Dive into the world of video on Apple platforms, exploring ways to integrate video functionalities within your iOS,iPadOS, macOS, tvOS, visionOS or watchOS app.

Video Documentation

Posts under Video subtopic

Post

Replies

Boosts

Views

Activity

Why AvAssetWriter adds pts drift when writing fMP4
Hi I'm working on a project that require video frame PTS to be consistent between original video and a transcoded one. It's working fairly well on regular mp4, however if I set preferredOutputSegmentInterval to have generate a fMP4 output, even I specified the initialSegmentStartTime as 0, it always add one frame pts offset to all frames. For example: if I use the code sample provided by Apple: https://developer.apple.com/videos/play/wwdc2020/10011/?time=406, useffprobe -select_streams v:0 -show_entries packet=pts_time -of csv ~/Downloads/fmp4/prog_index.m3u8 to display the pts of the output, it doesn't start from 0, but has some one frame pts offset. I also tried open with MP4Box, it also shows the first frames dts and cts are not start from 0. However, if I use AVAssetReader to read the same output video, and get the PTS from 1st frame, it's returning 0. So I can't use it to calculate the pts difference between 2 videos neither. Can I get some help to understand why there is difference between AVAssetWriter/Reader fMP4's pts and others like ffprobe?
0
0
592
Dec ’24
Custom Share Desination stopped working in FCP X 11
We integrate with FCP X using a custom share destination and the Apple Script interface. This has been working fine until the the recent version 11 update of FCP X. With this update we are no longer receiving the open event when the export has completed. We get the apple event to creat the Asset and the file is exported to the location we set in the response. There is just no open event after that. I suspect something is wrong with our scripting support but I have no idea what or how to troubleshoot. This works fine in 10.8.1 and below.
0
0
378
Dec ’24
Is Apple Log open to developers for 3rd party apps?
Hello! I am building a video camera app and trying to implement Apple log for iPhone 15 Pro and 16 Pro. I am not seeing a lot of documentation on it and notice the amount of apps that use it on the app is rather limited. Less an 5 to be exact. Is Apple Log recording a feature that is accessible to developers? Here is a link to documentation: https://developer.apple.com/documentation/avfoundation/avcapturecolorspace/applelog
1
0
493
Jan ’25
AVAssetWriter append audio/video streams concurrently in Real time recording setup
I see in most of the old sample codes from Apple that when using AVAssetWriter to append audio, video, and metadata samples in a real time camera recording setup, calls to .append(sampleBuffer) are either synchronised using an NSLock or all the samples are sent to the asset writer on the same dispatch queue thereby preventing concurrent writes. However I can't find any documentation that calls to assetWriterInput.append(sampleBuffer) for different media samples such as Audio and Video should not be done concurrently. Is it not valid for these methods to be executed in parallel for instance? `videoSamplesAssetWriterInput.append(videoSampleBuffer)` from DispatchQueue 1 `audioSamplesAssetWriterInput.append(audioSampleBuffer)` from DispatchQueue 2
1
0
634
Jan ’25
Exporting Higher Resolution Spatial Videos in Final Cut
How do we export a spatial video at a higher resolution than 2200 x 2200? For example, I want to export a video that I edited in a spatial timeline as 4096 x 4096 resolution with spatial metadata to output as an MV-HEVC file. I looked under both Final Cut and Compressor export settings, but couldn't find a way to do this. Thanks for your help!
1
0
482
Jan ’25
Help diagnosing crash with only system code in its stack trace
Hello, I work on a video streaming app, and I have been working on this crash that we are seeing quite frequently (it is our #2 crasher at the moment). The stack trace indicates that an AVPictureInPictureController is being deallocated on a background thread. This leads to dangling AutoLayout constraints getting cleaned up, further resulting in an exception being thrown from the framework about the layout engine being accessed from a background thread. We have internal analytics which indicate that the crash occurs after the user comes back to the app after putting it in the background, and switches playback from Picture in Picture mode back to the app's regular playback UI. What has me puzzled here is that, as I'm sure you know, we have no control over the PIP UI. It is entirely system-provided, and there is none of our app's code in the stack trace. The whole process is even initiated by a KVO on an AVPlayerController property, and our app doesn't use that class directly anywhere. So how did we manage to cause this process to happen on a background thread? Add to this the fact that the bug only appeared once we switched to compiling with Xcode 16, and is overwhelmingly present only on devices running iOS 18. These factors lead me to believe that this is probably an OS issue. But before I go to file a feedback, I thought I would post here in case anyone has any ideas. I have attached an instance of the crash log to this post. 2025-01-08_17-32-45.7003_-0800-ea8d5c3323e0f1fc059cf83f6ec86377bdae1788.crash
1
0
436
Jan ’25
Capacitor app on iOS inline video stops
Hello, I’m experiencing an issue with video playback in my Javascript (SvelteKit) application using Capacitor. The video plays and loops correctly on Android and web browsers (including Safari) but stops unexpectedly after a few iterations on iOS native App. <video src={videoPath} autoplay muted loop playsinline class="h-auto w-full max-w-full object-cover"></video> Has anyone encountered a similar issue or have insights into what might be causing this behavior on iOS? Any suggestions or workarounds would be greatly appreciated. Maybe it has something to do with the iOS power saving policy? Thank you in advance for your help!
