annotate h263.c @ 265:4e9e728021d8 libavcodec

use ac prediction in mpeg4 encoding (5% smaller intra-blocks/keyframes) some minor optimizations last non zero index cant be 64 bugfix

/*
 * H263/MPEG4 backend for ffmpeg encoder and decoder
 * Copyright (c) 2000,2001 Gerard Lantau.
 * H263+ support.
 * Copyright (c) 2001 Juan J. Sierralta P.
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * 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.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */
#include "common.h"
#include "dsputil.h"
#include "avcodec.h"
#include "mpegvideo.h"
#include "h263data.h"
#include "mpeg4data.h"

//rounded divison & shift
#define RDIV(a,b) ((a) > 0 ? ((a)+((b)>>1))/(b) : ((a)-((b)>>1))/(b))
#define RSHIFT(a,b) ((a) > 0 ? ((a) + (1<<((b)-1)))>>(b) : ((a) + (1<<((b)-1))-1)>>(b))
#define ABS(a) (((a)>=0)?(a):(-(a)))

static void h263_encode_block(MpegEncContext * s, DCTELEM * block,
			      int n);
static void h263_encode_motion(MpegEncContext * s, int val);
static void h263p_encode_umotion(MpegEncContext * s, int val);
static void mpeg4_encode_block(MpegEncContext * s, DCTELEM * block,
			       int n, int dc, UINT8 *scan_table);
static int h263_decode_motion(MpegEncContext * s, int pred, int fcode);
static int h263p_decode_umotion(MpegEncContext * s, int pred);
static int h263_decode_block(MpegEncContext * s, DCTELEM * block,
                             int n, int coded);
static int mpeg4_decode_block(MpegEncContext * s, DCTELEM * block,
                              int n, int coded);
static inline int mpeg4_pred_dc(MpegEncContext * s, int n, UINT16 **dc_val_ptr, int *dir_ptr);
static void mpeg4_inv_pred_ac(MpegEncContext * s, INT16 *block, int n,
                              int dir);

extern UINT32 inverse[256];

int h263_get_picture_format(int width, int height)
{
    int format;

    if (width == 128 && height == 96)
	format = 1;
    else if (width == 176 && height == 144)
	format = 2;
    else if (width == 352 && height == 288)
	format = 3;
    else if (width == 704 && height == 576)
	format = 4;
    else if (width == 1408 && height == 1152)
	format = 5;
    else
        format = 7;
    return format;
}

void h263_encode_picture_header(MpegEncContext * s, int picture_number)
{
    int format;

    align_put_bits(&s->pb);

    /* Update the pointer to last GOB */
    s->ptr_lastgob = pbBufPtr(&s->pb);
    s->gob_number = 0;

    put_bits(&s->pb, 22, 0x20); /* PSC */
    put_bits(&s->pb, 8, (((INT64)s->picture_number * 30 * FRAME_RATE_BASE) / 
                         s->frame_rate) & 0xff);

    put_bits(&s->pb, 1, 1);	/* marker */
    put_bits(&s->pb, 1, 0);	/* h263 id */
    put_bits(&s->pb, 1, 0);	/* split screen off */
    put_bits(&s->pb, 1, 0);	/* camera  off */
    put_bits(&s->pb, 1, 0);	/* freeze picture release off */
    
    format = h263_get_picture_format(s->width, s->height);
    if (!s->h263_plus) {
        /* H.263v1 */
        put_bits(&s->pb, 3, format);
        put_bits(&s->pb, 1, (s->pict_type == P_TYPE));
        /* By now UMV IS DISABLED ON H.263v1, since the restrictions
        of H.263v1 UMV implies to check the predicted MV after
        calculation of the current MB to see if we're on the limits */
        put_bits(&s->pb, 1, 0);	/* unrestricted motion vector: off */
        put_bits(&s->pb, 1, 0);	/* SAC: off */
        put_bits(&s->pb, 1, 0);	/* advanced prediction mode: off */
        put_bits(&s->pb, 1, 0);	/* not PB frame */
        put_bits(&s->pb, 5, s->qscale);
        put_bits(&s->pb, 1, 0);	/* Continuous Presence Multipoint mode: off */
    } else {
        /* H.263v2 */
        /* H.263 Plus PTYPE */
        put_bits(&s->pb, 3, 7);
        put_bits(&s->pb,3,1); /* Update Full Extended PTYPE */
        if (format == 7)
            put_bits(&s->pb,3,6); /* Custom Source Format */
        else
            put_bits(&s->pb, 3, format);
            
