zerofilt.h

g729 coding ipaddressing

/**************************************************************************
*
* ROUTINE
*				zerofilt
*
* FUNCTION
*				Direct form all-zero filter
*
* SYNOPSIS
*				subroutine zerofilt(b, n, z, xy, len)
*
*	formal 
*
*						data	I/O
*		name			type	type	function
*		-------------------------------------------------------------------
*		b				real	i		N+1 filter coefficients
*		n				int 	i		Filter order 
*		z				real	i/o 	N+1 filter delay elements 
*										(maintained by the user)
*		xy				real	i/o 	Input/Output data array 
*		len 			int 	i		Number of samples to filter
*
***************************************************************************
*		
* DESCRIPTION
*
*		Nonrecursive all-zero in-place time-domain filter.
*		The filter is implemented in a direct form realisation.
*
*				N		-i
*		H(z) = SUM b(i)z
*			   i=0
*
*		x(t) ->---(z0)----- b0 >------+-----> y(t)
*				   |				  |
*				   z1------ b1 >------+
*				   |				  |
*				   z2------ b2 >------+
*				   |				  |
*				   :				  :
*				   |				  |
*				   zN------ bN >------+
*
***************************************************************************
*
* CALLED BY
*
*		celp	confg	postfilt
*
* CALLS
*
*
*
***************************************************************************
*
* REFERENCES
*
*		Oppenheim & Schafer, Digital Signal Processing, PH, 1975, p. 149.
*
**************************************************************************/

static void zerofilt(const float b[], int n, float z[], float xy[], int len)
{
  float ar;
  int t, j;

  for (t = 0; t < len; t++)
  {
	z[0] = xy[t];
	ar	 = 0.0;
	for (j = n; j > 0; j--)
	{
	  ar  += z[j] * b[j];
	  z[j] = z[j-1];
	}
	xy[t] = ar + z[0] * b[0];
  }
}