iOS AVFoundation - 录制视频
iOS 中的音视频处理库 AVFoundation 内容很丰富，功能也很强大，本文主要讲解使用 ffprobe 查看视频信息和 AVFoundation 录制视频的问题。
ffprobe 查看视频信息
先介绍下 ffprobe，ffprobe 是 ffmpeg 中的音视频内容分析工具，对于后面通过 AVFoundation 录制的视频， ffprobe 可以帮助我们理解一些视频信息。
下面是视频文件中需要关注的常见信息：
- 封装格式，时长，存储大小
- 视频编码，视频码率，分辨率，帧率
- 音频编码，音频采样率，音频码率，声道
通过 ffprobe 来查看：
$ ffprobe -show_format -pretty hd1280x720.mp4
ffprobe version N-93044-g2e2b44baba Copyright (c) 2007-2019 the FFmpeg developers
built with Apple LLVM version 10.0.0 (clang-1000.11.45.5)
configuration: --enable-gpl --enable-nonfree --enable-libfdk-aac --enable-libmp3lame --enable-libx264 --enable-libx265
libavutil 56. 26.100 / 56. 26.100
libavcodec 58. 46.100 / 58. 46.100
libavformat 58. 26.100 / 58. 26.100
libavdevice 58. 6.101 / 58. 6.101
libavfilter 7. 48.100 / 7. 48.100
libswscale 5. 4.100 / 5. 4.100
libswresample 3. 4.100 / 3. 4.100
libpostproc 55. 4.100 / 55. 4.100
Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'hd1280x720.mp4':
Metadata:
major_brand : qt
minor_version : 0
compatible_brands: qt
creation_time : 2019-09-09T03:39:32.000000Z
Duration: 00:00:10.11, start: 0.000000, bitrate: 5263 kb/s
Stream #0:0(und): Video: hevc (Main) (hvc1 / 0x31637668), yuv420p(tv, bt709), 1280x720, 5165 kb/s, 29.98 fps, 29.97 tbr, 600 tbn, 600 tbc (default)
Metadata:
rotate : 90
creation_time : 2019-09-09T03:39:32.000000Z
handler_name : Core Media Video
encoder : HEVC
Side data:
displaymatrix: rotation of -90.00 degrees
Stream #0:1(und): Audio: aac (LC) (mp4a / 0x6134706D), 44100 Hz, mono, fltp, 88 kb/s (default)
Metadata:
creation_time : 2019-09-09T03:39:32.000000Z
handler_name : Core Media Audio
[FORMAT]
filename=hd1280x720.mp4
nb_streams=2
nb_programs=0
format_name=mov,mp4,m4a,3gp,3g2,mj2
format_long_name=QuickTime / MOV
start_time=0:00:00.000000
duration=0:00:10.108333
size=6.342447 Mibyte
bit_rate=5.263410 Mbit/s
probe_score=100
TAG:major_brand=qt
TAG:minor_version=0
TAG:compatible_brands=qt
TAG:creation_time=2019-09-09T03:39:32.000000Z
[/FORMAT]可以看到此视频的情况如下：
- 封装格式：format_name=mov,mp4,m4a,3gp,3g2,mj2
- 时长：duration=0:00:10.108333
- 存储大小：size=6.342447 Mibyte
- 视频编码：hevc
- 视频码率：5165 kb/s
- 分辨率：1280x720
- 帧率：29.98 fps
- 音频编码：aac
- 音频采样率：44100 Hz
- 音频码率：88 kb/s
- 声道：mono
由此可见，在后面调节录制的参数后，将视频拷贝到 macOS 上，再通过 ffprobe 来分析将会很方便。
搭建 AVCaptureSession
完整代码：
上图中的第三路流程就是视频录制的 AVCaptureSession，这里使用 AVCaptureMovieFileOutput 作为输出。
第一步，获取相机使用权限：
switch AVCaptureDevice.authorizationStatus(for: .video) {
case .authorized:
break
case .notDetermined:
sessionQueue.suspend()
AVCaptureDevice.requestAccess(for: .video) { [weak self] granted in
guard let self = self else { return }
if !granted {
self.setupResult = .notAuthorized
}
self.sessionQueue.resume()
}
default:
setupResult = .notAuthorized
}第二步，选择摄像头，也就是 AVCaptureDevice：
var defaultVideoDevice: AVCaptureDevice?
var backCameraDevice: AVCaptureDevice?
var frontCameraDevice: AVCaptureDevice?
