Commit 08c43397 authored by Vitor Sessak's avatar Vitor Sessak

Use ff_dot_productf() in ra288.c

Originally committed as revision 15757 to svn://svn.ffmpeg.org/ffmpeg/trunk
parent 3ef08e32
......@@ -160,7 +160,7 @@ OBJS-$(CONFIG_QPEG_DECODER) += qpeg.o
OBJS-$(CONFIG_QTRLE_DECODER) += qtrle.o
OBJS-$(CONFIG_QTRLE_ENCODER) += qtrleenc.o
OBJS-$(CONFIG_RA_144_DECODER) += ra144.o celp_filters.o
OBJS-$(CONFIG_RA_288_DECODER) += ra288.o
OBJS-$(CONFIG_RA_288_DECODER) += ra288.o celp_math.o
OBJS-$(CONFIG_RAWVIDEO_DECODER) += rawdec.o
OBJS-$(CONFIG_RAWVIDEO_ENCODER) += rawenc.o
OBJS-$(CONFIG_RL2_DECODER) += rl2.o
......
......@@ -24,6 +24,7 @@
#include "bitstream.h"
#include "ra288.h"
#include "lpc.h"
#include "celp_math.h"
typedef struct {
float sp_lpc[36]; ///< LPC coefficients for speech data (spec: A)
......@@ -52,17 +53,6 @@ static av_cold int ra288_decode_init(AVCodecContext *avctx)
return 0;
}
static inline float scalar_product_float(const float * v1, const float * v2,
int size)
{
float res = 0.;
while (size--)
res += *v1++ * *v2++;
return res;
}
static void apply_window(float *tgt, const float *m1, const float *m2, int n)
{
while (n--)
......@@ -72,7 +62,7 @@ static void apply_window(float *tgt, const float *m1, const float *m2, int n)
static void convolve(float *tgt, const float *src, int len, int n)
{
for (; n >= 0; n--)
tgt[n] = scalar_product_float(src, src - n, len);
tgt[n] = ff_dot_productf(src, src - n, len);
}
......@@ -101,7 +91,7 @@ static void decode(RA288Context *ractx, float gain, int cb_coef)
for (i=0; i < 5; i++)
buffer[i] = codetable[cb_coef][i] * sumsum;
sum = scalar_product_float(buffer, buffer, 5) * ((1<<24)/5.);
sum = ff_dot_productf(buffer, buffer, 5) * ((1<<24)/5.);
sum = FFMAX(sum, 1);
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment