1mlib_SignalCepstral_S16_Adpm(e3dMiLaILBi)b Library Fumnlcitbi_oSnisgnalCepstral_S16_Adp(3MLIB)
2
3
4

NAME

6       mlib_SignalCepstral_S16_Adp - perform cepstral analysis
7

SYNOPSIS

9       cc [ flag... ] file... -lmlib [ library... ]
10       #include <mlib.h>
11
12       mlib_status mlib_SignalCepstral_S16_Adp(mlib_s16 *cepst,
13            mlib_s32 *cscale, const mlib_s16 *signal, void *state);
14
15

DESCRIPTION

17       The  mlib_SignalCepstral_S16_Adp() function performs cepstral analysis.
18       The scaling factor of the output data will be calculated based  on  the
19       actual data.
20
21
22       The basic operations to compute the cepstrum is shown below.
23
24               +-----------+      +--------+       +-----------+
25               |  Fourier  |      |        |       |  Inverse  |
26         ----->|           |----->| log|*| |------>|  Fourier  |----->
27          x(n) | Transform | X(k) |        | X'(k) | Transform | c(n)
28               +-----------+      +--------+       +-----------+
29
30
31
32       where  x(n)  is the input signal and c(n) is its cepstrum. In mathemat‐
33       ics, they are
34
35                N-1                2*PI*k*n
36         X(k) = SUM x(n) * exp(-j*----------)
37                n=0                   N
38
39         X'(k) = log|X(k)|
40
41                 1  N-1                2*PI*k*n
42         c(n) = --- SUM X'(k) * exp(j*----------)
43                 N  n=0                   N
44
45
46
47       Since X'(k) is real and even (symmetric), i.e.
48
49         X'(k) = X'(N - k)
50
51
52
53       the c(n) is real and the equation becomes Cosine transform.
54
55                 1  N-1              2*PI*k*n
56         c(n) = --- SUM X'(k) * cos(----------)
57                 N  n=0                 N
58
59
60
61       The cepstral coefficients in LPC is a special case of the above.
62
63
64       See Digital Signal Processing  by  Alan  V.  Oppenheim  and  Ronald  W.
65       Schafer, Prentice Hall, 1974.
66
67
68       See  Fundamentals  of  Speech Recognition by Lawrence Rabinerand Biing-
69       Hwang Juang, Prentice Hall, 1993.
70

PARAMETERS

72       The function takes the following arguments:
73
74       cepst     The cepstral coefficients.
75
76
77       cscale    The  scaling   factor   of   cepstral   coefficients,   where
78                 actual_data = output_data * 2**(-scaling_factor).
79
80
81       signal    The  input  signal vector, the signal samples are in Q15 for‐
82                 mat.
83
84
85       state     Pointer to the internal state structure.
86
87

RETURN VALUES

89       The function returns MLIB_SUCCESS if successful. Otherwise  it  returns
90       MLIB_FAILURE.
91

ATTRIBUTES

93       See attributes(5) for descriptions of the following attributes:
94
95
96
97
98       ┌─────────────────────────────┬─────────────────────────────┐
99       │      ATTRIBUTE TYPE         │      ATTRIBUTE VALUE        │
100       ├─────────────────────────────┼─────────────────────────────┤
101       │Interface Stability          │Committed                    │
102       ├─────────────────────────────┼─────────────────────────────┤
103       │MT-Level                     │MT-Safe                      │
104       └─────────────────────────────┴─────────────────────────────┘
105

SEE ALSO

107       mlib_SignalCepstralInit_S16(3MLIB),     mlib_SignalCepstral_S16(3MLIB),
108       mlib_SignalCepstralFree_S16(3MLIB), attributes(5)
109
110
111
112SunOS 5.11                        2 Mar 2007mlib_SignalCepstral_S16_Adp(3MLIB)
Impressum