參考自：https://blog.csdn.net/u012323667/article/details/79214336
https://blog.csdn.net/szfhy/article/details/52448906
G711也稱為PCM（脈沖編碼調制），是國際電信聯盟制定出來的一套語音壓縮标準，主要用于電話。G711編碼的聲音清晰度好，語音自然度高，但壓縮效率低，資料量大常在32Kbps以上(推薦使用64Kbps)。它主要用脈沖編碼調制對音頻采樣，采樣率為8KHz。它利用一個64Kbps未壓縮通道傳輸語音訊号。
其壓縮率為1：2，即把16位資料壓縮成8位
。G.711是主流的波形聲音編解碼器。
G.711 标準下主要有兩種壓縮算法。一種是µ-law algorithm （又稱often u-law, ulaw, mu-law），主要運用于北美和日本；另一種是a-law algorithm，主要運用于歐洲和世界其他地區。其中，後者是特别設計用來友善計算機處理的。這兩種算法都使用一個采樣率為8kHz的輸入來建立64Kbps的數字輸出。
1、a-law
a-law也叫g711a，輸入的是13位（
其實是S16的高13位
），使用在歐洲和其他地區，這種格式是經過特别設計的，便于數字裝置進行快速運算。在WAV檔案中的識别标志是
WAVE_FORMAT_ALAW
。
目前最主要的用途是将13bit的資料轉化成為8bit的資料，13bit的最高位是符号位，轉化成為的8bit并不是線性的，而是每幾位都有其意義。
- 第7位： 代表符号位，1表示正數，0表示負數，這點和一般的計算機系統正好相反
- 第4-6位： 這實際上是一個表示一種幂級數的位
- 第0-3位： 是具體的量化數值
運算過程如下：
（1） 取符号位并取反得到s，
（2） 擷取強度位eee，擷取方法如圖所示
（3） 擷取高位樣本位wxyz
（4） 組合為seeewxyz，将seeewxyz逢偶數位取反，編碼完畢
示例：
輸入pcm資料為3210，二進制對應為（0000 1100 1000 1010）
二進制變換下排列組合方式（0 0001 1001 0001010）
（1） 擷取符号位最高位為0，取反，s=1
（2） 擷取強度位0001，查表，編碼制應該是eee=100
（3） 擷取高位樣本wxyz=1001
（4） 組合為11001001，逢偶數位取反為10011100
編碼完畢。
編碼代碼如下：
#define MAX (32635)
void encode(unsigned char *dst, short *src, size_t len)
{
for(int i = 0; i < len ; i++)
{
// *dst++ = *src++;
short pcm = *src++;
int sign = (pcm & 0x8000) >> 8;
if(sign != 0)
pcm = -pcm;
if(pcm > MAX) pcm = MAX;
int exponent = 7;
int expMask;
for(expMask = 0x4000; (pcm & expMask) == 0 && exponent >0; exponent--,expMask >>= 1){}
int mantissa = (pcm >> ((exponent == 0) ? 4 : (exponent + 3))) & 0x0f;
unsigned char alaw = (unsigned char)(sign | exponent << 4 | mantissa);
*dst++ = (unsigned char)(alaw ^0xD5);
}
}
譯碼代碼如下：
void decode(short *dst, unsigned char *src, size_t len)
{
for(size_t i=0; i < len ; i++)
{
unsigned char alaw = *src++;
alaw ^= 0xD5;
int sign = alaw & 0x80;
int exponent = (alaw & 0x70) >> 4;
int data = alaw & 0x0f;
data <<= 4;
data += 8; //丢失的a 寫1
if(exponent != 0) //将wxyz前面的1補上
data += 0x100;
if(exponent > 1)
data <<= (exponent - 1);
*dst++ = (short)(sign == 0 ? data : -data);
}
}
從編解碼的過程中，不難看出這種編碼方式的思路：
首先，我們确定目的，原始的碼流是13bit，除去符号位，12bit；我們需要将其轉化成為8bit，除去符号位，還有7bit。
基本是思想就是：記錄符号位，記錄下最有效資料位（也就是除去符号位第一個1）後的4位，記錄下這4位移動到最低位所需的次數（放在0-3位中），解碼的時候根據這些資訊還原，也就是最高的5位會是準确的，後面緊跟一個1是用來補償舍棄的資料，這就是誤差的來源。
但是在實際的操作中，是會直接将最後4位舍棄，然後在開始保留資料操作，也就是要是有效資料小于4位，最後全舍棄了，就會補償為8。若大于4位小于8位，最後準确的資料可能就不會有5位了。
2、µ-law
µ-law也叫g711µ，使用在北美和日本，
輸入的是14位
，編碼算法就是查表，沒啥複雜算法，就是基礎值+平均偏移值，具體示例如下：
pcm=2345
（1）取得範圍值
+4062 to +2015 in 16 intervals of 128
（2）得到基礎值0x90，
（3）間隔數128，
（4）區間基本值4062，
（5）目前值2345和區間基本值差異4062-2345=1717，
（6）偏移值=1717/間隔數=1717/128，取整得到13，
（7）輸出為0x90+13=0x9D
為了簡化編碼過程，原始的線性幅度增加了33，使得編碼範圍從(0 - 8158)變為(33 - 8191)。結果如下表所示：
| Biased Linear Input Code | Compressed Code |
|---|---|
| 00000001wxyza | 000wxyz |
| 0000001wxyzab | 001wxyz |
| 000001wxyzabc | 010wxyz |
| 00001wxyzabcd | 011wxyz |
| 0001wxyzabcde | 100wxyz |
| 001wxyzabcdef | 101wxyz |
| 01wxyzabcdefg | 110wxyz |
| 1wxyzabcdefgh | 111wxyz |
每個偏置線性輸入碼都有一個前導1來辨別段号。段号的值等于7減去前導0的個數。量化間隔的個數是直接可用的四個位wxyz。後面的位(a - h)被忽略。
由上可見，這種編碼方式和a-Law是極為類似的，是将14bit的資料編碼成為8bit的資料（最高位都是符号位），8bit的每位資料的意義和a-Law編碼是一緻的。
但是和a-Law不同的是，這是對14bit的資料進行壓縮，并且壓縮完後的資料不僅僅是偶數位取反，而是每一位都需要做取反操作
。
并且，注意到表中，這裡的最高位都是從1開始，和a-Law相比，這裡加了一位，但是seg還是3bit，是以對照兩表的第一個數是有明顯差異的，于是每個數加33是很有必要的，然後在解碼的時候減去，相當于移位就直接減少了一種情況。
注意：在g711.c代碼中，BIAS的值是0x84，該值是由33<<2得來的。
g711.c代碼如下：
/*
* This source code is a product of Sun Microsystems, Inc. and is provided
* for unrestricted use. Users may copy or modify this source code without
* charge.
*
* SUN SOURCE CODE IS PROVIDED AS IS WITH NO WARRANTIES OF ANY KIND INCLUDING
* THE WARRANTIES OF DESIGN, MERCHANTIBILITY AND FITNESS FOR A PARTICULAR
* PURPOSE, OR ARISING FROM A COURSE OF DEALING, USAGE OR TRADE PRACTICE.
*
* Sun source code is provided with no support and without any obligation on
* the part of Sun Microsystems, Inc. to assist in its use, correction,
* modification or enhancement.
*
* SUN MICROSYSTEMS, INC. SHALL HAVE NO LIABILITY WITH RESPECT TO THE
* INFRINGEMENT OF COPYRIGHTS, TRADE SECRETS OR ANY PATENTS BY THIS SOFTWARE
* OR ANY PART THEREOF.
*
* In no event will Sun Microsystems, Inc. be liable for any lost revenue
* or profits or other special, indirect and consequential damages, even if
* Sun has been advised of the possibility of such damages.
*
* Sun Microsystems, Inc.
* 2550 Garcia Avenue
* Mountain View, California 94043
*/
/*
* g711.c
*
* u-law, A-law and linear PCM conversions.
