I’m building a professional camera app where users can customize the video recording format and color grading. In the func captureOutput(_ output: AVCaptureOutput, didOutput sampleBuffer: CMSampleBuffer, from connection: AVCaptureConnection) method, I handle video frames and use Metal for real-time color grading. This works well when device.activeColorSpace is sRGB or P3, and the results are great. However, when the color space is HLG_BT2020 or appleLog, the MTKTextureLoader.newTexture(cgImage: cgImage, options: options) method throws an error. After researching, I found that the video frame in these color spaces has a bit-per-channel (bpc) greater than 8 after being converted to CGImage, causing the texture creation to fail. I tried converting the CGImage to a lower bpc to successfully create the texture, but the final output image is garbled and not as expected. Is there a solution to this issue?

According to the documentation (https://developer.apple.com/documentation/avfoundation/avplayeritem/externalmetadata), AVPlayerItem should have an externalMetadata property. However it does not appear to be visible to my app. When I try, I get: Value of type 'AVPlayerItem' has no member 'externalMetadata' Documentation states iOS 12.2+; I am building with a minimum deployment target of iOS 18. Code snippet: import Foundation import AVFoundation /// ... in function ... // create metadata as described in https://developer.apple.com/videos/play/wwdc2022/110338 var title = AVMutableMetadataItem() title.identifier = .commonIdentifierAlbumName title.value = "My Title" as NSString? title.extendedLanguageTag = "und" var playerItem = await AVPlayerItem(asset: composition) playerItem.externalMetadata = [ title ]

I have a memory leak, when using AVAudioPlayer. I managed to narrow down the issue into a very simple app, which code I paste in at the end. The memory leak start immediately when I start playing sound, but only in the emylator. On the real iPhone there is no memory leak. The memory leak on the Simulator looks like this: import SwiftUI import AVFoundation struct ContentView_Audio: View { var sound: AVAudioPlayer? init() { guard let path = Bundle.main.path(forResource: "cd201", ofType: "mp3") else { return } let url = URL(fileURLWithPath: path) do { try AVAudioSession.sharedInstance().setCategory(.playback, mode: .default, options: [.mixWithOthers]) } catch { return } do { try AVAudioSession.sharedInstance().setActive(true) } catch { return } do { sound = try AVAudioPlayer(contentsOf: url) } catch { return } } var body: some View { HStack { Button { playSound() } label: { ZStack { Circle() .fill(.mint.opacity(0.3)) .frame(width: 44, height: 44) .shadow(radius: 8) Image(systemName: "play.fill") .resizable() .frame(width: 20, height: 20) } } .padding() Button { stopSound() } label: { ZStack { Circle() .fill(.mint.opacity(0.3)) .frame(width: 44, height: 44) .shadow(radius: 8) Image(systemName: "stop.fill") .resizable() .frame(width: 20, height: 20) } } .padding() } } private func playSound() { guard sound != nil else { return } sound?.volume = 1 // sound?.numberOfLoops = -1 sound?.play() } func stopSound() { sound?.stop() } }

I'm able to get text to speech to audio file using the following code for iOS 12 iPhone 8 to create a car file: audioFile = try AVAudioFile( forWriting: saveToURL, settings: pcmBuffer.format.settings, commonFormat: .pcmFormatInt16, interleaved: false) where pcmBuffer.format.settings is: [AVAudioFileTypeKey: kAudioFileMP3Type, AVSampleRateKey: 48000, AVEncoderBitRateKey: 128000, AVNumberOfChannelsKey: 2, AVFormatIDKey: kAudioFormatLinearPCM] However, this code does not work when I run the app in iOS 18 on iPhone 13 Pro Max. The audio file is created, but it doesn't sound right. It has a lot of static and it seems the speech is very low pitch. Can anyone give me a hint or an answer?

