144 lines
4.1 KiB
C
144 lines
4.1 KiB
C
|
/*
|
||
|
|
||
|
$Log$
|
||
|
Revision 1.15 2004/06/26 03:50:14 markster
|
||
|
Merge source cleanups (bug #1911)
|
||
|
|
||
|
Revision 1.14 2003/02/12 13:59:15 matteo
|
||
|
mer feb 12 14:56:57 CET 2003
|
||
|
|
||
|
Revision 1.1.1.1 2003/02/12 13:59:15 matteo
|
||
|
mer feb 12 14:56:57 CET 2003
|
||
|
|
||
|
Revision 1.2 2000/01/05 08:20:39 markster
|
||
|
Some OSS fixes and a few lpc changes to make it actually work
|
||
|
|
||
|
* Revision 1.1 1996/08/19 22:30:58 jaf
|
||
|
* Initial revision
|
||
|
*
|
||
|
|
||
|
*/
|
||
|
|
||
|
/* -- translated by f2c (version 19951025).
|
||
|
You must link the resulting object file with the libraries:
|
||
|
-lf2c -lm (in that order)
|
||
|
*/
|
||
|
|
||
|
#include "f2c.h"
|
||
|
|
||
|
#ifdef P_R_O_T_O_T_Y_P_E_S
|
||
|
extern int preemp_(real *inbuf, real *pebuf, integer *nsamp, real *coef, real *z__);
|
||
|
#endif
|
||
|
|
||
|
/* ******************************************************************* */
|
||
|
|
||
|
/* PREEMP Version 55 */
|
||
|
|
||
|
/* $Log$
|
||
|
* Revision 1.15 2004/06/26 03:50:14 markster
|
||
|
* Merge source cleanups (bug #1911)
|
||
|
*
|
||
|
* Revision 1.14 2003/02/12 13:59:15 matteo
|
||
|
* mer feb 12 14:56:57 CET 2003
|
||
|
*
|
||
|
* Revision 1.1.1.1 2003/02/12 13:59:15 matteo
|
||
|
* mer feb 12 14:56:57 CET 2003
|
||
|
*
|
||
|
* Revision 1.2 2000/01/05 08:20:39 markster
|
||
|
* Some OSS fixes and a few lpc changes to make it actually work
|
||
|
*
|
||
|
* Revision 1.1 1996/08/19 22:30:58 jaf
|
||
|
* Initial revision
|
||
|
* */
|
||
|
/* Revision 1.3 1996/03/14 23:16:29 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 1996/03/11 23:23:34 jaf */
|
||
|
/* Added a bunch of comments to an otherwise simple subroutine. */
|
||
|
|
||
|
/* Revision 1.1 1996/02/07 14:48:48 jaf */
|
||
|
/* Initial revision */
|
||
|
|
||
|
|
||
|
/* ******************************************************************* */
|
||
|
|
||
|
/* Preemphasize speech with a single-zero filter. */
|
||
|
/* (When coef = .9375, preemphasis is as in LPC43.) */
|
||
|
|
||
|
/* Inputs: */
|
||
|
/* NSAMP - Number of samples to filter */
|
||
|
/* INBUF - Input speech buffer */
|
||
|
/* Indices 1 through NSAMP are read. */
|
||
|
/* COEF - Preemphasis coefficient */
|
||
|
/* Input/Output: */
|
||
|
/* Z - Filter state */
|
||
|
/* Output: */
|
||
|
/* PEBUF - Preemphasized speech buffer (can be equal to INBUF) */
|
||
|
/* Indices 1 through NSAMP are modified. */
|
||
|
|
||
|
/* This subroutine has no local state. */
|
||
|
|
||
|
/* Subroutine */ int preemp_(real *inbuf, real *pebuf, integer *nsamp, real *
|
||
|
coef, real *z__)
|
||
|
{
|
||
|
/* System generated locals */
|
||
|
integer i__1;
|
||
|
|
||
|
/* Local variables */
|
||
|
real temp;
|
||
|
integer i__;
|
||
|
|
||
|
/* Arguments */
|
||
|
/* Local variables */
|
||
|
|
||
|
/* None of these need to have their values saved from one */
|
||
|
/* invocation to the next. */
|
||
|
|
||
|
/* Logically, this subroutine computes the output sequence */
|
||
|
/* pebuf(1:nsamp) defined by: */
|
||
|
|
||
|
/* pebuf(i) = inbuf(i) - coef * inbuf(i-1) */
|
||
|
|
||
|
/* where inbuf(0) is defined by the value of z given as input to */
|
||
|
/* this subroutine. */
|
||
|
|
||
|
/* What is this filter's frequency response and phase response? */
|
||
|
|
||
|
/* Why is this filter applied to the speech? */
|
||
|
|
||
|
/* Could it be more efficient to apply multiple filters */
|
||
|
/* simultaneously, by combining them into one equivalent filter? */
|
||
|
|
||
|
/* Are there ever cases when "factoring" one high-order filter into
|
||
|
*/
|
||
|
/* multiple smaller-order filter actually reduces the number of */
|
||
|
/* arithmetic operations needed to perform them? */
|
||
|
/* When I first read this subroutine, I didn't understand why the */
|
||
|
/* variable temp was used. It seemed that the statements in the do
|
||
|
*/
|
||
|
/* loop could be replaced with the following: */
|
||
|
|
||
|
/* pebuf(i) = inbuf(i) - coef * z */
|
||
|
/* z = inbuf(i) */
|
||
|
|
||
|
/* The reason for temp is so that even if pebuf and inbuf are the */
|
||
|
/* same arrays in memory (i.e., they are aliased), then this */
|
||
|
/* subroutine will still work correctly. I didn't realize this */
|
||
|
/* until seeing the comment after PEBUF above that says "(can be */
|
||
|
/* equal to INBUF)". */
|
||
|
/* Parameter adjustments */
|
||
|
--pebuf;
|
||
|
--inbuf;
|
||
|
|
||
|
/* Function Body */
|
||
|
i__1 = *nsamp;
|
||
|
for (i__ = 1; i__ <= i__1; ++i__) {
|
||
|
temp = inbuf[i__] - *coef * *z__;
|
||
|
*z__ = inbuf[i__];
|
||
|
pebuf[i__] = temp;
|
||
|
/* L10: */
|
||
|
}
|
||
|
return 0;
|
||
|
} /* preemp_ */
|