view ra288.c @ 3198:6b9f0c4fbdbe libavcodec

First part of a series of speed-enchancing patches. This one sets up a snow.h and makes snow use the dsputil function pointer framework to access the three functions that will be implemented in asm in the other parts of the patchset. Patch by Robert Edele < yartrebo AH earthlink POIS net> Original thread: Subject: [Ffmpeg-devel] [PATCH] Snow mmx+sse2 asm optimizations Date: Sun, 05 Feb 2006 12:47:14 -0500

/*
 * RealAudio 2.0 (28.8K)
 * Copyright (c) 2003 the ffmpeg project
 *
 * This library is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2 of the License, or (at your option) any later version.
 *
 * This library 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.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with this library; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

#include "avcodec.h"
#include "ra288.h"

typedef struct {
        float   history[8];
        float   output[40];
        float   pr1[36];
        float   pr2[10];
        int     phase, phasep;

        float   st1a[111],st1b[37],st1[37];
        float   st2a[38],st2b[11],st2[11];
        float   sb[41];
        float   lhist[10];
} Real288_internal;

static int ra288_decode_init(AVCodecContext * avctx)
{
        Real288_internal *glob=avctx->priv_data;
        memset(glob,0,sizeof(Real288_internal));
        return 0;
}

static void prodsum(float *tgt, float *src, int len, int n);
static void co(int n, int i, int j, float *in, float *out, float *st1, float *st2, const float *table);
static int pred(float *in, float *tgt, int n);
static void colmult(float *tgt, float *m1, const float *m2, int n);


/* initial decode */
static void unpack(unsigned short *tgt, unsigned char *src, unsigned int len)
{
  int x,y,z;
  int n,temp;
  int buffer[len];

  for (x=0;x<len;tgt[x++]=0)
    buffer[x]=9+(x&1);

  for (x=y=z=0;x<len/*was 38*/;x++) {
    n=buffer[y]-z;
    temp=src[x];
    if (n<8) temp&=255>>(8-n);
    tgt[y]+=temp<<z;
    if (n<=8) {
      tgt[++y]+=src[x]>>n;
      z=8-n;
    } else z+=8;
  }
}

static void update(Real288_internal *glob)
{
  int x,y;
  float buffer1[40],temp1[37];
  float buffer2[8],temp2[11];

  for (x=0,y=glob->phasep+5;x<40;buffer1[x++]=glob->output[(y++)%40]);
  co(36,40,35,buffer1,temp1,glob->st1a,glob->st1b,table1);
  if (pred(temp1,glob->st1,36))
    colmult(glob->pr1,glob->st1,table1a,36);

  for (x=0,y=glob->phase+1;x<8;buffer2[x++]=glob->history[(y++)%8]);
  co(10,8,20,buffer2,temp2,glob->st2a,glob->st2b,table2);
  if (pred(temp2,glob->st2,10))
    colmult(glob->pr2,glob->st2,table2a,10);
}

/* Decode and produce output */
static void decode(Real288_internal *glob, unsigned int input)
{
  unsigned int x,y;
  float f;
  double sum,sumsum;
  float *p1,*p2;
  float buffer[5];
  const float *table;

  for (x=36;x--;glob->sb[x+5]=glob->sb[x]);
  for (x=5;x--;) {
    p1=glob->sb+x;p2=glob->pr1;
    for (sum=0,y=36;y--;sum-=(*(++p1))*(*(p2++)));
    glob->sb[x]=sum;
  }

  f=amptable[input&7];
  table=codetable+(input>>3)*5;

  /* convert log and do rms */
  for (sum=32,x=10;x--;sum-=glob->pr2[x]*glob->lhist[x]);
  if (sum<0) sum=0; else if (sum>60) sum=60;

  sumsum=exp(sum*0.1151292546497)*f;    /* pow(10.0,sum/20)*f */
  for (sum=0,x=5;x--;) { buffer[x]=table[x]*sumsum; sum+=buffer[x]*buffer[x]; }
  if ((sum/=5)<1) sum=1;