I'm encountering errors while using AVAudioEngine with voice processing enabled (setVoiceProcessingEnabled(true)) in scenarios where the input and output audio devices are not the same. This issue arises specifically with mismatched devices, preventing the application from functioning as expected. Works: Paired devices (e.g., MacBook Pro mic → MacBook Pro speakers) Fails: Mismatched devices (e.g., AirPods mic → MacBook Pro speakers) When using paired input and output devices: The setup works as expected. Example: MacBook Pro microphone → MacBook Pro speakers. When using mismatched devices: AVAudioEngine setup fails during aggregate device construction. Example: AirPods microphone → MacBook Pro speakers. Error logs indicate a channel count mismatch. Here are the partial logs. Due to the content limit, I cannot post the entire logs. AUVPAggregate.cpp:1000 client-side input and output formats do not match (err=-10875) AUVPAggregate.cpp:1036 err=-10875 AVAEInternal.h:109 [AVAudioEngineGraph.mm:1344:Initialize: (err = PerformCommand(*outputNode, kAUInitialize, NULL, 0)): error -10875 AggregateDevice.mm:329 Failed expectation of constructed aggregate (312): mInput.streamChannelCounts == inputStreamChannelCounts AggregateDevice.mm:331 Failed expectation of constructed aggregate (312): mInput.totalChannelCount == std::accumulate(inputStreamChannelCounts.begin(), inputStreamChannelCounts.end(), 0U) AggregateDevice.mm:182 error fetching default pair AggregateDevice.mm:329 Failed expectation of constructed aggregate (336): mInput.streamChannelCounts == inputStreamChannelCounts AggregateDevice.mm:331 Failed expectation of constructed aggregate (336): mInput.totalChannelCount == std::accumulate(inputStreamChannelCounts.begin(), inputStreamChannelCounts.end(), 0U) AUHAL.cpp:1782 ca_verify_noerr: [AudioDeviceSetProperty(mDeviceID, NULL, 0, isInput, kAudioDevicePropertyIOProcStreamUsage, theSize, theStreamUsage), 560227702] AudioHardware-mac-imp.cpp:3484 AudioDeviceSetProperty: no device with given ID AUHAL.cpp:1782 ca_verify_noerr: [AudioDeviceSetProperty(mDeviceID, NULL, 0, isInput, kAudioDevicePropertyIOProcStreamUsage, theSize, theStreamUsage), 560227702] AggregateDevice.mm:182 error fetching default pair AggregateDevice.mm:329 Failed expectation of constructed aggregate (348): mInput.streamChannelCounts == inputStreamChannelCounts AggregateDevice.mm:331 Failed expectation of constructed aggregate (348): mInput.totalChannelCount == std::accumulate(inputStreamChannelCounts.begin(), inputStreamChannelCounts.end(), 0U) Is it possible to use voice processing with different input/output devices? If yes, are there any specific configurations required to handle mismatched devices? How can we resolve channel count mismatch errors during aggregate device construction? Are there settings or API adjustments to enforce compatibility between input/output devices? Are there any workarounds or alternative approaches to achieve voice processing functionality with mismatched devices? For instance, can we force an intermediate channel configuration or downmix input/output formats?

