freeswitch/libs/libg722_1/src/dct4.c

201 lines
5.9 KiB
C

/*
* g722_1 - a library for the G.722.1 and Annex C codecs
*
* dct4.c
*
* Adapted by Steve Underwood <steveu@coppice.org> from the reference
* code supplied with ITU G.722.1, which is:
*
* (C)2004 Polycom, Inc.
* All rights reserved.
*
* This program 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.
*/
#if defined(HAVE_CONFIG_H)
#include <config.h>
#endif
#include <inttypes.h>
#include <stdlib.h>
#include <stdio.h>
#include <math.h>
#include <memory.h>
#include "g722_1/g722_1.h"
#include "defs.h"
#include "utilities.h"
#if !defined(G722_1_USE_FIXED_POINT)
typedef struct
{
float cosine;
float minus_sine;
} cos_msin_t;
#include "dct4.h"
static const cos_msin_t *cos_msin_table[] =
{
cos_msin_5,
cos_msin_10,
cos_msin_20,
cos_msin_40,
cos_msin_80,
cos_msin_160,
cos_msin_320,
cos_msin_640
};
/* Discrete Cosine Transform, Type IV */
void dct_type_iv(float input[], float output[], int dct_length)
{
float buffer_a[MAX_DCT_LENGTH];
float buffer_b[MAX_DCT_LENGTH];
float buffer_c[MAX_DCT_LENGTH];
float *in_ptr;
float *in_ptr_low;
float *in_ptr_high;
float *next_in_base;
float *out_ptr;
float *next_out_base;
float *out_buffer;
float *in_buffer;
float *buffer_swap;
float *fptr0;
float in_val_low;
float in_val_high;
float cos_even;
float cos_odd;
float msin_even;
float msin_odd;
const float *fptr2;
const float *core_a;
const cos_msin_t **table_ptr_ptr;
const cos_msin_t *cos_msin_ptr;
int set_span;
int set_count;
int set_count_log;
int pairs_left;
int sets_left;
int i;
int k;
int dct_length_log;
if (dct_length == MAX_DCT_LENGTH)
{
core_a = max_dct_core_a;
dct_length_log = MAX_DCT_LENGTH_LOG;
}
else
{
core_a = dct_core_a;
dct_length_log = DCT_LENGTH_LOG;
}
/* Do the sum/difference butterflies, the first part of
converting one N-point transform into N/2 two-point
transforms, where N = 1 << dct_length_log. */
in_buffer = input;
out_buffer = buffer_a;
for (set_count_log = 0; set_count_log <= dct_length_log - 2; set_count_log++)
{
/* Initialization for the loop over sets at the current size */
set_span = dct_length >> set_count_log;
set_count = 1 << set_count_log;
in_ptr = in_buffer;
next_out_base = out_buffer;
/* Loop over all the sets of this size */
for (sets_left = set_count; sets_left > 0; sets_left--)
{
/* Set up output pointers for the current set */
out_ptr = next_out_base;
next_out_base += set_span;
/* Loop over all the butterflies in the current set */
for (i = 0; i < (set_span >> 1); i++)
{
in_val_low = *in_ptr++;
in_val_high = *in_ptr++;
out_ptr[i] = in_val_low + in_val_high;
out_ptr[set_span - 1 - i] = in_val_low - in_val_high;
}
}
/* Decide which buffers to use as input and output next time.
Except for the first time (when the input buffer is the
subroutine input) we just alternate the local buffers. */
in_buffer = out_buffer;
out_buffer = (out_buffer == buffer_a) ? buffer_b : buffer_a;
}
/* Do dct_size/10 ten-point transforms */
fptr0 = in_buffer;
buffer_swap = buffer_c;
for (pairs_left = 1 << (dct_length_log - 1); pairs_left > 0; pairs_left--)
{
fptr2 = core_a;
for (k = 0; k < CORE_SIZE; k++)
{
buffer_swap[k] = vec_dot_prodf(fptr0, fptr2, CORE_SIZE);
fptr2 += CORE_SIZE;
}
fptr0 += CORE_SIZE;
buffer_swap += CORE_SIZE;
}
memcpy(in_buffer, buffer_c, dct_length*sizeof(float));
table_ptr_ptr = cos_msin_table;
/* Perform rotation butterflies */
for (set_count_log = dct_length_log - 2; set_count_log >= 0; set_count_log--)
{
/* Initialization for the loop over sets at the current size */
set_span = dct_length >> set_count_log;
set_count = 1 << set_count_log;
next_in_base = in_buffer;
next_out_base = (set_count_log == 0) ? output : out_buffer;
table_ptr_ptr++;
/* Loop over all the sets of this size */
for (sets_left = set_count; sets_left > 0; sets_left--)
{
/* Set up the pointers for the current set */
in_ptr_low = next_in_base;
in_ptr_high = in_ptr_low + (set_span >> 1);
out_ptr = next_out_base;
cos_msin_ptr = *table_ptr_ptr;
/* Loop over all the butterfly pairs in the current set */
for (i = 0; i < (set_span >> 1); i += 2)
{
cos_even = cos_msin_ptr[i].cosine;
msin_even = cos_msin_ptr[i].minus_sine;
cos_odd = cos_msin_ptr[i + 1].cosine;
msin_odd = cos_msin_ptr[i + 1].minus_sine;
out_ptr[i] = cos_even*in_ptr_low[i] - msin_even*in_ptr_high[i];
out_ptr[set_span - 1 - i] = msin_even*in_ptr_low[i] + cos_even*in_ptr_high[i];
out_ptr[i + 1] = cos_odd*in_ptr_low[i + 1] + msin_odd*in_ptr_high[i + 1];
out_ptr[set_span - 2 - i] = msin_odd*in_ptr_low[i + 1] - cos_odd*in_ptr_high[i + 1];
}
next_in_base += set_span;
next_out_base += set_span;
}
/* Swap input and output buffers for next time */
buffer_swap = in_buffer;
in_buffer = out_buffer;
out_buffer = buffer_swap;
}
}
/*- End of function --------------------------------------------------------*/
#endif
/*- End of file ------------------------------------------------------------*/