mirror of https://git.ffmpeg.org/ffmpeg.git
dither lpc cpeffs
Originally committed as revision 5727 to svn://svn.ffmpeg.org/ffmpeg/trunk
This commit is contained in:
parent
ebd7617ba7
commit
1484a46753
|
@ -661,7 +661,7 @@ static void quantize_lpc_coefs(double *lpc_in, int order, int precision,
|
||||||
int32_t *lpc_out, int *shift)
|
int32_t *lpc_out, int *shift)
|
||||||
{
|
{
|
||||||
int i;
|
int i;
|
||||||
double cmax;
|
double cmax, error;
|
||||||
int32_t qmax;
|
int32_t qmax;
|
||||||
int sh;
|
int sh;
|
||||||
|
|
||||||
|
@ -697,8 +697,11 @@ static void quantize_lpc_coefs(double *lpc_in, int order, int precision,
|
||||||
}
|
}
|
||||||
|
|
||||||
/* output quantized coefficients and level shift */
|
/* output quantized coefficients and level shift */
|
||||||
|
error=0;
|
||||||
for(i=0; i<order; i++) {
|
for(i=0; i<order; i++) {
|
||||||
lpc_out[i] = (int32_t)(lpc_in[i] * (1 << sh));
|
error += lpc_in[i] * (1 << sh);
|
||||||
|
lpc_out[i] = clip(lrintf(error), -qmax, qmax);
|
||||||
|
error -= lpc_out[i];
|
||||||
}
|
}
|
||||||
*shift = sh;
|
*shift = sh;
|
||||||
}
|
}
|
||||||
|
|
Loading…
Reference in New Issue