AudioToolbox--利用AudioQueue音频队列,通过缓存对声音进行采集与播放

都说iOS最恶心的部分是流媒体,其中恶心的恶心之处更在即时语音。

所以我们先不谈即时语音,研究一下,iOS中声音采集与播放的实现。

要在iOS设备上实现录音和播放功能,苹果提供了简单的做法,那就是利用AVAudioRecorder和AVAudioPlayer。度娘大多数 也是如此。但是这种方法有很大的局限性。单说说这种做法:录音,首先得设置录音文件路径,然后录音数据直接写入了文件。播放也是首先给出文件路径,等到音 频整个加载完成了,才能开始播放。这相当不灵活。

我的做法是利用音频队列AudioQueue,将声音暂存至缓冲区,然后从缓冲区取出音频数据,进行播放。

声音采集:

使用AudioQueue框架以队列的形式处理音频数据。因此使用时需要给队列分配缓存空间,由回调(Callback)函数完成向队列缓存读写音频数据的功能。

一个Recording Audio Queue,包括Buffer(缓冲器)组成的Buffer Queue(缓冲队列),以及一个Callback(回调)。实现主要步骤为:

  1. 设置音频的参数
  2. 准备并启动声音采集的音频队列
  3. 在回调函数中处理采集到的音频Buffer,在这里是暂存在了一个Byte数组里,提供给播放端使用
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Record.h
#import <Foundation/Foundation.h>
#import <AudioToolbox/AudioToolbox.h>
#import <CoreAudio/CoreAudioTypes.h>
#import "AudioConstant.h"
 
// use Audio Queue
 
typedef struct AQCallbackStruct
{
    AudioStreamBasicDescription mDataFormat;
    AudioQueueRef               queue;
    AudioQueueBufferRef         mBuffers[kNumberBuffers];
    AudioFileID                 outputFile;
     
    unsigned long               frameSize;
    long long                   recPtr;
    int                         run;
     
} AQCallbackStruct;
 
 
@interface Record : NSObject
{
    AQCallbackStruct aqc;
    AudioFileTypeID fileFormat;
    long audioDataLength;
    Byte audioByte[999999];
    long audioDataIndex;
}
- (id) init;
- (void) start;
- (void) stop;
- (void) pause;
- (Byte *) getBytes;
- (void) processAudioBuffer:(AudioQueueBufferRef) buffer withQueue:(AudioQueueRef) queue;
 
@property (nonatomic, assign) AQCallbackStruct aqc;
@property (nonatomic, assign) long audioDataLength;
@end

 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
Record.mm
#import "Record.h"
 
@implementation Record
@synthesize aqc;
@synthesize audioDataLength;
 
static void AQInputCallback (void                   * inUserData,
                             AudioQueueRef          inAudioQueue,
                             AudioQueueBufferRef    inBuffer,
                             const AudioTimeStamp   * inStartTime,
                             unsigned long          inNumPackets,
                             const AudioStreamPacketDescription * inPacketDesc)
{
     
    Record * engine = (__bridge Record *) inUserData;
    if (inNumPackets > 0)
    {
        [engine processAudioBuffer:inBuffer withQueue:inAudioQueue];
    }
     
    if (engine.aqc.run)
    {
        AudioQueueEnqueueBuffer(engine.aqc.queue, inBuffer, 0, NULL);
    }
}
 
- (id) init
{
    self = [super init];
    if (self)
    {
        aqc.mDataFormat.mSampleRate = kSamplingRate;
        aqc.mDataFormat.mFormatID = kAudioFormatLinearPCM;
        aqc.mDataFormat.mFormatFlags = kLinearPCMFormatFlagIsSignedInteger |kLinearPCMFormatFlagIsPacked;
        aqc.mDataFormat.mFramesPerPacket = 1;
        aqc.mDataFormat.mChannelsPerFrame = kNumberChannels;
        aqc.mDataFormat.mBitsPerChannel = kBitsPerChannels;
        aqc.mDataFormat.mBytesPerPacket = kBytesPerFrame;
        aqc.mDataFormat.mBytesPerFrame = kBytesPerFrame;
        aqc.frameSize = kFrameSize;
         
        AudioQueueNewInput(&aqc.mDataFormat, AQInputCallback, (__bridge void *)(self), NULL, kCFRunLoopCommonModes,0, &aqc.queue);
         
        for (int i=0;i<kNumberBuffers;i++)
        {
            AudioQueueAllocateBuffer(aqc.queue, aqc.frameSize, &aqc.mBuffers[i]);
            AudioQueueEnqueueBuffer(aqc.queue, aqc.mBuffers[i], 0, NULL);
        }
        aqc.recPtr = 0;
        aqc.run = 1;
    }
    audioDataIndex = 0;
    return self;
}
 
- (void) dealloc
{
    AudioQueueStop(aqc.queue, true);
    aqc.run = 0;
    AudioQueueDispose(aqc.queue, true);
}
 
- (void) start
{
    AudioQueueStart(aqc.queue, NULL);
}
 
- (void) stop
{
    AudioQueueStop(aqc.queue, true);
}
 
- (void) pause
{
    AudioQueuePause(aqc.queue);
}
 
- (Byte *)getBytes
{
    return audioByte;
}
 
- (void) processAudioBuffer:(AudioQueueBufferRef) buffer withQueue:(AudioQueueRef) queue
{
    NSLog(@"processAudioData :%ld", buffer->mAudioDataByteSize);
    //处理data:忘记oc怎么copy内存了,于是采用的C++代码,记得把类后缀改为.mm。同Play
    memcpy(audioByte+audioDataIndex, buffer->mAudioData, buffer->mAudioDataByteSize);
    audioDataIndex +=buffer->mAudioDataByteSize;
    audioDataLength = audioDataIndex;
}
 
