如何将AudioBufferList转换为CMSampleBuffer？

Question

我有一个连接到AVPlayerItem的AudioTapProcessor。它将在处理时调用static void tap_ProcessCallback(MTAudioProcessingTapRef tap, CMItemCount numberFrames, MTAudioProcessingTapFlags flags, AudioBufferList *bufferListInOut, CMItemCount *numberFramesOut, MTAudioProcessingTapFlags *flagsOut)。

我需要将AudioBufferList转换为CMSampleBuffer，这样我才能使用AVAssetWriterAudioInput.appendSampleBuffer将其写入电影文件。

那么如何将AudioBufferList转换为CMSampleBuffer呢？我试过了，但得到了-12731错误:Error :Optional("-12731")

func processAudioData(audioData: UnsafeMutablePointer<AudioBufferList>, framesNumber: UInt32) {
    var sbuf : Unmanaged<CMSampleBuffer>?
    var status : OSStatus?
    var format: Unmanaged<CMFormatDescription>?

    var formatId =  UInt32(kAudioFormatLinearPCM)
    var formatFlags = UInt32( kAudioFormatFlagIsSignedInteger | kAudioFormatFlagIsPacked )
    var audioFormat = AudioStreamBasicDescription(mSampleRate: 44100.00, mFormatID:formatId, mFormatFlags:formatFlags , mBytesPerPacket: 1, mFramesPerPacket: 1, mBytesPerFrame: 16, mChannelsPerFrame: 2, mBitsPerChannel: 2, mReserved: 0)

    status = CMAudioFormatDescriptionCreate(kCFAllocatorDefault, &audioFormat, 0, nil, 0, nil, nil, &format)
    if status != noErr {
        println("Error CMAudioFormatDescriptionCreater :\(status?.description)")
        return
    }


    var timing = CMSampleTimingInfo(duration: CMTimeMake(1, 44100), presentationTimeStamp: kCMTimeZero, decodeTimeStamp: kCMTimeInvalid)

    status = CMSampleBufferCreate(kCFAllocatorDefault,nil,Boolean(0),nil,nil,format?.takeRetainedValue(), CMItemCount(framesNumber), 1, &timing, 0, nil, &sbuf);
    if status != noErr {
        println("Error CMSampleBufferCreate :\(status?.description)")
        return
    }
    status =   CMSampleBufferSetDataBufferFromAudioBufferList(sbuf?.takeRetainedValue(), kCFAllocatorDefault , kCFAllocatorDefault, 0, audioData)
    if status != noErr {
        println("Error cCMSampleBufferSetDataBufferFromAudioBufferList :\(status?.description)")
        return
    }

    var currentSampleTime = CMSampleBufferGetOutputPresentationTimeStamp(sbuf?.takeRetainedValue());
    println(" audio buffer at time: \(CMTimeCopyDescription(kCFAllocatorDefault, currentSampleTime))")

    if !assetWriterAudioInput!.readyForMoreMediaData {
        return
    }else if assetWriter.status == .Writing {

        if !assetWriterAudioInput!.appendSampleBuffer(sbuf?.takeRetainedValue()) {
            println("Problem appending audio buffer at time: \(CMTimeCopyDescription(kCFAllocatorDefault, currentSampleTime))")
        }

    }else{
        println("assetWriterStatus:\(assetWriter.status.rawValue), Error: \(assetWriter.error.localizedDescription)")
        println("Could not write a frame")
    }




}

Answer 1

好了，我已经成功地解决了这个问题。

问题是我不应该自己构造AudioStreamBasicDescription结构。但请使用AudioProcessorTap的准备回调提供的回调。

static void tap_PrepareCallback(MTAudioProcessingTapRef tap, CMItemCount maxFrames, const AudioStreamBasicDescription *processingFormat) //retain this one