for cameraDevice in AVCaptureDevice.devices(for: .video) {
if cameraDevice.position == .back {
backCameraDevice = cameraDevice
}
if cameraDevice.position == .front {
frontCameraDevice = cameraDevice
}
}
if let backCameraDevice = backCameraDevice {
defaultVideoDevice = backCameraDevice
} else {
defaultVideoDevice = frontCameraDevice
}
guard let videoDevice = defaultVideoDevice else {
print("Could not find video device")
setupResult = .configurationFailed
session.commitConfiguration()
return
}第三步，配置预设，根据分辨率来选择，或者，根据类型是照片或视频来选择：
private func configSessionPreset(for videoDevice: AVCaptureDevice) {
guard let source = source else { return }
var presets: [AVCaptureSession.Preset] = []
if source == .capture {
if #available(iOS 9, *) {
presets.append(.hd4K3840x2160)
}
presets.append(.hd1920x1080)
presets.append(.photo)
for preset in presets {
if videoDevice.supportsSessionPreset(preset) {
session.sessionPreset = preset
break
}
}
} else {
presets.append(.hd1280x720)
presets.append(.medium)
for preset in presets {
if videoDevice.supportsSessionPreset(preset) {
session.sessionPreset = preset
break
}
}
}
}第四步，通过 AVCaptureDevice 来调整帧率（FPS）：
private func configRecordingFPS(for videoDevice: AVCaptureDevice) {
guard let source = source, source == .recording else { return }
let desiredFrameRate = mode.config.recordingFrameRate
var isFPSSupported = false
for range in videoDevice.activeFormat.videoSupportedFrameRateRanges {
if Double(desiredFrameRate) <= range.maxFrameRate,
Double(desiredFrameRate) >= range.minFrameRate {
isFPSSupported = true
}
}
if isFPSSupported {
do {
try videoDevice.lockForConfiguration()
videoDevice.activeVideoMinFrameDuration = CMTime(value: 1, timescale: CMTimeScale(desiredFrameRate))
videoDevice.activeVideoMaxFrameDuration = CMTime(value: 1, timescale: CMTimeScale(desiredFrameRate))
videoDevice.unlockForConfiguration()
} catch {
print("Could not config video device frame duration: \(error)")
}
}
}第五步，往 AVCaptureSession 添加视频设备输入 AVCaptureDeviceInput，当然设备输入来源是之前的视频设备 AVCaptureDevice：
do {
let videoDeviceInput = try AVCaptureDeviceInput(device: videoDevice)
if session.canAddInput(videoDeviceInput) {
session.addInput(videoDeviceInput)
self.videoDeviceInput = videoDeviceInput
} else {
print("Could not add video device input to the session")
setupResult = .configurationFailed
session.commitConfiguration()
return
}
} catch {
print("Could not create video device input: \(error)")
setupResult = .configurationFailed
session.commitConfiguration()
return
}第六步，往 AVCaptureSession 添加音频设备输入 AVCaptureDeviceInput：
guard let audioDevice = AVCaptureDevice.default(for: .audio) else {
print("Could not find audio device")
setupResult = .configurationFailed
session.commitConfiguration()
return
}
do {
let audioDeviceInput = try AVCaptureDeviceInput(device: audioDevice)
if session.canAddInput(audioDeviceInput) {
session.addInput(audioDeviceInput)
} else {
print("Could not add audio device input to the session")
setupResult = .configurationFailed
session.commitConfiguration()
return
}
} catch {
print("Could not create audio device input: \(error)")
setupResult = .configurationFailed
session.commitConfiguration()
return
}第七步，往 AVCaptureSession 添加视频文件输出 AVCaptureMovieFileOutput：
private func configSessionOutput() {
guard let source = source else { return }
if source == .capture {
session.removeOutput(movieFileOutput)
if session.canAddOutput(stillImageOutput) {
session.addOutput(stillImageOutput)
} else {
print("Could not add still image output to the session")
setupResult = .configurationFailed
session.commitConfiguration()
return
}
} else {
session.removeOutput(stillImageOutput)
if session.canAddOutput(movieFileOutput) {
session.addOutput(movieFileOutput)
} else {
print("Could not add movie file output to the session")
setupResult = .configurationFailed
session.commitConfiguration()
return
}
}
}第八步，配置 AVCaptureMovieFileOutput 的编码和码率：
private func configRecordingBitRate() {
guard let source = source, source == .recording else { return }
if #available(iOS 12.0, *) {