        put_bits(&s->pb,1,0); /* Custom PCF: off */
        s->umvplus = (s->pict_type == P_TYPE) && s->unrestricted_mv;
        put_bits(&s->pb, 1, s->umvplus); /* Unrestricted Motion Vector */
        put_bits(&s->pb,1,0); /* SAC: off */
        put_bits(&s->pb,1,0); /* Advanced Prediction Mode: off */
        put_bits(&s->pb,1,0); /* Advanced Intra Coding: off */
        put_bits(&s->pb,1,0); /* Deblocking Filter: off */
        put_bits(&s->pb,1,0); /* Slice Structured: off */
        put_bits(&s->pb,1,0); /* Reference Picture Selection: off */
        put_bits(&s->pb,1,0); /* Independent Segment Decoding: off */
        put_bits(&s->pb,1,0); /* Alternative Inter VLC: off */
        put_bits(&s->pb,1,0); /* Modified Quantization: off */
        put_bits(&s->pb,1,1); /* "1" to prevent start code emulation */
        put_bits(&s->pb,3,0); /* Reserved */
		
        put_bits(&s->pb, 3, s->pict_type == P_TYPE);
		
        put_bits(&s->pb,1,0); /* Reference Picture Resampling: off */
        put_bits(&s->pb,1,0); /* Reduced-Resolution Update: off */
        put_bits(&s->pb,1,0); /* Rounding Type */
        put_bits(&s->pb,2,0); /* Reserved */
        put_bits(&s->pb,1,1); /* "1" to prevent start code emulation */
		
        /* This should be here if PLUSPTYPE */
        put_bits(&s->pb, 1, 0);	/* Continuous Presence Multipoint mode: off */
		
		if (format == 7) {
            /* Custom Picture Format (CPFMT) */
		
            put_bits(&s->pb,4,2); /* Aspect ratio: CIF 12:11 (4:3) picture */
            put_bits(&s->pb,9,(s->width >> 2) - 1);
            put_bits(&s->pb,1,1); /* "1" to prevent start code emulation */
            put_bits(&s->pb,9,(s->height >> 2));
        }
        
        /* Unlimited Unrestricted Motion Vectors Indicator (UUI) */
        if (s->umvplus)
            put_bits(&s->pb,1,1); /* Limited according tables of Annex D */
        put_bits(&s->pb, 5, s->qscale);
    }

    put_bits(&s->pb, 1, 0);	/* no PEI */
}

int h263_encode_gob_header(MpegEncContext * s, int mb_line)
{
    int pdif=0;
    
    /* Check to see if we need to put a new GBSC */
    /* for RTP packetization                    */
    if (s->rtp_mode) {
        pdif = pbBufPtr(&s->pb) - s->ptr_lastgob;
        if (pdif >= s->rtp_payload_size) {
            /* Bad luck, packet must be cut before */
            align_put_bits(&s->pb);
            flush_put_bits(&s->pb);
            /* Call the RTP callback to send the last GOB */
            if (s->rtp_callback) {
                pdif = pbBufPtr(&s->pb) - s->ptr_lastgob;
                s->rtp_callback(s->ptr_lastgob, pdif, s->gob_number);
            }
            s->ptr_lastgob = pbBufPtr(&s->pb);
            put_bits(&s->pb, 17, 1); /* GBSC */
            s->gob_number = mb_line / s->gob_index;
            put_bits(&s->pb, 5, s->gob_number); /* GN */
            put_bits(&s->pb, 2, s->pict_type == I_TYPE); /* GFID */
            put_bits(&s->pb, 5, s->qscale); /* GQUANT */
            //fprintf(stderr,"\nGOB: %2d size: %d", s->gob_number - 1, pdif);
            return pdif;
       } else if (pdif + s->mb_line_avgsize >= s->rtp_payload_size) {
           /* Cut the packet before we can't */
           align_put_bits(&s->pb);
           flush_put_bits(&s->pb);
           /* Call the RTP callback to send the last GOB */
           if (s->rtp_callback) {
               pdif = pbBufPtr(&s->pb) - s->ptr_lastgob;
               s->rtp_callback(s->ptr_lastgob, pdif, s->gob_number);
           }
           s->ptr_lastgob = pbBufPtr(&s->pb);
           put_bits(&s->pb, 17, 1); /* GBSC */
           s->gob_number = mb_line / s->gob_index;
           put_bits(&s->pb, 5, s->gob_number); /* GN */
           put_bits(&s->pb, 2, s->pict_type == I_TYPE); /* GFID */
           put_bits(&s->pb, 5, s->qscale); /* GQUANT */
           //fprintf(stderr,"\nGOB: %2d size: %d", s->gob_number - 1, pdif);
           return pdif;
       }
   }
   return 0;
}

