 18.1 Release build, but currently I have no such device to confirm. But in iOS 17.x the issue has never happened. Regarding to this termination, on top of my head is the "watchdog" mechanism that will terminate the process that is running on the LockedCameraCapture feature. However I can make sure that currently the app is running as the main target on the home screen. Has anybody encountered this kind of issue and has found some solutions? Thanks in advance.

I am talking about AVCaptureVideoDataOutput.recommendedVideoSettings. I found sometimes it return nil, there is my test result. hevc .mov with activeColorSpace sRGB 60FPS -> ok 120FPS -> ok hevc .mov with activeColorSpace displayP3_HLG 60FPS -> nil 120FPS -> nil h264 .mov 30FPS -> ok 60FPS -> nil 120FPS -> nil so, if you don't give a recommend setting, and you don't give a document, how does developer to use it?

I’ve built a custom media player using AVSampleBufferAudioRenderer and AVSampleBufferRenderSynchronizer, and overall, it works great! However, I’ve noticed some unusual logs popping up: Domain: NSOSStatusErrorDomain Error Codes: -16384, -16155, -16512 *That error -16512 keeps happening repeatedly for one of our users, preventing them from playing any media at all. I’ve searched around but can’t find any documentation explaining what these errors mean. Has anyone run into this issue or have any suggestions? Any help would be hugely appreciated! Thanks!

Hello, To create a test project, I want to understand how the video and audio settings would look for a destination video whose content comes from a source video. I obtained the output from the source video in the audio and video tracks as follows: let audioSettings = [ AVFormatIDKey: kAudioFormatLinearPCM, AVSampleRateKey: 44100, AVNumberOfChannelsKey: 2 ] as [String : Any] var audioOutput = AVAssetReaderTrackOutput(track: audioTrack!, outputSettings: audioSettings) // Video let videoSettings = [ kCVPixelBufferPixelFormatTypeKey: kCVPixelFormatType_32BGRA, kCVPixelBufferWidthKey: videoTrack!.naturalSize.width, kCVPixelBufferHeightKey: videoTrack!.naturalSize.height ] as [String: Any] var videoOutput = AVAssetReaderTrackOutput(track: videoTrack!, outputSettings: videoSettings) With this, I'm obtaining the CMSampleBuffer using AVAssetReader.copyNextSampleBuffer . How can I add it to the destination video? Should I use a while loop, considering I already have the AVAssetWriter set up? Something like this: while let buffer = videoOutput.copyNextSampleBuffer() { if let imgBuffer = CMSampleBufferGetImageBuffer(sampleBuffer) { let frame = imgBuffer as CVPixelBuffer let time = CMSampleBufferGetPresentationTimeStamp(sampleBuffer) adaptor.append(frame, withMediaTime: time) } } Lastly, regarding the destination video. How should the AVAssetWriterInput for audio and PixelBuffer of the destination video be set up? Provide an example, something like: let audioSettings = […] as [String: Any] Looking forward to your response.

We develop a video playback app on Apple TV which has the two following features: Its content browsing screen has installed a gesture recognizer for presses on the PlayPause Siri remote button in order to directly launch a playback. The gesture recognizer is attached to the content browsing UIViewController view. It presents its own custom playback UI with an AVPlayerLayer for the video and supports MPNowPlayingSession in order to publish current playback information and respond to remote commands. It also supports switching between fullscreen and Picture in Picture playback. Both features work fine, ie. the playback is launched when pressing the PlayPause Siri remote button and, during playback, the playback info are properly advertised on other devices and remote commands are also triggered as expected. However, when pressing the PlayPause Siri remote button while the video is playing in PiP, the "pause" remote command is sometimes triggered instead of the .playPause gesture recognizer. The issue may not occur the first time but for subsequent PlayPause presses. Navigating a bit in the app UI seems to help preventing the issue to occur. Finally, the issue only occurs if the video is playing. If the video is paused, the PlayPause Siri remote button gesture is always recognized instead of the remote command. Please note that, before using MPNowPlayingSession (and the corresponding MPRemoteCommandCenter), the app was using the default MPRemoteCommandCenter to support remote commands and the issue did not occur. We don't reproduce this issue with the Apple TV app so there's probably something we are not doing right. Has someone any clue?