0
0
496
Jan ’25
How start AVPictureInPicture when video is paused
I have AVPlayer with AVPictureInPictureController. Play video in app and picture In Picture works except one situation. Issue is: I pause video in application and during switch to background is not PiP activate. What do I wrong? import UIKit import AVKit import AVFoundation class ViewControllerSec: UIViewController,AVPictureInPictureControllerDelegate { var pipPlayer: AVPlayer! var avCanvas : UIView! var pipCanvas: AVPlayerLayer? var pipController: AVPictureInPictureController! var mainViewControler : UIViewController! var playerItem : AVPlayerItem! var videoAvasset : AVAsset! public func link(to parentViewController : UIViewController) { mainViewControler = parentViewController setup() } @objc func appWillResignActiveNotification(application: UIApplication) { guard let pipController = pipController else { print("PiP not supported") return } print("PIP isSuspend: \(pipController.isPictureInPictureSuspended)") print("PIP isPossible: \(pipController.isPictureInPicturePossible)" if playerItem.status == .readyToPlay { if pipPlayer.rate == 0 { pipPlayer.play() } pipController.startPictureInPicture(). ---> Errorin log: Failed to start picture in picture. } else { print("Player not ready for PiP.") } } private func setupAudio() { do { let session = AVAudioSession.sharedInstance() try session.setCategory(.playback, mode: .moviePlayback) try session.setActive(true) } catch { print("Audio session setup failed: \(error.localizedDescription)") } } @objc func playerItemDidFailToPlayToEnd(_ notification: Notification) { if let error = notification.userInfo?[AVPlayerItemFailedToPlayToEndTimeErrorKey] as? Error { print("Failed to play to end: \(error.localizedDescription)") } } func setup() { setupAudio() guard let videoURL = URL(string: "https://demo.unified-streaming.com/k8s/features/stable/video/tears-of-steel/tears-of-steel.mp4/.m3u8") else { return } videoAvasset = AVAsset(url: videoURL) playerItem = AVPlayerItem(asset: videoAvasset) addPlayerObservers() pipPlayer = AVPlayer(playerItem: playerItem) avCanvas = UIView(frame: view.bounds) pipCanvas = AVPlayerLayer(player: pipPlayer) guard let pipCanvas else { return } pipCanvas.frame = avCanvas.bounds //pipCanvas.videoGravity = .resizeAspectFill mainViewControler.view.addSubview(avCanvas) avCanvas.layer.addSublayer(pipCanvas) if AVPictureInPictureController.isPictureInPictureSupported() { pipController = AVPictureInPictureController(playerLayer: pipCanvas) pipController?.delegate = self pipController?.canStartPictureInPictureAutomaticallyFromInline = true } let playButton = UIButton(frame: CGRect(x: 20, y: 50, width: 100, height: 50)) playButton.setTitle("Play", for: .normal) playButton.backgroundColor = .blue playButton.addTarget(self, action: #selector(playTapped), for: .touchUpInside) mainViewControler.view.addSubview(playButton) let pauseButton = UIButton(frame: CGRect(x: 140, y: 50, width: 100, height: 50)) pauseButton.setTitle("Pause", for: .normal) pauseButton.backgroundColor = .red pauseButton.addTarget(self, action: #selector(pauseTapped), for: .touchUpInside) mainViewControler.view.addSubview(pauseButton) let pipButton = UIButton(frame: CGRect(x: 260, y: 50, width: 150, height: 50)) pipButton.setTitle("Start PiP", for: .normal) pipButton.backgroundColor = .green pipButton.addTarget(self, action: #selector(startPictureInPicture), for: .touchUpInside) mainViewControler.view.addSubview(pipButton) print("Error:\(String(describing: pipPlayer.error?.localizedDescription))") NotificationCenter.default.addObserver(forName: UIApplication.didEnterBackgroundNotification, object: nil, queue: nil) { [weak self] _ in guard let self = self else { return } if self.pipPlayer.rate == 0 { self.pipPlayer.play() pipController?.startPictureInPicture() } } func addPlayerObservers() { playerItem?.addObserver(self, forKeyPath: "status", options: [.old, .new], context: nil) NotificationCenter.default.addObserver(self, selector: #selector(playerDidFinishPlaying(_:)), name: .AVPlayerItemDidPlayToEndTime, object: playerItem) } override func observeValue(forKeyPath keyPath: String?, of object: Any?, change: [NSKeyValueChangeKey : Any]?, context: UnsafeMutableRawPointer?) { if keyPath == "status" { if let statusNumber = change?