static inline int decide_ac_pred(MpegEncContext * s, DCTELEM block[6][64], int dir[6])
{
    int score0=0, score1=0;
    int i, n;

    for(n=0; n<6; n++){
        int x, y, wrap;
        INT16 *ac_val, *ac_val1;

        if (n < 4) {
            x = 2 * s->mb_x + 1 + (n & 1);
            y = 2 * s->mb_y + 1 + (n >> 1);
            wrap = s->mb_width * 2 + 2;
            ac_val = s->ac_val[0][0];
        } else {
            x = s->mb_x + 1;
            y = s->mb_y + 1;
            wrap = s->mb_width + 2;
            ac_val = s->ac_val[n - 4 + 1][0];
        }
        ac_val += ((y) * wrap + (x)) * 16;
        ac_val1= ac_val;
        if(dir[n]){
            ac_val-= wrap*16;
            for(i=1; i<8; i++){
                const int level= block[n][block_permute_op(i   )];
                score0+= ABS(level);
                score1+= ABS(level - ac_val[i+8]);
                ac_val1[i  ]=    block[n][block_permute_op(i<<3)];
                ac_val1[i+8]= level;
            }
        }else{
            ac_val-= 16;
            for(i=1; i<8; i++){
                const int level= block[n][block_permute_op(i<<3)];
                score0+= ABS(level);
                score1+= ABS(level - ac_val[i]);
                ac_val1[i  ]= level;
                ac_val1[i+8]=    block[n][block_permute_op(i   )];
            }
        }
    }

    return score0 > score1 ? 1 : 0;    
}

void mpeg4_encode_mb(MpegEncContext * s,
		    DCTELEM block[6][64],
		    int motion_x, int motion_y)
{
    int cbpc, cbpy, i, cbp, pred_x, pred_y;
    
    //    printf("**mb x=%d y=%d\n", s->mb_x, s->mb_y);
    if (!s->mb_intra) {
        /* compute cbp */
        cbp = 0;
        for (i = 0; i < 6; i++) {
        if (s->block_last_index[i] >= 0)
            cbp |= 1 << (5 - i);
        }
        if ((cbp | motion_x | motion_y) == 0) {
            /* skip macroblock */
            put_bits(&s->pb, 1, 1);
            return;
        }
        put_bits(&s->pb, 1, 0);	/* mb coded */
        cbpc = cbp & 3;
        put_bits(&s->pb,
                inter_MCBPC_bits[cbpc],
                inter_MCBPC_code[cbpc]);
        cbpy = cbp >> 2;
        cbpy ^= 0xf;
        put_bits(&s->pb, cbpy_tab[cbpy][1], cbpy_tab[cbpy][0]);

        /* motion vectors: 16x16 mode only now */
        h263_pred_motion(s, 0, &pred_x, &pred_y);
      
        h263_encode_motion(s, motion_x - pred_x);
        h263_encode_motion(s, motion_y - pred_y);

        /* encode each block */
        for (i = 0; i < 6; i++) {
            mpeg4_encode_block(s, block[i], i, 0, zigzag_direct);
        }
    } else {
        int dc_diff[6];   //dc values with the dc prediction subtracted 
        int dir[6];  //prediction direction
        int zigzag_last_index[6];
        UINT8 *scan_table[6];

        for(i=0; i<6; i++){
            const int level= block[i][0];
            UINT16 *dc_ptr;

            dc_diff[i]= level - mpeg4_pred_dc(s, i, &dc_ptr, &dir[i]);
            if (i < 4) {
                *dc_ptr = level * s->y_dc_scale;
            } else {
                *dc_ptr = level * s->c_dc_scale;
            }
        }