We are processing videos with Core Image filters in our apps, using an AVMutableVideoComposition (for playback/preview and export). For older devices, we want to limit the resolution at which the video frames are processed for performance and memory reasons. Ideally, we would tell AVFoundation to give us video frames with a defined maximum size into our composition. We thought setting the renderSize property of the composition to the desired size would do that. However, this only changes the size of output frames, not the size of the source frames that come into the composition's handler block. For example: let composition = AVMutableVideoComposition(asset: asset, applyingCIFiltersWithHandler: { request in let input = request.sourceImage // <- this still has the video's original size // ... }) composition.renderSize = CGSize(width: 1280, heigth: 720) // for example So if the user selects a 4K video, our filter chain gets 4K input frames. Sure, we can scale them down inside our pipeline, but this costs resources and especially a lot of memory. It would be way better if AVFoundation could decode the video frames in the desired size already before passing it into the composition handler. Is there a way to tell AVFoundation to load smaller video frames?

Hello, I used AVPlayer in my project to play network movie. Most movie could play normally, but I found the sound will disappear sometimes if I play specified 4K video network stream. The video will continue playing but audio stops after video is played for a while. If I pause player and then resume, the sound will be back but disappeared again after several seconds Check AVPlayerItem status: isPlaybackLikelyToKeepUp` == true isPlaybackBufferEmpty` = false player.volume > 0 According the value above, it seems not cause by empty playback buffer or volume issue. I am so confused for this situation. Movie information Video Format : AVC Format/Info : Advanced Video Codec Format profile : High L5.1 Codec ID : avc1 Codec ID/Info : Advanced Video Coding Bit rate mode : Variable Bit rate : 100.0 Mb/s Width : 3 840 pixels Height : 2 160 pixels Display aspect ratio : 16:9 Frame rate mode : Constant Frame rate : 29.970 (30000/1001) FPS Audio Format : AAC LC Format/Info : Advanced Audio Codec Low Complexity Codec ID : mp4a-40-2 Duration : 5 min 19 s Bit rate mode : Constant Bit rate : 192 kb/s Nominal bit rate : 48.0 kb/s Channel(s) : 2 channels Channel layout : L R Sampling rate : 48.0 kHz Frame rate : 46.875 FPS (1024 SPF) Does anyone know if AVPlayer has this limitations when playing high-bitrate movie streams, and are there any solutions?

Hello, I.m deaf-blind programmer. I'm experiencing memory issues in my app. Essentially, I'm writing a video. In this output video, I get content from two sources. The first source is an already recorded video of 18 seconds (just for testing). It will be shown at the beginning of the output video. The second source is an array with photos and another array with audio buffers from AVSpeechSynthesizer.write(). The photos will be added along with the audio buffers to the output video, right after adding the 18-second video. So, in the end, the output video should be: 18-second video + array of photos as video images and, for audio, the buffers from AVSpeechSynthesizer.write(). However, my app crashes as soon as I start the first process. I'm using AVAssetWriter to write the video and AVAssetReader to read the video. Below, I'll show the code where I get the CMSampleBuffer. I'd like an example of how to add the 18-second video to the beginning of the output video. It doesn't need to be a big piece of code. Here it is: // Variables var audioReaderBuffers = [CMSAMPLEBUFFER]() var videoReaderBuffers = [(frame: CVPixelBuffer, time: CMTIME)]() // Get CMSampleBuffer of a video asset if let videoURL = videoURL { let videoAsset = AVAsset(url: videoURL) Task { let videoAssetTrack = try await videoAsset.loadTracks(withMediaType: .video).first! let audioTrack = try await videoAsset.loadTracks(withMediaType: .audio).first! let reader = try AVAssetReader(asset: videoAsset) let videoSettings = [ kCVPixelBufferPixelFormatTypeKey: kCVPixelFormatType_32BGRA, kCVPixelBufferWidthKey: videoAssetTrack.naturalSize.width, kCVPixelBufferHeightKey: videoAssetTrack.naturalSize.height ] as [String: Any] let readerVideoOutput = AVAssetReaderTrackOutput(track: videoAssetTrack, outputSettings: videoSettings) let audioSettings = [ AVFormatIDKey: kAudioFormatLinearPCM, AVSampleRateKey: 44100, AVNumberOfChannelsKey: 2 ] as [String : Any] let readerAudioOutput = AVAssetReaderTrackOutput(track: audioTrack, outputSettings: audioSettings) reader.add(readerVideoOutput) reader.add(readerAudioOutput) reader.startReading() // Video CMSampleBuffer while let sampleBuffer = readerVideoOutput.copyNextSampleBuffer() { autoreleasepool { if let imgBuffer = CMSampleBufferGetImageBuffer(sampleBuffer) { let pixBuf = imgBuffer as CVPixelBuffer let pTime = CMSampleBufferGetPresentationTimeStamp(sampleBuffer) videoReaderBuffers.append((frame: pixBuf, time: pTime)) } } } if let videoURL = videoURL { let videoAsset = AVAsset(url: videoURL) Task { let videoAssetTrack = try await videoAsset.loadTracks(withMediaType: .video).first! let audioTrack = try await videoAsset.loadTracks(withMediaType: .audio).first! let reader = try AVAssetReader(asset: videoAsset) let videoSettings = [ kCVPixelBufferPixelFormatTypeKey: kCVPixelFormatType_32BGRA, kCVPixelBufferWidthKey: videoAssetTrack.naturalSize.width, kCVPixelBufferHeightKey: videoAssetTrack.naturalSize.height ] as [String: Any] let readerVideoOutput = AVAssetReaderTrackOutput(track: videoAssetTrack, outputSettings: videoSettings) let audioSettings = [ AVFormatIDKey: kAudioFormatLinearPCM, AVSampleRateKey: 44100, AVNumberOfChannelsKey: 2 ] as [String : Any] let readerAudioOutput = AVAssetReaderTrackOutput(track: audioTrack, outputSettings: audioSettings) reader.add(readerVideoOutput) reader.add(readerAudioOutput) reader.startReading() while let sampleBuffer = readerVideoOutput.copyNextSampleBuffer() { autoreleasepool { if let imgBuffer = CMSampleBufferGetImageBuffer(sampleBuffer) { let pixBuf = imgBuffer as CVPixelBuffer let pTime = CMSampleBufferGetPresentationTimeStamp(sampleBuffer) }