*/
#define SIGN_BIT (0x80) /* Sign bit for a A-law byte. */
#define QUANT_MASK (0xf) /* Quantization field mask. */
#define NSEGS (8) /* Number of A-law segments. */
#define SEG_SHIFT (4) /* Left shift for segment number. */
#define SEG_MASK (0x70) /* Segment field mask. */
static short seg_end[8] = {0xFF, 0x1FF, 0x3FF, 0x7FF,
0xFFF, 0x1FFF, 0x3FFF, 0x7FFF};
/* copy from CCITT G.711 specifications */
unsigned char _u2a[128] = { /* u- to A-law conversions */
1, 1, 2, 2, 3, 3, 4, 4,
5, 5, 6, 6, 7, 7, 8, 8,
9, 10, 11, 12, 13, 14, 15, 16,
17, 18, 19, 20, 21, 22, 23, 24,
25, 27, 29, 31, 33, 34, 35, 36,
37, 38, 39, 40, 41, 42, 43, 44,
46, 48, 49, 50, 51, 52, 53, 54,
55, 56, 57, 58, 59, 60, 61, 62,
64, 65, 66, 67, 68, 69, 70, 71,
72, 73, 74, 75, 76, 77, 78, 79,
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, 115, 116, 117, 118, 119, 120,
121, 122, 123, 124, 125, 126, 127, 128};
unsigned char _a2u[128] = { /* A- to u-law conversions */
1, 3, 5, 7, 9, 11, 13, 15,
16, 17, 18, 19, 20, 21, 22, 23,
24, 25, 26, 27, 28, 29, 30, 31,
32, 32, 33, 33, 34, 34, 35, 35,
36, 37, 38, 39, 40, 41, 42, 43,
44, 45, 46, 47, 48, 48, 49, 49,
50, 51, 52, 53, 54, 55, 56, 57,
58, 59, 60, 61, 62, 63, 64, 64,
65, 66, 67, 68, 69, 70, 71, 72,
73, 74, 75, 76, 77, 78, 79, 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, 115, 116, 117, 118, 119,
120, 121, 122, 123, 124, 125, 126, 127};
static int
search(
int val,
short *table,
int size)
{
int i;
for (i = 0; i < size; i++) {
if (val <= *table++)
return (i);
}
return (size);
}
/*
* linear2alaw() - Convert a 16-bit linear PCM value to 8-bit A-law
*
* linear2alaw() accepts an 16-bit integer and encodes it as A-law data.
*
* Linear Input Code Compressed Code
* ------------------------ ---------------
* 0000000wxyza 000wxyz
* 0000001wxyza 001wxyz
* 000001wxyzab 010wxyz
* 00001wxyzabc 011wxyz
* 0001wxyzabcd 100wxyz
* 001wxyzabcde 101wxyz
* 01wxyzabcdef 110wxyz
* 1wxyzabcdefg 111wxyz
*
* For further information see John C. Bellamy's Digital Telephony, 1982,
* John Wiley & Sons, pps 98-111 and 472-476.