        s->ac_pred= decide_ac_pred(s, block, dir);

        if(s->ac_pred){
            for(i=0; i<6; i++){
                UINT8 *st;
                int last_index;

                mpeg4_inv_pred_ac(s, block[i], i, dir[i]);
                if (dir[i]==0) st = ff_alternate_vertical_scan; /* left */
                else           st = ff_alternate_horizontal_scan; /* top */

                for(last_index=63; last_index>=0; last_index--) //FIXME optimize
                    if(block[i][st[last_index]]) break;
                zigzag_last_index[i]= s->block_last_index[i];
                s->block_last_index[i]= last_index;
                scan_table[i]= st;
            }
        }else{
            for(i=0; i<6; i++)
                scan_table[i]= zigzag_direct;
        }

        /* compute cbp */
        cbp = 0;
        for (i = 0; i < 6; i++) {
            if (s->block_last_index[i] >= 1)
                cbp |= 1 << (5 - i);
        }

        cbpc = cbp & 3;
        if (s->pict_type == I_TYPE) {
            put_bits(&s->pb,
                intra_MCBPC_bits[cbpc],
                intra_MCBPC_code[cbpc]);
        } else {
            put_bits(&s->pb, 1, 0);	/* mb coded */
            put_bits(&s->pb,
                inter_MCBPC_bits[cbpc + 4],
                inter_MCBPC_code[cbpc + 4]);
        }
        put_bits(&s->pb, 1, s->ac_pred);
        cbpy = cbp >> 2;
        put_bits(&s->pb, cbpy_tab[cbpy][1], cbpy_tab[cbpy][0]);

        /* encode each block */
        for (i = 0; i < 6; i++) {
            mpeg4_encode_block(s, block[i], i, dc_diff[i], scan_table[i]);
        }

        /* restore ac coeffs & last_index stuff if we messed them up with the prediction */
        if(s->ac_pred){
            for(i=0; i<6; i++){
                int j;    
                int x, y, wrap;
                INT16 *ac_val;

                if (i < 4) {
                    x = 2 * s->mb_x + 1 + (i & 1);
                    y = 2 * s->mb_y + 1 + (i >> 1);
                    wrap = s->mb_width * 2 + 2;
                    ac_val = s->ac_val[0][0];
                } else {
                    x = s->mb_x + 1;
                    y = s->mb_y + 1;
                    wrap = s->mb_width + 2;
                    ac_val = s->ac_val[i - 4 + 1][0];
                }
                ac_val += ((y) * wrap + (x)) * 16;

                if(dir[i]){
                    for(j=1; j<8; j++) 
                        block[i][block_permute_op(j   )]= ac_val[j+8];
                }else{
                    for(j=1; j<8; j++) 
                        block[i][block_permute_op(j<<3)]= ac_val[j  ];
                }
                s->block_last_index[i]= zigzag_last_index[i];
            }
        }
    }
}

void h263_encode_mb(MpegEncContext * s,
		    DCTELEM block[6][64],
		    int motion_x, int motion_y)
{
    int cbpc, cbpy, i, cbp, pred_x, pred_y;
   
    //    printf("**mb x=%d y=%d\n", s->mb_x, s->mb_y);
   if (!s->mb_intra) {
	   /* compute cbp */
	   cbp = 0;
	   for (i = 0; i < 6; i++) {
	      if (s->block_last_index[i] >= 0)
		   cbp |= 1 << (5 - i);
	   }
	   if ((cbp | motion_x | motion_y) == 0) {
	      /* skip macroblock */
	      put_bits(&s->pb, 1, 1);
	      return;
	   }
	   put_bits(&s->pb, 1, 0);	/* mb coded */
	   cbpc = cbp & 3;
	   put_bits(&s->pb,
		inter_MCBPC_bits[cbpc],
		inter_MCBPC_code[cbpc]);
	   cbpy = cbp >> 2;
	   cbpy ^= 0xf;
	   put_bits(&s->pb, cbpy_tab[cbpy][1], cbpy_tab[cbpy][0]);