I would like to take YCbCr CVPixelBuffers from AVCaptureVideoDataOutput, apply some processing in RGB space, render to an MTKView, and pass to AVAssetWriter for recording. Right now, I'm doing this all manually – deswing the incoming data if necessary, choose the right matrix to convert to RGB, apply processing, etc. I also have to convert back to YCbCr before feeding the frames to AVAssetWriter because encoding performs much better if I do. Is there any efficient, built-in way to achieve the same? I can't use AVCaptureVideoPreviewLayer, since I need to do some further processing before display. I can't use AVCaptureVideoDataOutput's videoSettings to get automatic BGRA conversion because that would lose bit depth for 10 bit video formats (and isn't available on all formats anyway). I see these Accelerate functions, but they seemingly don't use the GPU, nor do they support all the formats and bit depths I'd need. I found reference to some undocumented MTLPixelFormats that seem to do exactly what I want, but I don't want to rely on something like this unless it's explicitly endorsed. This would also incur an RGB/YCbCr conversion on every texture read and write, right? Is there anything I'm missing here?

As of iOS 18.1 being released we are having issues with our users experiencing issues with our app that relies on strobing the device torch. We have narrowed this down to being caused on devices with adaptive true-tone flash and have submitted a radar: FB15787160. The issue seems to be caused by ambient light levels. If run in a dark room, the torch strobes exactly as effectively as in previous iOS versions, if run in a light room, or outdoors, or near a window, the strobe will run for ~1s and then the torch will get stuck on for half a second or so (less frequently it gets stuck off) and then it will strobe again for ~1s and this behaviour repeats indefinitely. If we go to a darker environment, and background and then foreground the app (this is required) the issue is resolved, until moving to an area with higher ambient light levels again. We have done a lot of debugging, and also discovered that turning off "Auto-Brightness" from Settings -> Accessibility -> Display & Text Size resolves the issue. We have also viewed logs from Console.app at the time of the issue occurring and it seems to be that there are quite sporadic ambient light level readings at the time at which the issue occurs. The light readings transition from ~100 Lux to ~8000 Lux at the point that the issue starts occurring (seemingly caused by the rear sensor being affected by the torch). With "Auto-Brightness" turned off, it seems these readings stay at lower levels. This is rendering the primary use case of our app essentially useless, would be great to get to the bottom of it! We can't even really detect it in-app as I believe using SensorKit is restricted to research applications and requires a review process with Apple before accessing? Edit: It's worth noting this is also affecting other apps with strobe functionality in the exact same way

Looking to output dv video to my JVC SR-VS30 video deck. I used to be able to do this, but with most firewire stuff being deprecated, I'm not sure how to go about this. I found this old developer sample code that seems to do exactly what I'd like. Surely this could be rolled or updated for current macOS? https://developer.apple.com/library/archive/samplecode/SimpleVideoOut/Introduction/Intro.html#//apple_ref/doc/uid/DTS10000809-Intro-DontLinkElementID_2