if let recordingConnection = movieFileOutput.connection(with: .video) {
let supportedSettingsKeys = movieFileOutput.supportedOutputSettingsKeys(for: recordingConnection)
var outputSettings: [String: Any] = [:]
for (settingKey, settingValue) in movieFileOutput.outputSettings(for: recordingConnection) {
if supportedSettingsKeys.contains(settingKey) {
if settingKey == AVVideoCompressionPropertiesKey {
var compressionProperties: [String: Any] = [:]
if let properties = settingValue as? [String: Any] {
for (key, value) in properties {
if key == AVVideoAverageBitRateKey {
compressionProperties[key] = mode.config.recordingBitRate
} else {
compressionProperties[key] = value
}
}
}
outputSettings[settingKey] = compressionProperties
} else {
outputSettings[settingKey] = settingValue
}
}
}
movieFileOutput.setOutputSettings(outputSettings, for: recordingConnection)
}
}
}可能可以调节的参数查看 Video Settings，尝试了发现能够调节的参数非常有限，视频编码有 h264 和 hevc，上面的代码没有设置此项，实际的情况是在新设备编码为 hevc，在旧设备编码为 h264。
视频预览
通过 AVCaptureVideoPreviewLayer 来实现视频画面的预览，再将前面搭建好的 AVCaptureSession 赋值给 AVCaptureVideoPreviewLayer：
class CameraPreviewView: UIView {
var session: AVCaptureSession? {
set {
videoPreviewLayer?.session = newValue
videoPreviewLayer?.videoGravity = .resizeAspectFill
}
get {
return videoPreviewLayer?.session
}
}
override class var layerClass: AnyClass {
return AVCaptureVideoPreviewLayer.self
}
private var videoPreviewLayer: AVCaptureVideoPreviewLayer? {
return layer as? AVCaptureVideoPreviewLayer
}
}
previewView.session = session录制视频
通过 AVCaptureMovieFileOutput 录制视频到指定文件：
private func recording() {
sessionQueue.async { [weak self] in
guard let self = self else { return }
if !self.movieFileOutput.isRecording {
if let outputURL = MediaViewController.getMediaFileURL(name: "video", ext: "mp4") {
DispatchQueue.main.async { [weak self] in
self?.toggleRecordingControls(isHidden: false)
}
if UIDevice.current.isMultitaskingSupported {
self.backgroundRecordingID = UIApplication.shared.beginBackgroundTask(expirationHandler: nil)
}
self.movieFileOutput.startRecording(to: outputURL, recordingDelegate: self)
} else {
print("Could not create video file")
}
} else {
DispatchQueue.main.async { [weak self] in
self?.toggleRecordingControls(isHidden: true)
}
self.movieFileOutput.stopRecording()
}
}
}这里还需要实现 AVCaptureFileOutputRecordingDelegate 来响应录制视频过程中一些状态的变化：
extension CameraViewController: AVCaptureFileOutputRecordingDelegate {
func fileOutput(_ output: AVCaptureFileOutput, didStartRecordingTo fileURL: URL, from connections: [AVCaptureConnection]) {
startTimer()
}
func fileOutput(_ output: AVCaptureFileOutput, didFinishRecordingTo outputFileURL: URL, from connections: [AVCaptureConnection], error: Error?) {
var success = true
if let error = error {
success = false
print("Movie file finishing error: \(error)")
}
if success {
DispatchQueue.main.async { [weak self] in
self?.previewVideo(outputFileURL)
}
} else {
cancelTimer()
DispatchQueue.main.async { [weak self] in
self?.toggleRecordingControls(isHidden: true)
self?.updateDuration()
}
}
cleanupRecording()
}
}获取视频帧的图像
通过 AVAssetImageGenerator 来实现：
func generatorVideoThumbnail(video: URL, completion: @escaping (UIImage?, TimeInterval) -> Void) {
DispatchQueue.global().async {
let time = CMTimeMakeWithSeconds(1, preferredTimescale: 1)
let asset = AVURLAsset(url: video, options: nil)
if asset.tracks(withMediaType: .video).isNotEmpty {
let duration = CMTimeGetSeconds(asset.duration)
let imageGenerator = AVAssetImageGenerator(asset: asset)
imageGenerator.requestedTimeToleranceAfter = CMTime.zero
imageGenerator.requestedTimeToleranceBefore = CMTime.zero
imageGenerator.appliesPreferredTrackTransform = true
imageGenerator.generateCGImagesAsynchronously(
forTimes: [NSValue(time: time)], completionHandler: { _, image, _, _, _ in
DispatchQueue.main.async {
guard let cgimg = image else {
completion(nil, TimeInterval(duration))
return
}
let image = UIImage(cgImage: cgimg)
completion(image, TimeInterval(duration))
}
}
)
} else {
DispatchQueue.main.async {
completion(nil, 0)
}
}
}
}