mload.f,v

lpc10-15为美军2400bps语音压缩标准的C语音源代码。

head	1.5;
access;
symbols;
locks; strict;
comment	@* @;


1.5
date	96.03.27.23.59.51;	author jaf;	state Exp;
branches;
next	1.4;

1.4
date	96.03.26.19.16.53;	author jaf;	state Exp;
branches;
next	1.3;

1.3
date	96.03.18.21.16.00;	author jaf;	state Exp;
branches;
next	1.2;

1.2
date	96.03.13.16.47.41;	author jaf;	state Exp;
branches;
next	1.1;

1.1
date	96.02.07.14.48.01;	author jaf;	state Exp;
branches;
next	;


desc
@@


1.5
log
@Added some more accurate comments about which indices of the argument
array SPEECH are read.  I thought that this might be the cause of a
problem I've been having, but it isn't.
@
text
@******************************************************************
*
*	MLOAD Version 48
*
* $Log: mload.f,v $
* Revision 1.4  1996/03/26  19:16:53  jaf
* Commented out the code at the end that copied the lower triangular
* half of PHI into the upper triangular half (making the resulting
* matrix symmetric).  The upper triangular half was never used by later
* code in subroutine ANALYS.
*
* Revision 1.3  1996/03/18  21:16:00  jaf
* Just added a few comments about which array indices of the arguments
* are used, and mentioning that this subroutine has no local state.
*
* Revision 1.2  1996/03/13  16:47:41  jaf
* Comments added explaining that none of the local variables of this
* subroutine need to be saved from one invocation to the next.
*
* Revision 1.1  1996/02/07 14:48:01  jaf
* Initial revision
*
*
******************************************************************
*
* Load a covariance matrix.
*
* Input:
*  ORDER            - Analysis order
*  AWINS            - Analysis window start
*  AWINF            - Analysis window finish
*  SPEECH(AWINF)    - Speech buffer
*                     Indices MIN(AWINS, AWINF-(ORDER-1)) through
*                             MAX(AWINF, AWINS+(ORDER-1)) read.
*                     As long as (AWINF-AWINS) .GE. (ORDER-1),
*                     this is just indices AWINS through AWINF.
* Output:
*  PHI(ORDER,ORDER) - Covariance matrix
*                     Lower triangular half and diagonal written, and read.
*                     Upper triangular half untouched.
*  PSI(ORDER)       - Prediction vector
*                     Indices 1 through ORDER written,
*                     and most are read after that.
*
* This subroutine has no local state.
*

	SUBROUTINE MLOAD( ORDER, AWINS, AWINF, SPEECH, PHI, PSI )

*       Arguments

	INTEGER ORDER, AWINS, AWINF
	REAL SPEECH(AWINF)
	REAL PHI(ORDER,ORDER), PSI(ORDER)

*       Local variables that need not be saved

	INTEGER R, C, I, START

*   Load first column of triangular covariance matrix PHI

	START = AWINS + ORDER
	DO R = 1,ORDER
	   PHI(R,1) = 0.
	   DO I = START,AWINF
	      PHI(R,1) = PHI(R,1) + SPEECH(I-1)*SPEECH(I-R)
	   END DO
	END DO

*   Load last element of vector PSI

	PSI(ORDER) = 0.
	DO I = START,AWINF
	   PSI(ORDER) = PSI(ORDER) + SPEECH(I)*SPEECH(I-ORDER)
	END DO

*   End correct to get additional columns of PHI

	DO R = 2,ORDER
	   DO C = 2,R
	      PHI(R,C) = PHI(R-1,C-1)
     1                  - SPEECH(AWINF+1-R)*SPEECH(AWINF+1-C)
     1                    + SPEECH(START-R)*SPEECH(START-C)
	   END DO
	END DO

*   End correct to get additional elements of PSI

	DO C = 1,ORDER-1
	   PSI(C) = PHI(C+1,1) - SPEECH(START-1)*SPEECH(START-1-C)
     1                          + SPEECH(AWINF)*SPEECH(AWINF-C)
	END DO

*   Copy lower triangular section into upper (why bother?)

*       I'm commenting this out, since the upper triangular half of PHI
*       is never used by later code, unless a sufficiently high level of
*       tracing is turned on.

*	DO R = 1,ORDER
*	   DO C = 1,R-1
*	      PHI(C,R) = PHI(R,C)
*	   END DO
*	END DO

	RETURN
	END
@


1.4
log
@Commented out the code at the end that copied the lower triangular
half of PHI into the upper triangular half (making the resulting
matrix symmetric).  The upper triangular half was never used by later
code in subroutine ANALYS.
@
text
@d6 6
d33 4
a36 1
*                     Indices AWINS through AWINF read.
@


1.3
log
@Just added a few comments about which array indices of the arguments
are used, and mentioning that this subroutine has no local state.
@
text
@d6 4
d31 1
a31 4
*                     After that, the matrix is made symmetric by copying
*                     lower triangular half into upper.  This step
*                     could easily be avoided, and it should be if the
*                     upper triangular half is not used later.
d87 9
a95 5
	DO R = 1,ORDER
	   DO C = 1,R-1
	      PHI(C,R) = PHI(R,C)
	   END DO
	END DO
@


1.2
log
@Comments added explaining that none of the local variables of this
subroutine need to be saved from one invocation to the next.
@
text
@d6 4
d16 18
a33 1
*  Load a covariance matrix
d35 1
a35 8
*  Inputs:
*    ORDER            - Analysis order
*    AWINS            - Analysis window start
*    AWINF            - Analysis window finish
*    SPEECH(AWINF)    - Speech buffer
*  Outputs:
*    PHI(ORDER,ORDER) - Covariance matrix
*    PSI(ORDER)       - Prediction vector
d40 1
a40 1
*       Parameters
d46 1
a46 4
*       Local variables
*       
*       None of these need to have their values saved from one
*       invocation to the next.
@


1.1
log
@Initial revision
@
text
@d5 4
a8 1
* $Log$
d25 3
d31 6
@