[.newKey] as? NSNumber { let status = AVPlayer.Status(rawValue: statusNumber.intValue)! switch status { case .readyToPlay: print("Player is ready to play") case .failed: print("Player failed: \(String(describing: playerItem?.error))") case .unknown: print("Player status is unknown") @unknown default: fatalError() } } } } @objc func playerDidFinishPlaying(_ notification: Notification) { print("Video finished playing.") } deinit { playerItem?.removeObserver(self, forKeyPath: "status") NotificationCenter.default.removeObserver(self) } @objc func playTapped() { pipPlayer.play() } @objc func pauseTapped() { pipPlayer.pause() } @objc func startPictureInPicture() { if let pipController = pipController, !pipController.isPictureInPictureActive { pipController.startPictureInPicture() } } @objc func stopPictureInPicture() { if let pipController = pipController, pipController.isPictureInPictureActive { pipController.stopPictureInPicture() } } func pictureInPictureController(_ pictureInPictureController: AVPictureInPictureController, failedToStartPictureInPictureWithError error: Error) { print("Failed to start PiP: \(error.localizedDescription)") if let underlyingError = (error as NSError).userInfo[NSUnderlyingErrorKey] { print("Underlying error: \(underlyingError)") } } }
1
0
544
Jan ’25
slow decoding of animated AVIF images
Safari is supposed to support animated AVIF images since version 16, but the ones I've tested perform very poorly, even on an M4 Mac Mini running Sequoia 15.1.1. I believe Safari delegates decoding to the operating system itself, so this issue also happens in Live Preview in the finder, when I try to preview a file. Sample file here: https://s3.us-west-2.amazonaws.com/cdn.paintera.org/test/sample.avif 322KB file, 5 seconds long, 12fps This plays perfectly on Chrome on Mac OS, but is slow and laggy on Safari and Live Preview (it takes about 6.5 seconds to finish the 5 second video). Does anyone know how to fix this or workaround this issue?
2
1
564
Jan ’25
AVAssetExportSession in iOS18- Thread 11: "*** -[AVAssetExportSession exportAsynchronouslyWithCompletionHandler:] Cannot call exportAsynchronouslyWithCompletionHandler: more than once."
I’m experiencing a crash at runtime when trying to extract audio from a video. This issue occurs on both iOS 18 and earlier versions. The crash is caused by the following error: *** Terminating app due to uncaught exception 'NSInternalInconsistencyException', reason: '*** -[AVAssetExportSession exportAsynchronouslyWithCompletionHandler:] Cannot call exportAsynchronouslyWithCompletionHandler: more than once.' *** First throw call stack: (0x1875475ec 0x184ae1244 0x1994c49c0 0x217193358 0x217199899 0x192e208b9 0x217192fd9 0x30204c88d 0x3019e5155 0x301e5fb41 0x301af7add 0x301aff97d 0x301af888d 0x301aff27d 0x301ab5fa5 0x301ab6101 0x192e5ee39) libc++abi: terminating due to uncaught exception of type NSException My previous code worked fine, but it's crashing with Swift 6. Does anyone know a solution for this? ## **Previous code:** func extractAudioFromVideo(from videoURL: URL, exportHandler: ((AVAssetExportSession, CurrentValueSubject<Float, Never>?) -> Void)? = nil, completion: @escaping (Swift.Result<URL, Error>) -> Void) { let asset = AVAsset(url: videoURL) // Create an AVAssetExportSession to export the audio track guard let exportSession = AVAssetExportSession(asset: asset, presetName: AVAssetExportPresetAppleM4A) else { completion(.failure(NSError(domain: "com.example.app", code: -1, userInfo: [NSLocalizedDescriptionKey: "Failed to create AVAssetExportSession"]))) return } // Set the output file type and path guard let filename = videoURL.lastPathComponent.components(separatedBy: ["."]).first else { return } let outputURL = VideoUtils.getTempAudioExportUrl(filename) VideoUtils.deleteFileIfExists(outputURL.path) exportSession.outputFileType = .m4a exportSession.outputURL = outputURL let audioExportProgressPublisher = CurrentValueSubject<Float, Never>(0.0) if let exportHandler = exportHandler { exportHandler(exportSession, audioExportProgressPublisher) } // Periodically check the progress of the export session let timer = Timer.scheduledTimer(withTimeInterval: 0.1, repeats: true) { _ in audioExportProgressPublisher.send(exportSession.progress) } // Export the audio track asynchronously exportSession.exportAsynchronously { switch exportSession.status { case .completed: completion(.success(outputURL)) case .failed: completion(.failure(exportSession.error ?? NSError(domain: "com.example.app", code: -1, userInfo: [NSLocalizedDescriptionKey: "Unknown error occurred while exporting audio"]))) case .cancelled: completion(.failure(NSError(domain: "com.example.app", code: -1, userInfo: [NSLocalizedDescriptionKey: "Export session was cancelled"]))) default: completion(.failure(NSError(domain: "com.example.app", code: -1, userInfo: [NSLocalizedDescriptionKey: "Unknown export session status"]))) } // Invalidate the timer when the export session completes or is cancelled timer.invalidate() } } ## New Code: func extractAudioFromVideo(from videoURL: URL, exportHandler: ((AVAssetExportSession, CurrentValueSubject<Float, Never>?) -> Void)? = nil, completion: @escaping (Swift.Result<URL, Error>) -> Void) async { let asset = AVAsset(url: videoURL) // Create an AVAssetExportSession to export the audio track guard let exportSession = AVAssetExportSession(asset: asset, presetName: AVAssetExportPresetAppleM4A) else { completion(.failure(NSError(domain: "com.example.app", code: -1, userInfo: [NSLocalizedDescriptionKey: "Failed to create AVAssetExportSession"]))) return } // Set the output file type and path guard let filename = videoURL.lastPathComponent.components(separatedBy: ["."]).first else { return } let outputURL = VideoUtils.getTempAudioExportUrl(filename) VideoUtils.deleteFileIfExists(outputURL.path) let audioExportProgressPublisher = CurrentValueSubject<Float, Never>(0.0) if let exportHandler { exportHandler(exportSession, audioExportProgressPublisher) } if #available(iOS 18.0, *) { do { try await exportSession.export(to: outputURL, as: .m4a) let states = exportSession.states(updateInterval: 0.1) for await state in states { switch state { case .pending, .waiting: break case .exporting(progress: let progress): print("Exporting: \(progress.fractionCompleted)") if progress.isFinished { completion(.success(outputURL)) }else if progress.isCancelled { completion(.failure(NSError(domain: "com.example.app", code: -1, userInfo: [NSLocalizedDescriptionKey: "Export session was cancelled"]))) }else { audioExportProgressPublisher.send(Float(progress.fractionCompleted)) } } } }catch let error { print(error.localizedDescription) } }else { // Periodically check the progress of the export session let publishTimer = Timer.publish(every: 0.1, on: .main, in: .common) .autoconnect() .sink { [weak exportSession] _ in guard let exportSession else { return } audioExportProgressPublisher.send(exportSession.progress) } exportSession.outputFileType = .m4a exportSession.outputURL = outputURL await exportSession.export() switch exportSession.status { case .completed: completion(.success(outputURL)) case .failed: completion(.failure(exportSession.error ?? NSError(domain: "com.example.app", code: -1, userInfo: [NSLocalizedDescriptionKey: "Unknown error occurred while exporting audio"]))) case .cancelled: completion(.failure(NSError(domain: "com.example.app", code: -1, userInfo: [NSLocalizedDescriptionKey: "Export session was cancelled"]))) default: completion(.failure(NSError(domain: "com.example.app", code: -1, userInfo: [NSLocalizedDescriptionKey: "Unknown export session status"]))) } // Invalidate the timer when the export session completes or is cancelled publishTimer.cancel() } }
1
0
697
Jan ’25
AVPlayerItem step(byCount:) callback or notification
Hello there, I need to move through video loaded in an AVPlayer one frame at a time back or forth. For that I tried to use AVPlayerItem's method step(byCount:) and it works just fine. However I need to know when stepping happened and as far as I observed it is not immediate using the method. If I check the currentTime() just after calling the method it's the same and if I do it slightly later (depending of the video itself) it shows the correct "jumped" time. To achieve my goal I tried subclassing AVPlayerItem and implement my own async method utilizing NotificationCenter and the timeJumpedNotification assuming it would deliver it as the time actually jumps but it's not the case. Here is my "stripped" and simplified version of the custom Player Item: import AVFoundation final class PlayerItem: AVPlayerItem { private var