*/
unsigned char
linear2alaw(
int pcm_val) /* 2's complement (16-bit range) */
{
int mask;
int seg;
unsigned char aval;
if (pcm_val >= 0) {
mask = 0xD5; /* sign (7th) bit = 1 */
} else {
mask = 0x55; /* sign bit = 0 */
pcm_val = -pcm_val - 8;
}
/* Convert the scaled magnitude to segment number. */
seg = search(pcm_val, seg_end, 8);
/* Combine the sign, segment, and quantization bits. */
if (seg >= 8) /* out of range, return maximum value. */
return (0x7F ^ mask);
else {
aval = seg << SEG_SHIFT;
if (seg < 2)
aval |= (pcm_val >> 4) & QUANT_MASK;
else
aval |= (pcm_val >> (seg + 3)) & QUANT_MASK;
return (aval ^ mask);
}
}
/*
* alaw2linear() - Convert an A-law value to 16-bit linear PCM
*
*/
int
alaw2linear(
unsigned char a_val)
{
int t;
int seg;
a_val ^= 0x55;
t = (a_val & QUANT_MASK) << 4;
seg = ((unsigned)a_val & SEG_MASK) >> SEG_SHIFT;
switch (seg) {
case 0:
t += 8;
break;
case 1:
t += 0x108;
break;
default:
t += 0x108;
t <<= seg - 1;
}
return ((a_val & SIGN_BIT) ? t : -t);
}
#define BIAS (0x84) /* Bias for linear code. */
/*
* linear2ulaw() - Convert a linear PCM value to u-law
*
* In order to simplify the encoding process, the original linear magnitude
* is biased by adding 33 which shifts the encoding range from (0 - 8158) to
* (33 - 8191). The result can be seen in the following encoding table:
*
* Biased Linear Input Code Compressed Code
* ------------------------ ---------------
* 00000001wxyza 000wxyz
* 0000001wxyzab 001wxyz
* 000001wxyzabc 010wxyz
* 00001wxyzabcd 011wxyz
* 0001wxyzabcde 100wxyz
* 001wxyzabcdef 101wxyz
* 01wxyzabcdefg 110wxyz
* 1wxyzabcdefgh 111wxyz
*
* Each biased linear code has a leading 1 which identifies the segment
* number. The value of the segment number is equal to 7 minus the number
* of leading 0's. The quantization interval is directly available as the
* four bits wxyz. * The trailing bits (a - h) are ignored.
*
* Ordinarily the complement of the resulting code word is used for
* transmission, and so the code word is complemented before it is returned.
*
* For further information see John C. Bellamy's Digital Telephony, 1982,
* John Wiley & Sons, pps 98-111 and 472-476.
*/
unsigned char
linear2ulaw(
int pcm_val) /* 2's complement (16-bit range) */
{
int mask;
int seg;
unsigned char uval;
/* Get the sign and the magnitude of the value. */
if (pcm_val < 0) {
pcm_val = BIAS - pcm_val;
mask = 0x7F;
} else {
pcm_val += BIAS;
mask = 0xFF;
}
/* Convert the scaled magnitude to segment number. */
seg = search(pcm_val, seg_end, 8);
/*
* Combine the sign, segment, quantization bits;
* and complement the code word.
*/
if (seg >= 8) /* out of range, return maximum value. */
return (0x7F ^ mask);
else {
uval = (seg << 4) | ((pcm_val >> (seg + 3)) & 0xF);
return (uval ^ mask);
}
}
/*
* ulaw2linear() - Convert a u-law value to 16-bit linear PCM
*
* First, a biased linear code is derived from the code word. An unbiased
* output can then be obtained by subtracting 33 from the biased code.
*
* Note that this function expects to be passed the complement of the
* original code word. This is in keeping with ISDN conventions.
*/
int
ulaw2linear(
unsigned char u_val)
{
int t;
/* Complement to obtain normal u-law value. */
u_val = ~u_val;
/*
* Extract and bias the quantization bits. Then
* shift up by the segment number and subtract out the bias.
*/
t = ((u_val & QUANT_MASK) << 3) + BIAS;
t <<= ((unsigned)u_val & SEG_MASK) >> SEG_SHIFT;
return ((u_val & SIGN_BIT) ? (BIAS - t) : (t - BIAS));
}
/* A-law to u-law conversion */
unsigned char
alaw2ulaw(
unsigned char aval)
{
aval &= 0xff;
return ((aval & 0x80) ? (0xFF ^ _a2u[aval ^ 0xD5]) :
(0x7F ^ _a2u[aval ^ 0x55]));
}
/* u-law to A-law conversion */
unsigned char
ulaw2alaw(
unsigned char uval)
{
uval &= 0xff;
return ((uval & 0x80) ? (0xD5 ^ (_u2a[0xFF ^ uval] - 1)) :
(0x55 ^ (_u2a[0x7F ^ uval] - 1)));
}