	   /* motion vectors: 16x16 mode only now */
      h263_pred_motion(s, 0, &pred_x, &pred_y);
      
      if (!s->umvplus) {  
         h263_encode_motion(s, motion_x - pred_x);
         h263_encode_motion(s, motion_y - pred_y);
      }
      else {
         h263p_encode_umotion(s, motion_x - pred_x);
         h263p_encode_umotion(s, motion_y - pred_y);
         if (((motion_x - pred_x) == 1) && ((motion_y - pred_y) == 1))
            /* To prevent Start Code emulation */
            put_bits(&s->pb,1,1);
      }
   } else {
	/* compute cbp */
	cbp = 0;
	for (i = 0; i < 6; i++) {
	    if (s->block_last_index[i] >= 1)
		cbp |= 1 << (5 - i);
	}

	cbpc = cbp & 3;
	if (s->pict_type == I_TYPE) {
	    put_bits(&s->pb,
		     intra_MCBPC_bits[cbpc],
		     intra_MCBPC_code[cbpc]);
	} else {
	    put_bits(&s->pb, 1, 0);	/* mb coded */
	    put_bits(&s->pb,
		     inter_MCBPC_bits[cbpc + 4],
		     inter_MCBPC_code[cbpc + 4]);
	}
	if (s->h263_pred) {
	    /* XXX: currently, we do not try to use ac prediction */
	    put_bits(&s->pb, 1, 0);	/* no ac prediction */
	}
	cbpy = cbp >> 2;
	put_bits(&s->pb, cbpy_tab[cbpy][1], cbpy_tab[cbpy][0]);
    }

    /* encode each block */
    if (s->h263_pred) {
	for (i = 0; i < 6; i++) {
//	    mpeg4_encode_block(s, block[i], i);
	}
    } else {
	for (i = 0; i < 6; i++) {
	    h263_encode_block(s, block[i], i);
	}
    }
}

void h263_pred_acdc(MpegEncContext * s, INT16 *block, int n)
{
    int x, y, wrap, a, c, pred_dc, scale, i;
    INT16 *dc_val, *ac_val, *ac_val1;

    /* find prediction */
    if (n < 4) {
        x = 2 * s->mb_x + 1 + (n & 1);
        y = 2 * s->mb_y + 1 + ((n & 2) >> 1);
        wrap = s->mb_width * 2 + 2;
        dc_val = s->dc_val[0];
        ac_val = s->ac_val[0][0];
        scale = s->y_dc_scale;
    } else {
        x = s->mb_x + 1;
        y = s->mb_y + 1;
        wrap = s->mb_width + 2;
        dc_val = s->dc_val[n - 4 + 1];
        ac_val = s->ac_val[n - 4 + 1][0];
        scale = s->c_dc_scale;
    }
    
    ac_val += ((y) * wrap + (x)) * 16;
    ac_val1 = ac_val;
    
    /* B C
     * A X 
     */
    a = dc_val[(x - 1) + (y) * wrap];
    c = dc_val[(x) + (y - 1) * wrap];
    
    pred_dc = 1024;
    if (s->ac_pred) {
        if (s->h263_aic_dir) {
            /* left prediction */
            if (a != 1024) {
                ac_val -= 16;
                for(i=1;i<8;i++) {
                    block[block_permute_op(i*8)] += ac_val[i];
                }
                pred_dc = a;
            }
        } else {
            /* top prediction */
            if (c != 1024) {
                ac_val -= 16 * wrap;
                for(i=1;i<8;i++) {
                    block[block_permute_op(i)] += ac_val[i + 8];
                }
                pred_dc = c;
            }
        }
    } else {
        /* just DC prediction */
        if (a != 1024 && c != 1024)
            pred_dc = (a + c) >> 1;
        else if (a != 1024)
            pred_dc = a;
        else
            pred_dc = c;
    }
    
    /* we assume pred is positive */
    block[0]=block[0]*scale + pred_dc;
    
    if (block[0] < 0)
        block[0] = 0;
    else if (!(block[0] & 1))
        block[0]++;
    
    /* Update AC/DC tables */
    dc_val[(x) + (y) * wrap] = block[0];
    
