init.m

语音编码

% MATLAB SIMULATION OF FS-1015 LPC-10e
% COPYRIGHT (C) 1996-99 ANDREAS SPANIAS and TED PAINTER
%
% This Copyright applies only to this particular MATLAB implementation
% of the LPC-10e coder.  The MATLAB software is intended only for educational
% purposes.  No other use is intended or authorized.  This is not a public
% domain program and unauthorized distribution to individuals or networks 
% is prohibited. Be aware that use of the standard in any form is goverened
% by rules of the US DoD.  
% This program is free software. It is distributed in the hope that it will
% be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  There is no commitment 
% or even implied commitment on behalf of Andreas Spanias or Ted Painter
% for maintenance or support of this code.
%
% MATLAB is trademark of The Mathworks Inc
%
% ALL DERIVATIVE WORKS MUST INCLUDE THIS COPYRIGHT NOTICE.
%
% ******************************************************************
% INIT
%
% PORTED TO MATLAB FROM LPC-55 C RELEASE
% 2-14-94
%
% ******************************************************************
%
% DESCRIPTION
%
% Initializes all global and static data objects
%
% DESIGN NOTES
%
% "Statics" were implemented as packed matricies for several functions
% early in the development process.  Although these were preserved,
% the practice was abandoned in favor of using MATLAB globals where
% static behavior was required.
%
% INTERNAL VARIABLES
%   h0 - h15        -   LPF31 lowpass tap weights
%
% GLOBAL VARIABLES BY FUNCTION
%
% LPC10.M
%   SimRunning      -   First or subsequent call flag
%
% HP100.M
%   hp100z1         -   Highpass filter memory, first stage
%   hp100z2         -   Highpass filter memory, second stage
%
% ANALYS.M
%   inbuf           -   Raw speech, scaled to 12-bits + sign
%   pebuf           -   Preemphasized speech
%   lpbuf           -   Lowpass filtered speech buffer
%   awin            -   Analysis window indicies (3 frames)
%   ewin            -   Energy window indicies        "
%   rmsbuf          -   RMS energy                    "
%   rcbuf           -   Reflection coefficients       "
%   ivbuf           -   Inverse filtered speech
%   bias            -   DC bias estimate for input speech
%   tau             -   Table of AMDF pitch lags
%   voibuf          -   Voicing decisions for 3 frames
%
% LPFILT31.M
%   alpf31          -   Lowpass filter H(Z) denominator taps
%   blpf31          -   Lowpass filter H(Z) numerator taps
%   zlpf31          -   Lowpass filter H(Z) memory
%
% PREEMP.M
%   zpre            -   Preemphasis filter memory
%
% IVFILT.M
%   ziv             -   Inverse filter memory
%
% TBDM.M
%   tau2            -   Table of high-resolution lags
%
% VOICE.M
%   nvdcl           -   Number of voicing decision coefficient vectors (VDCs)
%   ref             -   Reference energy for for initialization/dither threshold
%   vstate          -   Decimal interpretation of binary voicing classification
%   dither          -   Zero crossing threshold level
%   sfbue           -   Static copy of full band unvoiced energy
%   slbue           -   Static copy of low band unvoiced energy
%   lbve            -   Low band voiced energy
%   fbve            -   Full band voiced energy
%   fbue            -   Full band unvoiced energy
%   ofbue           -   Previous full band unvoiced energy
%   lbue            -   Low band unvoiced energy
%   olbue           -   Previous low band unvoiced energy
%   snr             -   Estimated SNR
%   vdcl            -   SNR values corresponding to the set of VDCs
%   voice           -   History of linear discriminant analysis results
%   vdc             -   2-D voicing decision coefficient vector
%
% DYPTRK.M
%   ipoint          -   Current frame position pointer
%   alphax          -   Scaled version of alpha
%   s               -   Winner array
%   p               -   Pitch pointer array
%
% PLACEA.M
%   allv            -   True if five most recent half-frame decisions = V
%
% ENCODE.M
%   enctab          -   Table of Hamming 8,4 redundancy bits used for
%                       protection of RC1-RC4 during unvoiced frames.
%   entau           -   Pitch encoding table
%   enadd           -   Table of negative values of the bias of the
%                       probability density functions (pdf) of each of
%                       the last eight RCs.
%   enscl           -   Table of scale factors for each of the last eight
%                       RCs.
%   enbits          -   Table of number of shifts nescessary to properly
%                       scale each of the last 8 RCs to their respective
%                       bit allocations.
%   entab6          -   Log-area-ratio (LAR) table for encoding of RC1 and RC2
%   rmst            -   Energy encoding table
%
% DECODE.M