Hi all, with my app ScreenFloat, you can record your screen, along with system- and microphone audio. Those two audio feeds are recorded into separate audio tracks in order to individually remove or edit them later on. Now, these recordings you create with ScreenFloat can be drag-and-dropped to other apps instantly. So far, so good, but some apps, like Slack, or VLC, or even websites like YouTube, do not play back multiple audio tracks, just one. So what I'm trying to do is, on dragging the video recording file out of ScreenFloat, instantly baking together the two individual audio tracks into one, and offering that new file as the drag and drop file, so that all audio is played in the target app. But it's slow. I mean, it's actually quite fast, but for drag and drop, it's slow. My approach is this: "Bake together" the two audio tracks into a one-track m4a audio file using AVMutableAudioMix and AVAssetExportSession Take the video track, add the new audio file as an audio track to it, and render that out using AVAssetExportSession For a quick benchmark, a 3'40'' movie, step 1 takes ~1.7 seconds, and step two adds another ~1.5 seconds, so we're at ~3.2 seconds. That's an eternity for a drag and drop, where the user might cancel if there's no immediate feedback. I could also do it in one step, but then I couldn't use the AV*Passthrough preset, and that makes it take around 32 seconds then, because I assume it touches the video data (which is unnecessary in this case, so I think the two-step approach here is the fastest). So, my question is, is there a faster way? The best idea I can come up with right now is, when initially recording the screen with system- and microphone audio as separate tracks, to also record both of them into a third, muted, "hidden" track I could use later on, basically eliminating the need for step one and just ripping the two single audio tracks out of the movie and only have the video and the "hidden" track (then unmuted), but I'd still have a ~1.5 second delay there. Also, there's the processing and data overhead (basically doubling the movie's audio data). All this would be great for an export operation (where one expects it to take a little time), but for a drag-and-drop operation, it's not ideal. I've discarded the idea of doing a promise file drag, because many apps do not accept those, and I want to keep wide compatibility with all sorts of apps. I'd appreciate any ideas or pointers. Thank you kindly, Matthias

I am trying to stream audio from local filesystem. For that, I am trying to use an AVAssetResourceLoaderDelegate for an AVURLAsset. However, Content-Length is not known at the start. To overcome this, I tried several methods: Set content length as nil, in the AVAssetResourceLoadingContentInformationRequest Set content length to -1, in the ContentInformationRequest Both of these cause the AVPlayerItem to fail with an error. I also tried setting Content-Length as INT_MAX, and setting a renewalDate = Date(timeIntervalSinceNow: 5). However, that seems to be buggy. Even after updating the Content-Length to the correct value (e.g. X bytes) and finishing that loading request, the resource loader keeps getting requests with requestedOffset = X with dataRequest.requestsAllDataToEndOfResource = true. These requests keep coming indefinitely, and as a result it seems that the next item in the queue does not get played. Also, .AVPlayerItemDidPlayToEndTime notification does not get called. I wanted to check if this is an expected behavior or is there a bug in this implementation. Also, what is the recommended way to stream audio of unknown initial length from local file system? Thanks!

FairPlay-Protected HLS Files Not Transferred via Quick Start I have an iOS app that downloads HLS files, which are protected by FairPlay. These files are stored locally, and their locations are managed using Core Data. When playing these tracks, I use AVURLAsset to access the stored file paths. Recently, a client upgraded to a new iPhone and used Quick Start to transfer data from his old device. While all other app data was successfully transferred, including Core Data records and UserDefaults, the actual HLS files were missing. As a result, the app retained metadata about the downloaded content, but the files themselves were gone, causing playback failures. Does Quick Start exclude certain types of locally stored files, especially DRM-protected HLS downloads, or is the issue related to how FairPlay-protected content is handled during the transfer of locally stored files?

Since iOS and tvOS 18, CMCD can now be automatically sent by AVPlayer (https://developer.apple.com/streaming/Whats-new-HLS.pdf). However, after enabling CMCD, our streams occasionally fail with the following error: CoreMediaErrorDomain Error -17383 This issue appears to affect only DRM-protected (FairPlay) streams so far. We activate CMCD via the resource loader of an AVURLAsset, before assigning the item to an AVPlayer. Unfortunately, we haven’t found a reliable way to reproduce the issue, and we’ve been unable to gather any useful diagnostic information. Has anyone else observed this behavior when enabling CMCD on FairPlay streams?