@end

声音播放:

同采集一样,播放主要步骤如下:

  1. 设置音频参数(需和采集时设置参数一样)
  2. 取得缓存的音频Buffer
  3. 准备并启动声音播放的音频队列
  4. 在回调函数中处理Buffer
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Play.h
#import <Foundation/Foundation.h>
#import <AudioToolbox/AudioToolbox.h>
 
#import "AudioConstant.h"
 
@interface Play : NSObject
{
    //音频参数
    AudioStreamBasicDescription audioDescription;
    // 音频播放队列
    AudioQueueRef audioQueue;
    // 音频缓存
    AudioQueueBufferRef audioQueueBuffers[QUEUE_BUFFER_SIZE];
}
 
-(void)Play:(Byte *)audioByte Length:(long)len;
 
@end

 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
Play.mm
 
#import "Play.h"
 
@interface Play()
{
    Byte *audioByte;
    long audioDataIndex;
    long audioDataLength;
}
@end
 
@implementation Play
 
//回调函数(Callback)的实现
static void BufferCallback(void *inUserData,AudioQueueRef inAQ,AudioQueueBufferRef buffer){
     
    NSLog(@"processAudioData :%u", (unsigned int)buffer->mAudioDataByteSize);
     
    Play* player=(__bridge Play*)inUserData;
     
    [player FillBuffer:inAQ queueBuffer:buffer];
}
 
//缓存数据读取方法的实现
-(void)FillBuffer:(AudioQueueRef)queue queueBuffer:(AudioQueueBufferRef)buffer
{
    if(audioDataIndex + EVERY_READ_LENGTH < audioDataLength)
    {
        memcpy(buffer->mAudioData, audioByte+audioDataIndex, EVERY_READ_LENGTH);
        audioDataIndex += EVERY_READ_LENGTH;
        buffer->mAudioDataByteSize =EVERY_READ_LENGTH;
        AudioQueueEnqueueBuffer(queue, buffer, 0, NULL);
    }
     
}
 
-(void)SetAudioFormat
{
    ///设置音频参数
    audioDescription.mSampleRate  = kSamplingRate;//采样率
    audioDescription.mFormatID    = kAudioFormatLinearPCM;
    audioDescription.mFormatFlags =  kAudioFormatFlagIsSignedInteger;//|kAudioFormatFlagIsNonInterleaved;
    audioDescription.mChannelsPerFrame = kNumberChannels;
    audioDescription.mFramesPerPacket  = 1;//每一个packet一侦数据
    audioDescription.mBitsPerChannel   = kBitsPerChannels;//av_get_bytes_per_sample(AV_SAMPLE_FMT_S16)*8;//每个采样点16bit量化
    audioDescription.mBytesPerFrame    = kBytesPerFrame;
    audioDescription.mBytesPerPacket   = kBytesPerFrame;
     
    [self CreateAudioQueue];
}
 
-(void)CreateAudioQueue
{
    [self Cleanup];
    //使用player的内部线程播
    AudioQueueNewOutput(&audioDescription, BufferCallback, (__bridge void *)(self), nil, nil, 0, &audioQueue);
    if(audioQueue)
    {
        ////添加buffer区
        for(int i=0;i<QUEUE_BUFFER_SIZE;i++)
        {
            int result =  AudioQueueAllocateBuffer(audioQueue, EVERY_READ_LENGTH, &audioQueueBuffers[i]);
            ///创建buffer区,MIN_SIZE_PER_FRAME为每一侦所需要的最小的大小,该大小应该比每次往buffer里写的最大的一次还大
            NSLog(@"AudioQueueAllocateBuffer i = %d,result = %d",i,result);
        }
    }
}
 
-(void)Cleanup
{
    if(audioQueue)
    {
        NSLog(@"Release AudioQueueNewOutput");
         
        [self Stop];
        for(int i=0; i < QUEUE_BUFFER_SIZE; i++)
        {
            AudioQueueFreeBuffer(audioQueue, audioQueueBuffers[i]);
            audioQueueBuffers[i] = nil;
        }
        audioQueue = nil;
    }
}
 
-(void)Stop
{
    NSLog(@"Audio Player Stop");
     
    AudioQueueFlush(audioQueue);
    AudioQueueReset(audioQueue);
    AudioQueueStop(audioQueue,TRUE);
}
 
-(void)Play:(Byte *)byte Length:(long)len
{
    [self Stop];
    audioByte = byte;
    audioDataLength = len;
     
    NSLog(@"Audio Play Start >>>>>");
     
    [self SetAudioFormat];
     
    AudioQueueReset(audioQueue);
    audioDataIndex = 0;
    for(int i=0; i<QUEUE_BUFFER_SIZE; i++)
    {
        [self FillBuffer:audioQueue queueBuffer:audioQueueBuffers[i]];
    }
    AudioQueueStart(audioQueue, NULL);
}
 
@end

以上,实现了通过内存缓存,声音的采集和播放,包括了声音采集,暂停,结束,播放等主要流程。

PS:由于本人水品有限加之这方面资料较少,只跑通了正常流程,暂时没做异常处理。采集的声音Buffer限定大小每次只有十来秒钟的样子,这个留给需要的人自己去优化了。

demo的下载地址:demo

 

来源:http://www.cnblogs.com/anjohnlv/p/3383908.html

你可能感兴趣的:(Queue)