Format related:
There are four supported output formats:
[“wav”, “aif”, “caf”, “m4a”]
Caf, which is core Audio Format, supports uncompressed formats, and compressed formats AAC
Wav, uncompressed format
M4a, compressed format, where the general data format is AAC compressed format
Does not support mp3 transfer out,
Audio to MP3, usually using lame library
Supported input formats:
[“wav”, “aif”, “caf”, “m4a”
“mp3”, “snd”, “au”, “sd2”,
“aif”, “aiff”, “aifc”, “aac”,
“mp4”, “m4v”, “mov” ]
And there’s no ending
Audio format conversion generally falls into four categories:
- Conversion between uncompressed audio
- Compressed audio between the conversion
Mp3 transfer is not supported
- Uncompressed audio, go to compressed audio
Mp3 transfer is not supported
- Compressed audio, uncompressed audio
In Audio Kit, combine (compressed Audio -> uncompressed Audio) and (uncompressed Audio (format A) -> uncompressed Audio (format B)) into one method
func convertToPCM
In Audio Kit, there are 3 methods for converting Audio formats:
- Compressed audio between the conversion
func convertCompressed
- Uncompressed audio, go to compressed audio
func convertAsset
- Audio (including two kinds), to uncompressed audio
func convertToPCM
Determine whether to compress the audio
To determine whether the audio is compressed, look at the end of the file
internal func isCompressed(url: URL) -> Bool {
let ext = url.pathExtension.lowercased()
return (ext == "m4a" || ext == "mp3" || ext == "mp4" || ext == "m4v" || ext == "mpg")
}
Copy the codeimplementation
Audio (including two kinds), to uncompressed audio
The ExtAudioFile Service is used
- Configure format first
/ / to create a blank format template ASBD / / input ASBD var srcFormat = AudioStreamBasicDescription ASBD var dstFormat = () / / output AudioStreamBasicDescription() // ... // Get the ASBD from the source file SrcFormat error = ExtAudioFileGetProperty(inputFile, kExtAudioFileProperty_FileDataFormat, &thePropertySize, &srcFormat) // ... / / configuration, Output of ASBD dstFormat. MSampleRate = outputSampleRate dstFormat. MFormatID = formatKey dstFormat. MChannelsPerFrame = outputChannels dstFormat.mBitsPerChannel = outputBitRate dstFormat.mBytesPerPacket = outputBytesPerPacket dstFormat.mBytesPerFrame = outputBytesPerFrame dstFormat.mFramesPerPacket = 1 dstFormat.mFormatFlags = KLinearPCMFormatFlagIsPacked | kAudioFormatFlagIsSignedInteger / / visible, AIFF is compressed audio formats, PCM big end up if the format = = kAudioFileAIFFType {dstFormat. MFormatFlags = dstFormat. MFormatFlags | kLinearPCMFormatFlagIsBigEndian }Copy the code- Reconversion data
A while loop that reads to the end, numFrames 0
Use ExtAudioFileRead to read data,
Write data using ExtAudioFileWrite
var srcBuffer = [UInt8](repeating: 0, count: Int(bufferByteSize)) var sourceFrameOffset: UInt32 = 0 srcBuffer.withUnsafeMutableBytes { ptr in while true { let mBuffer = AudioBuffer(mNumberChannels: srcFormat.mChannelsPerFrame, mDataByteSize: bufferByteSize, mData: ptr.baseAddress) var fillBufList = AudioBufferList(mNumberBuffers: 1, mBuffers: mBuffer) var numFrames: UInt32 = 0 if dSTformat. mBytesPerFrame > 0 {numFrames = bufferByteSize/dstFormat.mBytesPerFrame} // Read data error = ExtAudioFileRead(inputFile, &numFrames, &fillBufList) if error ! = noErr { completionHandler? (createError(message: "Unable to read input file.")) return} // Unable to read input file. If numFrames == 0 {error = noErr break} sourceFrameOffset += numFrames ExtAudioFileWrite(outputFile, numFrames, &fillBufList) if error ! = noErr { completionHandler? (createError(message: "Unable to write output file.")) return } } }Copy the codeThe remaining two conversion methods can, of course, be usedExtAudioFile Service
Audio Kit does not use ExtAudioFile Service in the other two conversion methods
The code that Audio Kit uses is much simpler
Compressed audio format between the conversion, the code is very simple
Use the AVAssetExportSession
let asset = AVURLAsset(url: inputURL) guard let session = AVAssetExportSession(asset: asset, presetName: AVAssetExportPresetAppleM4A) else { return } session.outputURL = outputURL session.outputFileType = .m4a session.exportAsynchronously { completionHandler? (nil) }Copy the codeOne more. Uncompressed audio. Go to compressed audio