I noticed that AVSampleBufferDisplayLayerContentLayer is not released when the AVSampleBufferDisplayLayer is removed and released. It is possible to reproduce the issue with the simple code: import AVFoundation import UIKit class ViewController: UIViewController { var displayBufferLayer: AVSampleBufferDisplayLayer? override func viewDidLoad() { super.viewDidLoad() let displayBufferLayer = AVSampleBufferDisplayLayer() displayBufferLayer.videoGravity = .resizeAspectFill displayBufferLayer.frame = view.bounds view.layer.insertSublayer(displayBufferLayer, at: 0) self.displayBufferLayer = displayBufferLayer DispatchQueue.main.asyncAfter(deadline: .now() + 1) { self.displayBufferLayer?.flush() self.displayBufferLayer?.removeFromSuperlayer() self.displayBufferLayer = nil } } } In my real project I have mutliple AVSampleBufferDisplayLayer created and removed in different view controllers, this is problematic because the amount of leaked AVSampleBufferDisplayLayerContentLayer keeps increasing. I wonder that maybe I should use a pool of AVSampleBufferDisplayLayer and reuse them, however I'm slightly afraid that this can also lead to strange bugs. Edit: It doesn't cause leaks on iOS 18 device but leaks on iPad Pro, iOS 17.5.1

Hi, I am looking for a good way to play sounds at a high frequency. At the moment I am using the AVAudioEngine, and create a couple AVAudioPlayerNode and for each sound I need to play I create a AVAudioPCMBuffer. When the app needs to play a sound, I get the correct AVAudioPCMBuffer for the sound and use the first available AVAudioPlayerNode and feed it to the buffer. The timing for a metronome app has to be very precise because if it's of by about 16ms the user can hear that it is not playing had the right interval. For low speeds this is working without any problems, but at high speeds it is getting worse. Maybe anyone has an idea on how I can improve my method. Its a Plugin for Flutter. import AVFoundation class FastSoundPlayer { private var audioPlayers: [SoundPlayer?] = [] private var sounds: [String: Sound] = [:] private var engine = AVAudioEngine() let session = AVAudioSession.sharedInstance() init() { do { try session.setCategory(AVAudioSession.Category.playback, mode: AVAudioSession.Mode.default, options: [AVAudioSession.CategoryOptions.mixWithOthers]) try session.setActive(true) createSoundPlayers(count: 20) try engine.start() } catch { print("Error starting audio engine: \(error.localizedDescription)") } } // Selector method to handle applicationDidBecomeActiveNotification func applicationDidBecomeActive() { // Reinitialize AVAudioEngine and reattach all nodes do { engine.reset() objc_sync_enter(audioPlayers) audioPlayers.removeAll() createSoundPlayers(count: 20) objc_sync_exit(audioPlayers) try engine.start() } catch { print("Error starting audio engine: \(error.localizedDescription)") } } func createSoundPlayers(count: Int) { for _ in 0..<count { let player = SoundPlayer() engine.attach(player.player) engine.connect(player.player, to: engine.mainMixerNode, format: nil) audioPlayers.append(player) } } func load(sound: Data, name: String) { let sound = Sound(soundData: sound) sounds[name] = sound } func play(name: String) { if !engine.isRunning { applicationDidBecomeActive() } guard let sound = sounds[name] else { print("Sound not found") return } if let player = getAvailablePlayer() { player.play(sound: sound) } } func getAvailablePlayer() -> SoundPlayer? { for player in audioPlayers { if !player!.isPlaying { return player } } return nil } } class SoundPlayer { let player = AVAudioPlayerNode() var isPlaying = false init() { player.volume = 1.0 } func play(sound: Sound) { player.scheduleBuffer(sound.sound!, at: nil, options: .interrupts, completionCallbackType: .dataPlayedBack) { _ in self.complete() } if (player.engine != nil && player.engine!.isRunning) { player.play() isPlaying = true } } func complete() { isPlaying = false } } class Sound { var sound: AVAudioPCMBuffer? init(soundData: Data) { do { let temporaryURL = FileManager.default.temporaryDirectory.appendingPathComponent("tempSound.wav") try soundData.write(to: temporaryURL) // Create AVAudioFile from the temporary file URL let audioFile = try AVAudioFile(forReading: temporaryURL) // Define the format for the PCM buffer (44100Hz, stereo) let format = AVAudioFormat(commonFormat: .pcmFormatInt16, sampleRate: 44100, channels: 2, interleaved: false) // Create AVAudioPCMBuffer guard let pcmBuffer = AVAudioPCMBuffer(pcmFormat: format!, frameCapacity: AVAudioFrameCount(audioFile.length)) else { // Failed to create PCM buffer self.sound = nil return } // Read audio file into PCM buffer try audioFile.read(into: pcmBuffer) // Assign the created AVAudioPCMBuffer to the sound property self.sound = pcmBuffer } catch { print("Error loading sound file: \(error.localizedDescription)") self.sound = nil } } } Thanks!