jumpCompletion: ( (CMTime) -> () )? override init(asset: AVAsset, automaticallyLoadedAssetKeys: [String]?) { super .init(asset: asset, automaticallyLoadedAssetKeys: automaticallyLoadedAssetKeys) NotificationCenter.default.addObserver(self, selector: #selector(timeDidChange(_:)), name: AVPlayerItem.timeJumpedNotification, object: self) } deinit { NotificationCenter.default.removeObserver(self, name: AVPlayerItem.timeJumpedNotification, object: self) jumpCompletion = nil } @discardableResult func step(by count: Int) async -> CMTime { await withCheckedContinuation { continuation in step(by: count) { time in continuation.resume(returning: time) } } } func step(by count: Int, completion: @escaping ( (CMTime) -> () )) { guard jumpCompletion == nil else { completion(currentTime()) return } jumpCompletion = completion step(byCount: count) } @objc private func timeDidChange(_ notification: Notification) { switch notification.name { case AVPlayerItem.timeJumpedNotification where notification.object as? AVPlayerItem [==](https://www.example.com/) self: jumpCompletion?(currentTime()) jumpCompletion = nil default: return } } } In short the notification never gets called thus the above is not working. I guess the key there is that in the docs about the timeJumpedNotification: is said: "A notification the system posts when a player item’s time changes discontinuously." so the step(byCount:) is not considered as discontinuous operation and doesn't trigger it. I'd be really helpful if somebody can help as I don't want to use seek(to:toleranceBefore:toleranceAfter:) mainly cause it's not accurate in terms of the exact next/previous frame as the video might have VFR and that causes repeating frames sometimes or even skipping one or another. Thanks a lot
2
0
636
Jan ’25
4k 120fps Showing Black Screen on iPhone 16
Hey - I am developing an app that uses the camera for recording video. I put the ability to choose a framerate and resolution and all combinations work perfectly fine, except for 4k 120fps for the new iPhone 16 pro. This just shows black on the preview. I tried to record even though the preview was black, but the recording is also just a black screen. Is there anything special that needs to be done in the camera setup for 4k 120fps to work? I have my camera setup code attached. Is it possible this is a bug in Apple's code, since this works with every other combination (1080p up to 240fps and 4k up to 60fps)? Thanks so much for the help. class CameraManager: NSObject { enum Errors: Error { case noCaptureDevice case couldNotAddInput case unsupportedConfiguration } enum Resolution { case hd1080p case uhd4K var preset: AVCaptureSession.Preset { switch self { case .hd1080p: return .hd1920x1080 case .uhd4K: return .hd4K3840x2160 } } var dimensions: CMVideoDimensions { switch self { case .hd1080p: return CMVideoDimensions(width: 1920, height: 1080) case .uhd4K: return CMVideoDimensions(width: 3840, height: 2160) } } } enum CameraType { case wide case ultraWide var captureDeviceType: AVCaptureDevice.DeviceType { switch self { case .wide: return .builtInWideAngleCamera case .ultraWide: return .builtInUltraWideCamera } } } enum FrameRate: Int { case fps60 = 60 case fps120 = 120 case fps240 = 240 } let orientationManager = OrientationManager() let captureSession: AVCaptureSession let previewLayer: AVCaptureVideoPreviewLayer let movieFileOutput = AVCaptureMovieFileOutput() let videoDataOutput = AVCaptureVideoDataOutput() private var videoCaptureDevice: AVCaptureDevice? override init() { self.captureSession = AVCaptureSession() self.previewLayer = AVCaptureVideoPreviewLayer(session: self.captureSession) super.init() self.previewLayer.videoGravity = .resizeAspect } func configureSession(resolution: Resolution, frameRate: FrameRate, stabilizationEnabled: Bool, cameraType: CameraType, sampleBufferDelegate: AVCaptureVideoDataOutputSampleBufferDelegate?) throws { assert(Thread.isMainThread) captureSession.beginConfiguration() defer { captureSession.commitConfiguration() } captureSession.sessionPreset = resolution.preset if captureSession.canAddOutput(movieFileOutput) { captureSession.addOutput(movieFileOutput) } else { throw Errors.couldNotAddInput } videoDataOutput.setSampleBufferDelegate(sampleBufferDelegate, queue: DispatchQueue(label: "VideoDataOutputQueue")) if