%   ivtab           -   Table containing all nescessary logical flow
%                       information for this routine.
%   corth           -   Smoothing threshold table, one row per error rate
%                       class (2-D, 4x8).
%   detau           -   Pitch decoding table
%   detab7          -   Table for inverse LAR decoding of RC1, RC2
%   descl           -   Table of scale factors for each of the last 8 RCs
%   deadd           -   Table of pdf biases for the last 8 RCs
%   qb              -   Quantization bias table of the last 8 RCs
%   nbit            -   Number of bits allocated to each of the RCs
%   zrc             -   Zero forcing table for RC5 through RC10
%   drc             -   Matrix of future, present, past, and last frame RCs
%   dpit            -   Vector of future, present, past, and last frame pitch
%   drms            -   Vector of future, present, past, and last frame RMS
%   ivp2h           -   Voicing condition of previous second half frame
%   erate           -   Measured bit error rate
%   iovoic          -   2-bit voicing condition code for present frame
%   first           -   Flag set only for first frame of data
%
% HAM84.M
%   dactab          -   Hamming 8,4 decode table
%
% SYNTHS.M
%   ivuv            -   Epoch voicing decisions
%   ipiti           -   Pitch epoch lengths
%   rmsi            -   Pitch epoch energy
%   rci             -   Pitch epoch reflection coefficients
%   pc              -   Pitch epoch predictor coefficients
%   sout            -   Pitch epoch output speech
%
% PITSYN.M
%   ivoico          -   Voicing decision of previous second half frame
%   ipito           -   Pitch of last epoch from previous frame
%   rmso            -   Energy of previous frame
%   rco             -   Reflection coefficients of previous frame
%   yarc            -   Temp storage for present frame RCs during voiced
%                       section of transition frame to unvoiced.
%   psFirst         -   Flag to indicate first call to PITSYN.M
%   jsamp           -   Number of samples from end of previous frame not
%                       yet assigned to an epoch.
%
% BSYNZ.M
%   kexc            -   Voiced excitation base sequence (predefined)
%   exc             -   Input excitation for LPC synthesis filter I (all zero)
%   exc2            -   Input and output for LPC synthesis filter II (all pole)
%   ipo             -   Pitch epoch length, previous epoch
%   bsRmso          -   Previous epoch energy
%   Zlpf            -   Filter memory, excitation LPF
%   Zhpf            -   Filter memory, excitation HPF
%
% RANDOM.M
%   j               -   Random tap index 1
%   k               -   Random tap index 2
%   y               -   Random generator taps
%
% DEEMPH.M
%   Zdemph          -   Deemphasis filter memory
%
% BUFMAN.M
%   BigBuf          -   Circular matching buffer - bsynz -> 180 samples/frame
%   sPtr            -   Head pointer
%   ePtr            -   Tail pointer
%
% GUI.M
%   guiDisplayMode  -   Frequency or time mode flag
%   guiDisplayType  -   Speech I/O, excitation, or pole plot flag
%   guiState        -   State flag: startup, runtime, or shutdown
%   guiSline        -   Handle to status line graphic object
%   guiSbuf1        -   Buffering allows gui to track LPC-10 buffering
%   guiSbuf2        -         "               "                "
%   guiSin          -   Input signal buffer for plotting
%   guiFrames       -   Number of frames in current input file
%   guiExcit        -   Excitation signal buffer for plotting
%
% "STATICS" - USED GLOBALS IN MOST PLACES RATHER THAN THESE
%   analysState     -   Static variables for analys.m
%   onsetState      -   Static variables for onset.m
%
% CONSTANTS
%   LFRAME          -   Input speech frame size
%   AF              -   Analysis frame index (into 3 frame buffer)
%   READ            -   Read flag to control diskio
%   WRITE           -   Write flag to control diskio
%   EOF             -   Status flag for diskio
%   OSLEN           -   Onset buffer size
%   MAXWIN          -   The maximum length of a window
%   LTAU            -   Number of points (lags) in the AMDF
%   SBUFL           -   The lower index of speech buffers INBUF and PEBUF
%   LBUFL           -   The lower index of speech buffer LPBUF
%   MAXORD          -   Maximum predictor order
%   ORDER           -   Predictor order
%   MIDTAP          -   Random number generator middle tap
%   MAXTAP          -   Random number generator last tap
%   MAXPIT          -   Maximum pitch period
%   DVWINL          -   Default lower index of the voicing window
%   DVWINH          -   Default upper index of the voicing window
%   MXVDCL          -   Voicing detection SNR array size
%   BBUFSIZE        -   Circular matching buffer size
%
% ******************************************************************

