如何在 iOS 中添加带有 PCM 数据/缓冲区的可播放(如 wav、wmv)标头?

Posted

技术标签:

【中文标题】如何在 iOS 中添加带有 PCM 数据/缓冲区的可播放(如 wav、wmv)标头?【英文标题】:How to add playable(such as wav,wmv) header with PCM data/buffer in iOS? 【发布时间】:2015-01-30 07:28:06 【问题描述】:

我正在尝试在原始 PCM 数据之上添加一个 wav 标头,以使其可通过 AVAudioPlayer 播放。但我在 ios 上使用 Objective-C/Swift 找不到任何解决方案或源代码。虽然我找到了this,但它没有正确答案。

但我发现了一段代码here,它在 C 中,也包含一些问题。从该代码生成的 wav 文件无法正常播放。

我已经给出了我迄今为止编码的代码。

int NumChannels = AUDIO_CHANNELS_PER_FRAME;
short BitsPerSample = AUDIO_BITS_PER_CHANNEL;
int SamplingRate = AUDIO_SAMPLE_RATE;
int numOfSamples = [[NSData dataWithContentsOfFile:filePath] length];

int ByteRate = NumChannels*BitsPerSample*SamplingRate/8;
short BlockAlign = NumChannels*BitsPerSample/8;
int DataSize = NumChannels*numOfSamples*BitsPerSample/8;
int chunkSize = 16;
int totalSize = 36 + DataSize;
short audioFormat = 1;

if((fout = fopen([wavFilePath cStringUsingEncoding:1], "w")) == NULL)

    printf("Error opening out file ");


fwrite("RIFF", sizeof(char), 4,fout);
fwrite(&totalSize, sizeof(int), 1, fout);
fwrite("WAVE", sizeof(char), 4, fout);
fwrite("fmt ", sizeof(char), 3, fout);
fwrite(&chunkSize, sizeof(int),1,fout);
fwrite(&audioFormat, sizeof(short), 1, fout);
fwrite(&NumChannels, sizeof(short),1,fout);
fwrite(&SamplingRate, sizeof(int), 1, fout);
fwrite(&ByteRate, sizeof(int), 1, fout);
fwrite(&BlockAlign, sizeof(short), 1, fout);
fwrite(&BitsPerSample, sizeof(short), 1, fout);
fwrite("data", sizeof(char), 3, fout);
fwrite(&DataSize, sizeof(int), 1, fout);

文件播放太快,声音失真,仅播放前 10 到 20(大约)秒。我认为,wav 标头没有正确生成(因为我能够使用 AudioUnit/AudioQueue 播放相同的 PCM 数据/缓冲区)。那么我的代码中缺少什么?任何帮助将不胜感激。

提前致谢。

【问题讨论】:

没人能回答我的问题???!!!我只是很惊讶...有人请回复.... iam 也面临同样的问题..您找到解决此问题的方法了吗? 【参考方案1】:

好的,如果它对其他人有帮助,我正在回答我自己的问题。经过几天的不懈努力,我终于让它工作了。下面是一个用 Objective-C 和 C 编写的完整函数。它将文件路径作为参数,其中包含直接从麦克风捕获的 RAW PCM 数据,并返回包含 PCM 数据和适当的 wav 头信息的文件路径。然后您可以使用 AVAudioPlayer 或 AVPlayer 播放该文件。这是代码...

- (NSURL *) getAndCreatePlayableFileFromPcmData:(NSString *)filePath

    NSString *wavFileName = [[filePath lastPathComponent] stringByDeletingPathExtension];
    NSString *wavFileFullName = [NSString stringWithFormat:@"%@.wav",wavFileName];

    [self createFileWithName:wavFileFullName];
    NSArray *dirPaths = NSSearchPathForDirectoriesInDomains(NSDocumentDirectory, NSUserDomainMask, YES);
    NSString *docsDir = [dirPaths objectAtIndex:0];
    NSString *wavFilePath = [docsDir stringByAppendingPathComponent:wavFileFullName];

    NSLog(@"PCM file path : %@",filePath);

    FILE *fout;

    short NumChannels = AUDIO_CHANNELS_PER_FRAME;
    short BitsPerSample = AUDIO_BITS_PER_CHANNEL;
    int SamplingRate = AUDIO_SAMPLE_RATE;
    int numOfSamples = [[NSData dataWithContentsOfFile:filePath] length];

    int ByteRate = NumChannels*BitsPerSample*SamplingRate/8;
    short BlockAlign = NumChannels*BitsPerSample/8;
    int DataSize = NumChannels*numOfSamples*BitsPerSample/8;
    int chunkSize = 16;
    int totalSize = 46 + DataSize;
    short audioFormat = 1;

    if((fout = fopen([wavFilePath cStringUsingEncoding:1], "w")) == NULL)
    
        printf("Error opening out file ");
    

    fwrite("RIFF", sizeof(char), 4,fout);
    fwrite(&totalSize, sizeof(int), 1, fout);
    fwrite("WAVE", sizeof(char), 4, fout);
    fwrite("fmt ", sizeof(char), 4, fout);
    fwrite(&chunkSize, sizeof(int),1,fout);
    fwrite(&audioFormat, sizeof(short), 1, fout);
    fwrite(&NumChannels, sizeof(short),1,fout);
    fwrite(&SamplingRate, sizeof(int), 1, fout);
    fwrite(&ByteRate, sizeof(int), 1, fout);
    fwrite(&BlockAlign, sizeof(short), 1, fout);
    fwrite(&BitsPerSample, sizeof(short), 1, fout);
    fwrite("data", sizeof(char), 4, fout);
    fwrite(&DataSize, sizeof(int), 1, fout);

    fclose(fout);