  /* shift and store */
  for (x=10;--x;glob->lhist[x]=glob->lhist[x-1]);
  *glob->lhist=glob->history[glob->phase]=10*log10(sum)-32;

  for (x=1;x<5;x++) for (y=x;y--;buffer[x]-=glob->pr1[x-y-1]*buffer[y]);

  /* output */
  for (x=0;x<5;x++) {
    f=glob->sb[4-x]+buffer[x];
    if (f>4095) f=4095; else if (f<-4095) f=-4095;
    glob->output[glob->phasep+x]=glob->sb[4-x]=f;
  }
}

/* column multiply */
static void colmult(float *tgt, float *m1, const float *m2, int n)
{
  while (n--)
    *(tgt++)=(*(m1++))*(*(m2++));
}

static int pred(float *in, float *tgt, int n)
{
  int x,y;
  float *p1,*p2;
  double f0,f1,f2;
  float temp;

  if (in[n]==0) return 0;
  if ((f0=*in)<=0) return 0;

  for (x=1;;x++) {
    if (n<x) return 1;

    p1=in+x;
    p2=tgt;
    f1=*(p1--);
    for (y=x;--y;f1+=(*(p1--))*(*(p2++)));

    p1=tgt+x-1;
    p2=tgt;
    *(p1--)=f2=-f1/f0;
    for (y=x>>1;y--;) {
      temp=*p2+*p1*f2;
      *(p1--)+=*p2*f2;
      *(p2++)=temp;
    }
    if ((f0+=f1*f2)<0) return 0;
  }
}

static void co(int n, int i, int j, float *in, float *out, float *st1, float *st2, const float *table)
{
  int a,b,c;
  unsigned int x;
  float *fp;
  float buffer1[37];
  float buffer2[37];
  float work[111];

  /* rotate and multiply */
  c=(b=(a=n+i)+j)-i;
  fp=st1+i;
  for (x=0;x<b;x++) {
    if (x==c) fp=in;
    work[x]=*(table++)*(*(st1++)=*(fp++));
  }

  prodsum(buffer1,work+n,i,n);
  prodsum(buffer2,work+a,j,n);

  for (x=0;x<=n;x++) {
    *st2=*st2*(0.5625)+buffer1[x];
    out[x]=*(st2++)+buffer2[x];
  }
  *out*=1.00390625; /* to prevent clipping */
}

/* product sum (lsf) */
static void prodsum(float *tgt, float *src, int len, int n)
{
  unsigned int x;
  float *p1,*p2;
  double sum;

  while (n>=0)
  {
    p1=(p2=src)-n;
    for (sum=0,x=len;x--;sum+=(*p1++)*(*p2++));
    tgt[n--]=sum;
  }
}

static void * decode_block(AVCodecContext * avctx, unsigned char *in, signed short int *out,unsigned len)
{
  int x,y;
  Real288_internal *glob=avctx->priv_data;
  unsigned short int buffer[len];

  unpack(buffer,in,len);
  for (x=0;x<32;x++)
  {
    glob->phasep=(glob->phase=x&7)*5;
    decode(glob,buffer[x]);
    for (y=0;y<5;*(out++)=8*glob->output[glob->phasep+(y++)]);
    if (glob->phase==3) update(glob);
  }
  return out;
}

/* Decode a block (celp) */
static int ra288_decode_frame(AVCodecContext * avctx,
            void *data, int *data_size,
            uint8_t * buf, int buf_size)
{
    void *datao;

    if (buf_size < avctx->block_align)
    {
        av_log(avctx, AV_LOG_ERROR, "ffra288: Error! Input buffer is too small [%d<%d]\n",buf_size,avctx->block_align);
        return 0;
    }

    datao = data;
    data = decode_block(avctx, buf, (signed short *)data, avctx->block_align);

    *data_size = (char *)data - (char *)datao;
    return avctx->block_align;
}

AVCodec ra_288_decoder =
{
    "real_288",
    CODEC_TYPE_AUDIO,
    CODEC_ID_RA_288,
    sizeof(Real288_internal),
    ra288_decode_init,
    NULL,
    NULL,
    ra288_decode_frame,
};