captureSession.canAddOutput(videoDataOutput) { captureSession.addOutput(videoDataOutput) // Set the video orientation if needed if let connection = videoDataOutput.connection(with: .video) { //connection.videoOrientation = .portrait } } else { throw Errors.couldNotAddInput } guard let videoCaptureDevice = AVCaptureDevice.default(cameraType.captureDeviceType, for: .video, position: .back) else { throw Errors.noCaptureDevice } let useDimensions = resolution.dimensions guard let format = videoCaptureDevice.formats.first(where: { format in let dimensions = CMVideoFormatDescriptionGetDimensions(format.formatDescription) let isRes = dimensions.width == useDimensions.width && dimensions.height == useDimensions.height let frameRates = format.videoSupportedFrameRateRanges return isRes && frameRates.contains(where: { $0.maxFrameRate >= Float64(frameRate.rawValue) }) }) else { throw Errors.unsupportedConfiguration } self.videoCaptureDevice = videoCaptureDevice do { let videoInput = try AVCaptureDeviceInput(device: videoCaptureDevice) if captureSession.canAddInput(videoInput) { captureSession.addInput(videoInput) } else { throw Errors.couldNotAddInput } try videoCaptureDevice.lockForConfiguration() videoCaptureDevice.activeFormat = format videoCaptureDevice.activeVideoMinFrameDuration = CMTime(value: 1, timescale: CMTimeScale(frameRate.rawValue)) videoCaptureDevice.activeVideoMaxFrameDuration = CMTime(value: 1, timescale: CMTimeScale(frameRate.rawValue)) videoCaptureDevice.activeMaxExposureDuration = CMTime(seconds: 1.0 / 960, preferredTimescale: 1000000) videoCaptureDevice.exposureMode = .locked videoCaptureDevice.unlockForConfiguration() } catch { throw error } configureStabilization(enabled: stabilizationEnabled) }`
0
0
463
Jan ’25
Transparent overlay changes color in HDR video
Overlay changes color in HDR video When I’m using trying to add an overlay to an image with AVMutableVideoComposition, When the video is in HDR the overlay colors are changing and white becomes grey screen shot from original HDR video result from the code with the wrong overlay colorthe result when reducing to SDR (the right overlay color) the distorted colorsthe way it should look(sdr) Im creating the overlay with a CGContext class CustomHdrCompositor: NSObject, AVVideoCompositing { private let coreImageContext = CIContext(options: [CIContextOption.cacheIntermediates: false]) let combinedFilter = CIFilter(name: "CISourceOverCompositing")! var sourcePixelBufferAttributes: [String: Any]? = [String(kCVPixelBufferPixelFormatTypeKey): [kCVPixelFormatType_420YpCbCr10BiPlanarVideoRange]] var requiredPixelBufferAttributesForRenderContext: [String: Any] = [String(kCVPixelBufferPixelFormatTypeKey): [kCVPixelFormatType_420YpCbCr10BiPlanarVideoRange]] var supportsWideColorSourceFrames = true var supportsHDRSourceFrames = true func renderContextChanged(_ newRenderContext: AVVideoCompositionRenderContext) { return } func startRequest(_ request: AVAsynchronousVideoCompositionRequest) { guard let outputPixelBuffer = request.renderContext.newPixelBuffer() else { print("No valid pixel buffer found. Returning.") request.finish(with: CustomCompositorError.ciFilterFailedToProduceOutputImage) return } guard let requiredTrackIDs = request.videoCompositionInstruction.requiredSourceTrackIDs, !requiredTrackIDs.isEmpty else { print("No valid track IDs found in composition instruction.") return } let sourceCount = requiredTrackIDs.count if sourceCount > 1 { request.finish(with: CustomCompositorError.notSupportingMoreThanOneSources) return } if sourceCount == 1 { let sourceID = requiredTrackIDs[0] let sourceBuffer = request.sourceFrame(byTrackID: sourceID.value(of: Int32.self)!)