In the past, when using Lightning, many external devices had to go through MFi certification. However, since the iPhone 15 switched from Lightning to USB-C, is MFi certification still required? Our company has developed several UVC devices, and we have confirmed that iPads can read frames from external cameras through the external device type in AVFoundation. However, this is not supported on iPhones. We are currently exploring feasible ways to enable UVC device support on iPhones. Is MFi certification the only option? If so, is the MFi certification process for USB-C the same as it was for Lightning? Does it still require purchasing an MFi chip and manufacturing specially designed USB-C cables?

I'm working on an application that uses the iPhone camera for scientific purposes - and, as a result would like to receive video in as unprocessed format as possible. In particular, I'm interested in getting pixel buffers that contain pretty much the bayer data as the sensor sees it - with the minimum processing of color possible. Currently we configure the AVCaptureDevice to fix the focus and exposure, use a low ISO with no gain and set the white balance gains to 1. AVCaptureVideoDataOutput is using 32BGRA. What I'd like to do is remove any additional color and brightness processing such that the data is effectively processed with a linear transfer function (i.e. gamma function is 1). I thought that this might be down to using the AVCaptureDevice activeColorSpace - we currently use P3_D65 for this. But there only seems to be a few choices (e.g. sRGB, HLG_BT2020) all of which I think affect the gamma. So: is it possible to control or specify the gamma / transfer function when using CaptureVideoDelegate? if not, does one of the color space settings have a defined gamma function that I can effectively reverse it from the pixel data without losing too much information? or is there a better way to capture video-ish speed images (15-30fps) from the camera sensor that skips processing like this? Many thanks for any suggestions.

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() } } }

If you have two video segments, one HDR and one SDR, next to one another in a composition, the SDR one appears dark, since its max luminance will be lower than the max luminance of the HDR clip. iMovie handles this well by (reverse) tone mapping the SDR content to make it look HDR in an HDR composition. This is what I want to achieve. I've looked into algorithms to do this, and the best that I can find is the conversion from RGB to YCbCr described in Table 4 of BT.2020, followed by conversion method A (Section 4.2, table 4) of BT.2446-1. I have these implemented in a Core Image kernel, available at this repo. The issue that I'm seeing is that the colors are still much too hot, and while there are frames that appear close to properly tone mapped, it doesn't come close to the accuracy of iMovie's approach. If someone is an expert in colorimetry, etc., I'd really appreciate a breakdown of what I'm doing wrong here. To be specific: Assuming non-constant luminance for the YCbCr conversion Using a Metal Core Image kernel for the actual tone mapping The video composition uses Core Image filters directly The video composition is using BT.2020 colorimetry and the PQ transfer function. Below is a comparison of two screenshots, one using an unaltered HDR asset, and the other using the same asset transcoded to SDR with QuickTime, and reverse tone mapped to HDR for playback. Is there something I'm missing?