    /* left copy */
    for(i=1;i<8;i++)
        ac_val1[i] = block[block_permute_op(i * 8)];
    /* top copy */
    for(i=1;i<8;i++)
        ac_val1[8 + i] = block[block_permute_op(i)];
}


static inline int mid_pred(int a, int b, int c)
{
    int vmin, vmax;
    vmax = vmin = a;
    if (b < vmin)
        vmin = b;
    else
	vmax = b;

    if (c < vmin)
        vmin = c;
    else if (c > vmax)
        vmax = c;

    return a + b + c - vmin - vmax;
}

INT16 *h263_pred_motion(MpegEncContext * s, int block, 
                        int *px, int *py)
{
    int xy, y, wrap;
    INT16 *A, *B, *C, *mot_val;
    static const int off[4]= {2, 1, 1, -1};

    wrap = 2 * s->mb_width + 2;
    y = xy = 2 * s->mb_y + 1 + (block >> 1); // y
    xy *= wrap; // y * wrap
    xy += 2 * s->mb_x + 1 + (block & 1); // x + y * wrap

    mot_val = s->motion_val[xy];

    /* special case for first line */
    if (y == 1 || s->first_slice_line || s->first_gob_line) {
        A = s->motion_val[xy - 1];
        *px = A[0];
        *py = A[1];
    } else {
        A = s->motion_val[xy - 1];
        B = s->motion_val[xy - wrap];
        C = s->motion_val[xy + off[block] - wrap];
        *px = mid_pred(A[0], B[0], C[0]);
        *py = mid_pred(A[1], B[1], C[1]);
    }
    return mot_val;
}

static void h263_encode_motion(MpegEncContext * s, int val)
{
    int range, l, m, bit_size, sign, code, bits;

    if (val == 0) {
        /* zero vector */
        code = 0;
        put_bits(&s->pb, mvtab[code][1], mvtab[code][0]);
    } else {
        bit_size = s->f_code - 1;
        range = 1 << bit_size;
        /* modulo encoding */
        l = range * 32;
        m = 2 * l;
        if (val < -l) {
            val += m;
        } else if (val >= l) {
            val -= m;
        }

        if (val >= 0) {
            val--;
            code = (val >> bit_size) + 1;
            bits = val & (range - 1);
            sign = 0;
        } else {
            val = -val;
            val--;
            code = (val >> bit_size) + 1;
            bits = val & (range - 1);
            sign = 1;
        }

        put_bits(&s->pb, mvtab[code][1] + 1, (mvtab[code][0] << 1) | sign); 
        if (bit_size > 0) {
            put_bits(&s->pb, bit_size, bits);
        }
    }
}

/* Encode MV differences on H.263+ with Unrestricted MV mode */
static void h263p_encode_umotion(MpegEncContext * s, int val)
{
    short sval = 0; 
    short i = 0;
    short n_bits = 0;
    short temp_val;
    int code = 0;
    int tcode;
    
    if ( val == 0)
        put_bits(&s->pb, 1, 1);
    else if (val == 1)
        put_bits(&s->pb, 3, 0);
    else if (val == -1)
        put_bits(&s->pb, 3, 2);
    else {
        
        sval = ((val < 0) ? (short)(-val):(short)val);
        temp_val = sval;
        
        while (temp_val != 0) {
            temp_val = temp_val >> 1;
            n_bits++;
        }
        
        i = n_bits - 1;
        while (i > 0) {
            tcode = (sval & (1 << (i-1))) >> (i-1);
            tcode = (tcode << 1) | 1;
            code = (code << 2) | tcode;
            i--;
        }
        code = ((code << 1) | (val < 0)) << 1;
        put_bits(&s->pb, (2*n_bits)+1, code);
        //printf("\nVal = %d\tCode = %d", sval, code);
    }
}

void h263_encode_init_vlc(MpegEncContext *s)
{
    static int done = 0;

    if (!done) {
        done = 1;
        init_rl(&rl_inter);
        init_rl(&rl_intra);
    }
}

static void h263_encode_block(MpegEncContext * s, DCTELEM * block, int n)
{
    int level, run, last, i, j, last_index, last_non_zero, sign, slevel;
    int code;
    RLTable *rl = &rl_inter;

    if (s->mb_intra) {
        /* DC coef */
	    level = block[0];
        /* 255 cannot be represented, so we clamp */
        if (level > 254) {
            level = 254;
            block[0] = 254;
        }
        /* 0 cannot be represented also */
        else if (!level) {
            level = 1;
            block[0] = 1;
        }
	    if (level == 128)
	        put_bits(&s->pb, 8, 0xff);
	    else
	        put_bits(&s->pb, 8, level & 0xff);
	    i = 1;
    } else {
	    i = 0;
    }