! let sourceCIImage = CIImage(cvPixelBuffer: sourceBuffer) var textImage = TextLayerPlayer.instance.getTextLayerAtTimesStamp(ts:request.compositionTime.seconds) combinedFilter.setValue(textImage, forKey: "inputImage") if let outputImage = combinedFilter.outputImage { let renderDestination = CIRenderDestination(pixelBuffer: outputPixelBuffer) do { try coreImageContext.startTask(toRender: outputImage, to: renderDestination) } catch { } } } request.finish(withComposedVideoFrame: outputPixelBuffer) } } func regularCompositionHdr(asset: AVAsset) -> AVVideoComposition { self.isHdr = checkHdr(asset: asset) let avComposition = AVMutableComposition() let composition = AVMutableVideoComposition() composition.colorPrimaries = AVVideoColorPrimaries_ITU_R_2020 composition.colorTransferFunction = AVVideoTransferFunction_ITU_R_2100_HLG composition.colorYCbCrMatrix = AVVideoYCbCrMatrix_ITU_R_2020 composition.renderSize = assetSize composition.frameDuration = CMTime(value: 1, timescale: 30) composition.customVideoCompositorClass = CustomHdrCompositor.self composition.perFrameHDRDisplayMetadataPolicy = .propagate return composition } I’m using this function to transfer the transparent CGImage to CIImage that supports HDR func convertToHDRCIImage(from cgImage: CGImage, maxBrightness: CGFloat = 3.0) -> CIImage? { // Create a CIImage from the input CGImage let baseImage = CIImage(cgImage: cgImage) // Create HDR color adjustment filter let colorAdjust = CIFilter(name: "CIColorMatrix")! colorAdjust.setValue(baseImage, forKey: kCIInputImageKey) // Calculate HDR multipliers based on maxBrightness // This will maintain color ratios while increasing brightness colorAdjust.setValue(CIVector(x: maxBrightness, y: 0, z: 0, w: 0), forKey: "inputRVector") colorAdjust.setValue(CIVector(x: 0, y: maxBrightness, z: 0, w: 0), forKey: "inputGVector") colorAdjust.setValue(CIVector(x: 0, y: 0, z: maxBrightness, w: 0), forKey: "inputBVector") // Maintain alpha channel colorAdjust.setValue(CIVector(x: 0, y: 0, z: 0, w: 1), forKey: "inputAVector") guard let adjustedImage = colorAdjust.outputImage else { return nil } // Apply color space transformation using CIImage's colorSpace property let transformedImage = adjustedImage.matchedFromWorkingSpace(to: hdrWorkingSpace)! // Create context with HDR color space let context = CIContext(options: [ .workingColorSpace: hdrColorSpace, .outputColorSpace: hdrColorSpace ]) // Get the image bounds let bounds = transformedImage.extent // Create a new pixel buffer with HDR format var pixelBuffer: CVPixelBuffer? let pixelBufferAttributes = [ kCVPixelBufferPixelFormatTypeKey: kCVPixelFormatType_64RGBAHalf, kCVPixelBufferMetalCompatibilityKey: true ] as CFDictionary CVPixelBufferCreate(kCFAllocatorDefault, Int(bounds.width), Int(bounds.height), kCVPixelFormatType_64RGBAHalf, pixelBufferAttributes, &pixelBuffer) guard let destinationBuffer = pixelBuffer else { return nil } context.render(transformedImage, to: destinationBuffer, bounds: bounds, colorSpace: hdrColorSpace) // Create final CIImage from the HDR pixel buffer let finalImage = CIImage(cvPixelBuffer: destinationBuffer, options: [.colorSpace: hdrColorSpace]) return finalImage } When reducing the HDR to SDR it keeps the right color of the overlay with, but than it reduces the HDR effect which I want to keep
0
0
393
Jan ’25
Playing fMP4 Raw Chunks in AVPlayer on iOS
Hello Apple Community, We are working on a real-time streaming feature where we receive chunks of raw MP4 data through a custom protocol and store them in a buffer (array). Our goal is to use these data chunks to play a continuous video stream in AVPlayer. What We've Tried: Custom URL Scheme with AVAssetResourceLoaderDelegate: We implemented a custom URL scheme (customscheme://) to serve the buffered data using AVAssetResourceLoaderDelegate. The method shouldWaitForLoadingOfRequestedResource is called only during the initial allocation. It doesn't get triggered when new chunks are appended to the buffer. Despite appending new data to the buffer, AVPlayer doesn’t request further chunks from the delegate. What We Need: We are looking for a solution where: The player continuously fetches data from the buffer as new chunks are added. The playback remains smooth and uninterrupted, even with real-time data being appended. Ideally, this solution works with AVPlayer while adhering to HLS-like behavior without implementing an HLS server. Questions: Is AVAssetResourceLoaderDelegate the right approach for this use case? If so, how can we ensure shouldWaitForLoadingOfRequestedResource is called whenever new data is available in the buffer? Are there alternative APIs or recommended patterns for playing real-time MP4 data chunks in AVPlayer? Would implementing a custom FFmpeg-based player be necessary, or can this be achieved using AVPlayer and its APIs? We appreciate any guidance, suggestions, or examples that can help us achieve this. Thank you!