I have DNG files that I want to open and show as EDR content in my app. It seems like the DNG files should have enough per pixel information to show more colors that Display P3 but whenever I load the images using CIRawFilter and then inspect the outputImage color space it is always "DisplayP3", not something like "ITU-R BT.2100 PQ" there doesn't seem to be any way to make it load with a different color space for displaying EDR images. Does this make sense for DNG files, it seems like it should? If I open the same file using CIImage with the expandToHDR option e.g. CIImage(contentsOf: rawURL, options: [.expandToHDR: true]) then it does have the desired EDR color space, but then I don't get any of the properties that are available via the CIRAWFilter class to manipulate the data. Basically I just want to be able to open the DNG file via CIRAWFilter and then display it in my SwiftUI app as an EDR image by adding the allowedDynamicRange(.high) property. Image("my-dng-image").allowedDynamicRange(.high) Or do DNG files (just RAW not ProRAW) not contain enough information to be displayed as EDR images, seems like they should.

I'm using an iPhone 15 Pro, which has switched from Lightning to USB Type-C. My iOS version is 18.3. According to Apple's documentation, AVCaptureDevice.DeviceType should support external device types. 🔗 Apple's Official Documentation: https://developer.apple.com/documentation/avfoundation/avcapturedevice/devicetype-swift.struct/external The documentation clearly states that iPadOS 17.0+ and iOS 17.0+ support external devices. However, in my actual tests: On iPhone, discoverySession does not detect any external devices. On iPad, discoverySession can detect external devices without any issues. My Question: Does iPhone USB-C actually support external devices (e.g., UVC cameras)? If not, why does Apple's documentation claim that iOS 17 supports external devices instead of specifying iPadOS 17 only?

I have an iOS app that downloads HLS files, which are protected by FairPlay. These files are stored locally, and their locations are managed using Core Data. When playing these tracks, I use AVURLAsset to access the stored file paths. Recently, a client upgraded to a new iPhone and used Quick Start to transfer data from his old device. While all other app data was successfully transferred, including Core Data records and UserDefaults, the actual HLS files were missing. As a result, the app retained metadata about the downloaded content, but the files themselves were gone, causing playback failures. Does Quick Start exclude certain types of locally stored files, especially DRM-protected HLS downloads, or is the issue related to how FairPlay-protected content is handled during the transfer of locally stored files?

Our streaming app uses FairPlay-protected video streams, which previously worked fine when using AVAssetResourceLoaderDelegate to provide CKCs. Recently, we migrated to AVContentKeySession, and while everything works as expected during regular playback, we encountered an issue with AirPlay. Our CKC has a 120-second expiry, so we renew it by calling renewExpiringResponseData.. This trigger the didProvideRenewingContentKeyRequest delegate and we respond with updated CKC. However, when streaming via AirPlay, both video and audio freeze exactly after 120 seconds. To validate the issue, I tested with AVAssetResourceLoaderDelegate and found that I can reproduce the same freeze if I do not renew the key. This suggests that AirPlay is not accepting the renewed CKC when using AVContentKeySession. Additional Details: This issue occurs across different iOS versions and various AirPlay devices. The same content plays without issues when played directly on the device. The renewal process is successful, and segments continue to load, but playback remains frozen. Tried renewing the CKC bit early (100s). I also tried setting player.usesExternalPlaybackWhileExternalScreenIsActive = true, but the issue persists. We don't use persistentKey. Is there anything else that needs to be considered for proper key renewal when AirPlaying? Any help on how to fix this or confirmation if this is a known issue would be greatly appreciated.

I was trying to set custom audio output device for a generated audio on macCatalyst. While using let status = AudioUnitSetProperty(outputUnit, kAudioOutputUnitProperty_CurrentDevice, kAudioUnitScope_Global, 0, &outputDeviceID, UInt32(MemoryLayout.size)) kAudioOutputUnitProperty_CurrentDevice is invalid, and status = -10879, indicating an error. STEPS TO REPRODUCE Set Run Destination to MacOS and run the program. "AudioUnitSetProperty: 0" should be printed, indicating it works fine. Set Run Destination to Mac Catalyst and run the program. "Error setting output device: -10879" should be printed, indicating an error.