转载：http://blog.csdn.net/dancing_night/article/details/53080819
Wav文件直接反映了一个声音在每个时刻的大小值，比如说以下一段波形： 我们按每人0.1秒取一点，得到的wav文件数值就是0,1,1,-1,0,1。因此，假如我们能把许多Wav文件的数据直接相加，你听到的就是所有的声音，这就是混音器的原理。
Step 1, Get the Raw data of the two files, (Example, of the sample 8bit and 8Kh, means one sample is of 8bit) Step 2 Let the two audio signal be A and B respectively, the range is between 0 and 255. Where A and B are the Sample Values (Each raw data) And store the resultant into the Y If Both the samples Values are possitv Y = A + B - A * B / 255 Where Y is the resultant signal which contains both signal A and B, merging two audio streams into single stream by this method solves the problem of overflow and information loss to an extent. If the range of 8-bit sampling is between -127 to 128
If both A and B are negative Y = A +B - (A * B / (-127)) Else Y = A + B - A * B / 128
Similarly for the nbit (ex 16bit data) For n-bit sampling audio signal If both A and B are negative Y = A + B - (A * B / (-(2 pow(n-1) -1))) Else Y = A + B - (A * B / (2 pow(n-1)) Step 3. Add the Header to the Resultant (mixed) data and play back. If some thing is unclear and ambigious let me know. Regards Ranjeet Gupta.
还有简单C程序示意代码，但是其中包含了核心算法：
#include #include #include #include int main(int argc,char *argv[]) { char mixname[255]; FILE *pcm1, *pcm2, *mix; char sample1, sample2; int value; pcm1 = fopen(argv[1],"r"); pcm2 = fopen(argv[2],"r"); strcpy (mixname, argv[1]); strcat (mixname, "_temp.wav"); mix = fopen(mixname, "w"); while(!feof(pcm1)) { sample1 = fgetc(pcm1); sample2 = fgetc(pcm2); if ((sample1 < 0) && (sample2 < 0)) { value = sample1 + sample2 - (sample1 * sample2 / -(pow(2,16-1)-1)); }else{ value = sample1 + sample2 - (sample1 * sample2 / (pow(2,16-1)-1)); } fputc(value, mix); } fclose(pcm1); fclose(pcm2); fclose(mix); return 0; } 自己的混音(混音麦克风和扬声器)：16位的数据，双声道 //将PCM叠加 for (int i = 0; i < oAcc->frame_size*2; i=i+2) { uint8_t* pMicOut = frame_audioMicOut->extended_data[0] + i; uint8_t* pMicIn = frame_audioMicIn->extended_data[0] + i; short tempMicOut = *(short*)pMicOut; short tempMicIn = *(short*)pMicIn; int tempOut = 0; if (tempMicOut < 0 && tempMicIn < 0) tempOut = tempMicOut + tempMicIn - tempMicOut*tempMicIn / (-(pow(2, 15) - 1)); else if (tempMicOut > 0 && tempMicIn > 0) tempOut = tempMicOut + tempMicIn - tempMicOut*tempMicIn / (pow(2, 15)); pMicIn = (uint8_t*)tempOut; }
优点：不会产生溢出，噪音较小； 缺点：衰减过大，影响通话质量；
short remix(short buffer1,short buffer2) { int value = buffer1 + buffer2; return (short)(value/2); } 12345 12345使用更多的位数(32 bit)来表示音频数据的一个样本，混完音后在想办法降低其振幅，使其仍旧分布在16 bit所能表示的范围之内，这种方法叫做归一法.
为避免发生溢出，使用一个可变的衰减因子对语音进行衰减。这个衰减因子也就代表语音的权重，衰减因子随着音频数据的变化而变化，所以称为自适应加权混音。当溢出时，衰减因子较小，使得溢出的数据在衰减后能够处于临界值以内，而在没有溢出时，又让衰减因子慢慢增大，使数据较为平缓的变化. 代码:
void Mix(char sourseFile[10][SIZE_AUDIO_FRAME],int number,char *objectFile) { //归一化混音 int const MAX=32767; int const MIN=-32768; double f=1; int output; int i = 0,j = 0; for (i=0;i<SIZE_AUDIO_FRAME/2;i++) { int temp=0; for (j=0;j<number;j++) { temp+=*(short*)(sourseFile[j]+i*2); } output=(int)(temp*f); if (output>MAX) { f=(double)MAX/(double)(output); output=MAX; } if (output<MIN) { f=(double)MIN/(double)(output); output=MIN; } if (f<1) { f+=((double)1-f)/(double)32; } *(short*)(objectFile+i*2)=(short)output; } } 12345678910111213141516171819202122232425262728293031323334 12345678910111213141516171819202122232425262728293031323334可以把各个通道的声音叠到一起，让声音的采样率按倍增加，如果提高声音的播放频率，声音可以正常的播放，声音实现了叠加；如果不想修改声音的播放输出频率，可以通过声音的重采样后输出自己想要的输出频率；