0
1
569
Jan ’25
Use Metal to conver HDR Pixelbuffer to SDR Pixelbuffer
I see some demo show convert HDR video to SDR Pixelbuffer,such AVAssetReader、 AVVideoComposition 、AVComposition 、AVFoundation. But In some cases,I want to render HDR Pixelbuffer and record video. AVCaptureSession *session = [[AVCaptureSession alloc] init]; session.sessionPreset = AVCaptureSessionPresetHigh; AVCaptureDevice *videoDevice = [AVCaptureDevice defaultDeviceWithMediaType:AVMediaTypeVideo]; if ([videoDevice isVideoHDRSupported]) { NSError *error = nil; if ([videoDevice lockForConfiguration:&error]) { videoDevice.automaticallyAdjustsVideoHDREnabled = NO; videoDevice.videoHDREnabled = YES; // 开启 HDR [videoDevice unlockForConfiguration]; } else { NSLog(@"Error: %@", error.localizedDescription); } } Real-time processing of HDR data requires processing of video frame data (such as filters), ensuring that the processing chain supports 10-bit color depth and HDR metadata. And use imagesBuffer to object tracking, etc. How to solve this problem?
1
0
459
Jan ’25
Encoder Performance Unable to Achieve 4k 120fps on iPhone 16 Pro
Hello, I am trying to get the new iPhone 16 pro to achieve 4k 120fps encoding when we are getting the video feed from the default, wide angle camera on the back. We are using the apple API to capture the individual frames from the camera as they are processed and we get them in this callback: // this is the main callback function to handle video frames captured func captureOutput(_ output: AVCaptureOutput, didOutput sampleBuffer: CMSampleBuffer, from connection: AVCaptureConnection) { We are then taking these frames as they come in and encoding them using VideoToolBox. After they are encoded, they are added to a ring buffer so we can access them after they have been encoded. The problem is that when we are encoding these frames on an iPhone 16 Pro, we are only reaching 80-90fps instead of 120fps. We have removed as much processing as we can. We get some small attributes about the frame when it comes in, encode the frame, and then add it to our ring buffer. I have attached a sample project that is broken down as much as possible to the basic task of encoding 4k 120fps footage. Inside the sample app, there is an fps and pps display showing how many frames we are encoding per second. FPS represents how many frames we are coming in per second from the camera, and PPS represents how many frames we are processing (encoding) per second. Link to sample project: https://github.com/jake-fishtech/EncoderPerformance Thanks you for any help or suggestions.
2
0
581
Jan ’25
AVURLAsset with AVURLAssetHTTPCookiesKey - Cookies not persisting on retry requests
I'm experiencing an unexpected behavior with AVURLAsset and cookies. When setting cookies through AVURLAssetHTTPCookiesKey option, they seem to be sent only on the initial request but not on retry attempts. Here's my current implementation: let cookieProperties: [HTTPCookiePropertyKey: Any] = [ .name: "sessionCookie", .value: "testValue", .domain: url.host ?? "", .path: "/", .secure: true ] if let cookie = HTTPCookie(properties: cookieProperties) { let asset = AVURLAsset(url: url, options: [ AVURLAssetHTTPCookiesKey: [cookie], ]) } According to the documentation, AVURLAssetHTTPCookiesKey should apply the cookies to all requests made by this asset. However, when the initial request fails and AVPlayer retries, the cookies are not included in subsequent requests. Only when I store the cookie with HTTPCookieStorage.shared.setCookie, then it persists. Questions: Is this the expected behavior? If not, what could be causing the cookies to not persist for retry attempts? Is using HTTPCookieStorage.shared the recommended approach instead? Environment: iOS 16+ Using AVPlayer with AVURLAsset Streaming HLS content Any insights would be greatly appreciated.
0
0
345
Feb ’25