- Get avasse Reader and Avasse Riter ready
let outputFormat = options? .format ?? outputURL.pathExtension.lowercased() let asset = AVAsset(url: inputURL) do { self.reader = try AVAssetReader(asset: asset) } catch let err as NSError { return } switch outputFormat { case "m4a", "mp4": format = .m4a formatKey = kAudioFormatMPEG4AAC case "aif": format = .aiff formatKey = kAudioFormatLinearPCM case "caf": format = .caf formatKey = kAudioFormatLinearPCM case "wav": format = .wav formatKey = kAudioFormatLinearPCM default: print("Unsupported output format: \(outputFormat)") return } do { self.writer = try AVAssetWriter(outputURL: outputURL, fileType: format) } catch let err as NSError { return }Copy the code- use
AVAssetReaderTrackOutputComplete the transformation
Let writerInput = AVAssetWriterInput(mediaType:. Audio, outputSettings: outputSettings) writer.add(writerInput) guard let track = asset.tracks(withMediaType: .audio). First else {return} let readerOutput = AVAssetReaderTrackOutput(track: track, outputSettings: nil) guard reader.canAdd(readerOutput) else { return } reader.add(readerOutput) if ! writer.startWriting() { return } writer.startSession(atSourceTime: CMTime.zero) if ! reader.startReading() { return } let queue = DispatchQueue(label: "Com. Audiodesigndesk. FormatConverter. ConvertAsset") / / thread, perform the audio format conversion writerInput. RequestMediaDataWhenReady (on: queue, using: { var processing = true while writerInput.isReadyForMoreMediaData, processing { if reader.status == .reading, let buffer = readerOutput.copyNextSampleBuffer() { writerInput.append(buffer) } else { writerInput.markAsFinished() switch reader.status { case .failed: print("Conversion failed with error", reader.error ?? "") writer.cancelWriting() case .cancelled: print("Conversion cancelled") case .completed: // writer.endSession(atSourceTime: asset.duration) writer.finishWriting { } default: break } processing = false } } })Copy the codeThe effect
Compressed audio, uncompressed audio, MP 3 to WAV
Original file MP3, 7.6m, bit rate 320 KB /s
ffprobe -hide_banner -print_format json 1_Le\ Papillon.mp3 { Input #0, mp3, from '1_Le Papillon.mp3': Metadata: Encoder: Lavf57.83.100 Duration: 00:03:08.73, start: 0.025057, bitrate: 320 KB /s Stream #0:0: Audio: Stereo, 44100 Hz, FLTP, 320 KB /s Metadata: encoder: Lavc57.10}Copy the codeThe converted WAV, 66.6m, bit rate 2822 KB /s
Ffprobe-hide_banner -print_format json one.wav {Input #0, wav, from 'one.wav': Duration: 00:03:08.71, bitrate: 2822 kb/s Stream #0:0: Audio: pcm_s32le ([1][0][0][0] / 0x0001), 44100 Hz, 2 channels, s32, 2822 kb/s }Copy the codeUncompressed audio, go to compressed audio, WAV to M4a
The input is the output of the previous one.
The output is 3.1m, bit rate 129 KB /s
ffprobe -hide_banner -print_format json one.m4a { Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'one.m4a': Metadata: major_brand : M4A minor_version : 0 compatible_brands: M4A isommp42 creation_time : The 2020-12-14 T08:05:13. 000000 z iTunSMPB: 00000000 00000840 0000004C 00000000007EFB74 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 Duration: 00:03:08.71, start: 0.047891, bitrate: 129 KB /s Stream #0:0(und): Audio: aac (LC) (mp4a / 0x6134706D), 44100 Hz, stereo, fltp, 127 kb/s (default) Metadata: creation_time : 2020-12-14T08:05:13.000000z Handler_name: Core Media Audio}Copy the codeCompressed audio format conversion between MP3 to M4A
The input is the raw MP3, the input for the first step
The output is 6.1m and the bit rate is 256 KB /s
ffprobe -hide_banner -print_format json two.m4a { Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'two.m4a': Metadata: major_brand : M4A minor_version : 0 compatible_brands: M4A isommp42 creation_time : The 2020-12-14 T08: outrebounded SAN Antonio 44-36. 000000 z iTunSMPB: 00000000 00000840 0000004C 00000000007EFB74 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 Duration: 00:03:08.71, start: 0.047891, bitrate: 256 KB /s Stream #0:0(und): Audio: aac (LC) (mp4a / 0x6134706D), 44100 Hz, stereo, fltp, 254 kb/s (default) Metadata: creation_time : 2020-12-14T08:44:36.000000z Handler_name: Core Media Audio}Copy the code