% CLEAR MEMORY
clear global
clear
clear all
clc

% DECLARE ALL GLOBAL CONSTANTS
global LFRAME AF OSLEN SBUFL LBUFL MAXORD MIDTAP MAXTAP MAXPIT;
global EOF READ WRITE MAXWIN LTAU ORDER;

% DEFINE GLOBAL CONSTANTS
READ        = 0;
WRITE       = 1;
EOF         = -99;
LFRAME      = 180;
AF          = 3;
OSLEN       = 10;
MAXWIN      = 156;
LTAU        = 60;
SBUFL       = 181;
LBUFL       = 25;
MAXORD      = 10;
ORDER       = 10;
MIDTAP      = 2;
MAXTAP      = 5;
MAXPIT      = 156;

% DECLARE ALL LOCAL CONSTANTS
DVWINL      = 307;
DVWINH      = 462;

% GLOBAL VARIABLES ARE DECLARED ADJACENT TO THEIR DEFINITIONS

% LPC10.M
global LPC10path
LPC10path = cd;

% LPCEXEC.M
global FrameCnt
FrameCnt = 0;

% HP100.M
global hp100z1 hp100z2;
hp100z1 = zeros( 2, 1 );
hp100z2 = zeros( 2, 1 );

% ANALYS.M
global inbuf pebuf lpbuf ivbuf bias zpre tau voibuf awin ewin rmsbuf rcbuf;
inbuf = zeros( LFRAME * AF, 1 );
pebuf = zeros( LFRAME * AF, 1 );
lpbuf = zeros( 696, 1 );
awin = zeros( 2, AF );
awin(1,AF) = DVWINL;
awin(2,AF) = DVWINH;
ewin = zeros( 2, AF );
rmsbuf = zeros( AF, 1 );
rcbuf = zeros( MAXORD, AF );
ivbuf = zeros( 312, 1 );
bias = 0.00;
zpre = zeros( 2, 1 );
tau = [ ...
	20,21,22,23,24,25,26,27,28,29,30,31,32,33,34, ...
	35,36,37,38,39,40,42,44,46,48,50,52,54,56,58,60,62,64,66, ...
	68,70,72,74,76,78,80,84,88,92,96,100,104,108,112,116,120, ...
	124,128,132,136,140,144,148,152,156 ...
      ];
tau = tau';
voibuf = zeros( 2, AF+1 );

% LPFILT31.M
global alpf31 blpf31 zlpf31;

% FILTER TAP WEIGHTS
h0          = -0.0097201988;
h1          = -0.0105179986;
h2          = -0.0083479648;
h3          =  0.0005860774;
h4          =  0.0130892089;
h5          =  0.0217052232;
h6          =  0.0184161253;
h7          =  0.0003397230;
h8          = -0.0260797087;