dmG711(3dm) dmG711(3dm)
dmG711MulawEncode, dmG711MulawDecode, dmG711MulawZeroTrapEncode,
dmG711MulawZeroTrapDecode, dmG711AlawEncode, dmG711AlawDecode,
dmG711MulawToAlaw, dmG711AlawToMulaw, dmSunMulawEncode, dmSunMulawDecode,
dmNeXTMulawEncode, dmNeXTMulawDecode - G.711 mu-law, A-law conversion
routines
#include <dmedia/dm_audioutil.h>
void dmG711MulawEncode(short *samples, unsigned char *mulawdata,
int numsamples)
void dmG711MulawDecode(unsigned char *mulawdata, short *samples,
int numsamples)
void dmG711MulawZeroTrapEncode(short *samples, unsigned char *mulawdata,
int numsamples)
void dmG711MulawZeroTrapDecode(unsigned char *mulawdata, short *samples,
int numsamples)
void dmG711AlawEncode(short *samples, unsigned char *Alawdata,
int numsamples)
void dmG711AlawDecode(unsigned char *Alawdata, short *samples,
int numsamples)
void dmG711MulawToAlaw(unsigned char *mulawdata, unsigned char *Alawdata,
int numsamples)
void dmG711AlawToMulaw(unsigned char *Alawdata, unsigned char *mulawdata,
int numsamples)
void dmSunMulawEncode(short *samples, unsigned char *mulawdata,
int numsamples)
void dmSunMulawDecode(unsigned char *mulawdata, short *samples,
int numsamples)
void dmNeXTMulawEncode(short *samples, unsigned char *mulawdata,
int numsamples)
void dmNeXTMulawDecode(unsigned char *mulawdata, short *samples,
int numsamples)
samples pointer to input sample data buffer. The data format is
short (16-bit). The samples are assumed to be two's
complement. Sampling rate could range from 8 kHz to 48 kHz.
mulawdata pointer to output mulaw data buffer. The data format is
unsigned char (8-bit).
Alawdata pointer to output Alaw data buffer. The data format is
unsigned char (8-bit).
Page 1
dmG711(3dm) dmG711(3dm)
numsamples length of the input and output buffers, in samples.
None.
dmG711MulawEncode(3dm), dmG711MulawDecode(3dm),
dmG711MulawZeroTrapEncode(3dm), dmG711MulawZeroTrapDecode(3dm),
dmG711AlawEncode(3dm), dmG711AlawDecode(3dm), dmG711MulawToAlaw(3dm),
dmG711AlawToMulaw(3dm) implement the ITU (International Telecommunication
Union) Recommendation G.711 compressions. They conform to the
specifications in the Recommendation as well as the optional
implementations for linear to mu-law and linear to A-law conversions, as
described in the Recommendation G.721, 4.2.8 sub block COMPRESS. See
NOTE for further explanation.
dmG711MulawEncode(3dm), dmG711MulawDecode(3dm) convert a 16-bit linear
PCM value to 8-bit mu-law value and vice versa. The input signed 16-bit
linear sample is first truncated to 14-bit by right shift of 2 bits by
dmG711MulawEncode and then converted to a 8-bit mu-law value by table
lookup. The output of decompression is 16-bit PCM value, which is done
by dmG711MulawDecode(3dm) through another table lookup.
dmG711MulawZeroTrapEncode(3dm), dmG711MulawZeroTrapDecode(3dm) implement
the ITU G.711 mu-law compression with zero trap during compression.
Selection of this option is necessary in networks where suppression of
the all 0 character signal is required. It reduces signal quality for
large negative values. dmG711MulawZeroTrapDecode(3dm) is the same as
dmG711MulawDecode(3dm).
dmG711AlawEncode(3dm), dmG711AlawDecode(3dm) convert a 16-bit linear PCM
value to 8-bit A-law value and vice versa. The input signed 16-bit
linear sample is first truncated to 13-bit by right shift of 3 bits by
dmG711AlawEncode and then converted to a 8-bit A-law value by table
lookup. The output of decompression is 16-bit PCM value, which is done
by dmG711AlawDecode(3dm) through another table lookup.
dmG711MulawToAlaw(3dm), dmG711AlawToMulaw(3dm) implement direct mu-law to
A-law or A-law to mu-law value conversions.
dmSunMulawEncode(3dm), dmSunMulawDecode(3dm), dmNeXTMulawEncode(3dm),
dmNeXTMulawDecode(3dm) convert a 16-bit linear PCM value to 8-bit mu-law
value and vice versa but using the conversion tables of Sun Microsystems'
and NeXT Computer's. The input signed 16-bit linear sample is first
truncated to 13-bit by right shift of 3 bits and then converted to a 8-
bit mu-law value by table lookup. The output of decompression is 16-bit
PCM value, which is done through another table lookup. These tables are
different from ITU Recommendation G.711 in that 13-bit precision table is
used in Sun's and NeXT's implementation instead of 14-bit specified by
Page 2
dmG711(3dm) dmG711(3dm)
G.711 for linear to mu-law conversion. The table size (8192) is half of
the size of G.711 (16384). Sun Microsystems' tables are the same as
those found on the 56k DSP in hardware. NeXT's table of mu-law to linear
conversion agrees with G.711. These routines can be used in cases of
processing Sun or NeXT mu-law audio files. The errors caused by using
above different tables are generally not audible because they are even
smaller than the mu-law quantization noise.
ITU Recommendation G.711 states that "the conversion to A-law or mu-law
values from uniform PCM values corresponding to the decision values, is
left to the individual equipment specification." Therefore, there might
be different implementations of G.711 that are all valid in term of mulaw
values corresponding to the decision values, which are the values of
the boundaries of the quantization intervals. Despite of the above
variations, 14-bit precision linear to mu-law conversion is still
required in order to conform the G.711 specifications.
ITU (International Telecommunication Union) Recommendations G.711 and
G.721.
PPPPaaaaggggeeee 3333 [ Back ]
|