    NSMutableData *pamdata = [NSMutableData dataWithContentsOfFile:filePath];
    NSFileHandle *handle;
    handle = [NSFileHandle fileHandleForUpdatingAtPath:wavFilePath];
    [handle seekToEndOfFile];
    [handle writeData:pamdata];
    [handle closeFile];

    return [NSURL URLWithString:wavFilePath];

但该功能仅适用于以下音频设置。

// Audio settings.
#define AUDIO_SAMPLE_RATE 8000
#define AUDIO_FRAMES_PER_PACKET 1
#define AUDIO_CHANNELS_PER_FRAME 1
#define AUDIO_BITS_PER_CHANNEL 16
#define AUDIO_BYTES_PER_PACKET 2
#define AUDIO_BYTES_PER_FRAME 2

【讨论】:

什么是 [self createFileWithName:wavFileFullName];? 非常有帮助。谢谢! 首先非常感谢您发布伟大的代码,这对我很有帮助,您能告诉我如何从波形文件中获取 PCM 数据吗? iam 也面临同样的问题。您找到解决方案了吗? 谢谢!非常方便,正是我需要的。【参考方案2】:

非常有帮助的问题和答案,非常感谢。

这个快速版本是为有需要的人准备的:

static func createWAV(from pcmFilePath: String, to wavFilePath: String) -> Bool 

    // Make sure that the path does not contain non-ascii characters
    guard let fout = fopen(wavFilePath.cString(using: .ascii), "w") else  return false 

    guard let pcmData = try? Data(contentsOf: URL(fileURLWithPath: pcmFilePath)) else  return false 

    var numChannels: CShort = 1
    let numChannelsInt: CInt = 1
    var bitsPerSample: CShort = 16
    let bitsPerSampleInt: CInt = 16
    var samplingRate: CInt = 16000
    let numOfSamples = CInt(pcmData.count)
    var byteRate = numChannelsInt * bitsPerSampleInt * samplingRate / 8
    var blockAlign = numChannelsInt * bitsPerSampleInt / 8
    var dataSize = numChannelsInt * numOfSamples * bitsPerSampleInt / 8
    var chunkSize: CInt = 16
    var totalSize = 46 + dataSize
    var audioFormat: CShort = 1

    fwrite("RIFF".cString(using: .ascii), MemoryLayout<CChar>.size, 4, fout)
    fwrite(&totalSize, MemoryLayout<CInt>.size, 1, fout)
    fwrite("WAVE".cString(using: .ascii), MemoryLayout<CChar>.size, 4, fout);
    fwrite("fmt ".cString(using: .ascii), MemoryLayout<CChar>.size, 4, fout);
    fwrite(&chunkSize, MemoryLayout<CInt>.size,1,fout);
    fwrite(&audioFormat, MemoryLayout<CShort>.size, 1, fout);
    fwrite(&numChannels, MemoryLayout<CShort>.size,1,fout);
    fwrite(&samplingRate, MemoryLayout<CInt>.size, 1, fout);
    fwrite(&byteRate, MemoryLayout<CInt>.size, 1, fout);
    fwrite(&blockAlign, MemoryLayout<CShort>.size, 1, fout);
    fwrite(&bitsPerSample, MemoryLayout<CShort>.size, 1, fout);
    fwrite("data".cString(using: .ascii), MemoryLayout<CChar>.size, 4, fout);
    fwrite(&dataSize, MemoryLayout<CInt>.size, 1, fout);
    fclose(fout);

    guard let handle = FileHandle(forUpdatingAtPath: wavFilePath) else  return false 

    handle.seekToEndOfFile()
    handle.write(pcmData)
    handle.closeFile()

    return true

【讨论】:

【参考方案3】:

根据 quiz 对 swift 5 的回答修改

func extractSubchunks(data:Data) -> RiffFile? 
    var data = data
    var chunks = [SubChunk]()
    let position = data.subdata(in: 8..<12)
    let filelengthBytes = data.subdata(in: 4..<8).map  UInt32($0) 
    let filelength: UInt32 = filelengthBytes[0] << 24 + filelengthBytes[1] << 16 + filelengthBytes[2] << 8 + filelengthBytes[3]
    let wave = String(bytes: position, encoding: .utf8) ?? "NoName"
    guard wave == "WAVE" else 
        print("File is \(wave) not WAVE")
        return nil
    
    data.removeSubrange(0..<12)
    print("Found chunks")
    while data.count != 0
        let position = data.subdata(in: 0..<4)
        let lengthBytes = data.subdata(in: 4..<8).map  UInt32($0) 
        let length: UInt32 = lengthBytes[0] << 24 + lengthBytes[1] << 16 + lengthBytes[2] << 8 + lengthBytes[3]
        guard let current = String(bytes: position, encoding: .utf8) else
            return nil
        
        data.removeSubrange(0..<8)
        let chunkData = data.subdata(in: 0..<Int(length))
        data.removeSubrange(0..<Int(length))
        let subchunk = SubChunk(name: current, size: Int(length), data: chunkData)
        chunks.append(subchunk)
        print(subchunk.debugDescription)
    
    let riff = RiffFile(size: Int(filelength), subChunks: chunks)
    return riff

【讨论】:

以上是关于如何在 iOS 中添加带有 PCM 数据/缓冲区的可播放(如 wav、wmv)标头?的主要内容,如果未能解决你的问题,请参考以下文章

iOS录制时如何将PCM缓冲区实时转换为AAC数据?

如何正确使用 iOS AudioUnit 渲染回调

如何将 PCM 音频流转换为在线播放

如何仅使用 AVAudioPlayer 或 AVPlayer 播放 PCM 数据/缓冲区?

在 iOS 中实时将 pcm 样本编码为 mp3,可能吗?

如何播放来自服务器的流式 PCM 数据,如 Android AudioTrack 类?