    /* AC coefs */
    last_index = s->block_last_index[n];
    last_non_zero = i - 1;
    for (; i <= last_index; i++) {
	j = zigzag_direct[i];
	level = block[j];
	if (level) {
	    run = i - last_non_zero - 1;
	    last = (i == last_index);
	    sign = 0;
	    slevel = level;
	    if (level < 0) {
		sign = 1;
		level = -level;
	    }
            code = get_rl_index(rl, last, run, level);
            put_bits(&s->pb, rl->table_vlc[code][1], rl->table_vlc[code][0]);
            if (code == rl->n) {
                put_bits(&s->pb, 1, last);
                put_bits(&s->pb, 6, run);
                put_bits(&s->pb, 8, slevel & 0xff);
            } else {
                put_bits(&s->pb, 1, sign);
            }
	    last_non_zero = i;
	}
    }
}

/***************************************************/

static void mpeg4_stuffing(PutBitContext * pbc)
{
    int length;
    put_bits(pbc, 1, 0);
    length= (-get_bit_count(pbc))&7;
    put_bits(pbc, length, (1<<length)-1);
}

static void put_string(PutBitContext * pbc, char *s)
{
    while(*s){
        put_bits(pbc, 8, *s);
        s++;
    }
    put_bits(pbc, 8, 0);
}

static void mpeg4_encode_vol_header(MpegEncContext * s)
{
    int vo_ver_id=1; //must be 2 if we want GMC or q-pel

    if(get_bit_count(&s->pb)!=0) mpeg4_stuffing(&s->pb);
    put_bits(&s->pb, 16, 0);
    put_bits(&s->pb, 16, 0x100);        /* video obj */
    put_bits(&s->pb, 16, 0);
    put_bits(&s->pb, 16, 0x120);        /* video obj layer */

    put_bits(&s->pb, 1, 0);		/* random access vol */
    put_bits(&s->pb, 8, 1);		/* video obj type indication= simple obj */
    put_bits(&s->pb, 1, 1);		/* is obj layer id= yes */
      put_bits(&s->pb, 4, vo_ver_id);	/* is obj layer ver id */
      put_bits(&s->pb, 3, 1);		/* is obj layer priority */
    put_bits(&s->pb, 4, 1);		/* aspect ratio info= sqare pixel */ //FIXME real aspect
    put_bits(&s->pb, 1, 0);		/* vol control parameters= no */
    put_bits(&s->pb, 2, RECT_SHAPE);	/* vol shape= rectangle */
    put_bits(&s->pb, 1, 1);		/* marker bit */
    put_bits(&s->pb, 16, s->time_increment_resolution=30000);
    s->time_increment_bits = av_log2(s->time_increment_resolution - 1) + 1;
    if (s->time_increment_bits < 1)
        s->time_increment_bits = 1;
    put_bits(&s->pb, 1, 1);		/* marker bit */
    put_bits(&s->pb, 1, 0);		/* fixed vop rate=no */
    put_bits(&s->pb, 1, 1);		/* marker bit */
    put_bits(&s->pb, 13, s->width);	/* vol width */
    put_bits(&s->pb, 1, 1);		/* marker bit */
    put_bits(&s->pb, 13, s->height);	/* vol height */
    put_bits(&s->pb, 1, 1);		/* marker bit */
    put_bits(&s->pb, 1, 0);		/* interlace */
    put_bits(&s->pb, 1, 1);		/* obmc disable */
    if (vo_ver_id == 1) {
        put_bits(&s->pb, 1, s->vol_sprite_usage=0);		/* sprite enable */
    }else{ /* vo_ver_id == 2 */
        put_bits(&s->pb, 2, s->vol_sprite_usage=0);		/* sprite enable */
    }
    put_bits(&s->pb, 1, 0);		/* not 8 bit */
    put_bits(&s->pb, 1, 0);		/* quant type= h263 style*/
    if (vo_ver_id != 1)
        put_bits(&s->pb, 1, s->quarter_sample=0);
    put_bits(&s->pb, 1, 1);		/* complexity estimation disable */
    put_bits(&s->pb, 1, 1);		/* resync marker disable */
    put_bits(&s->pb, 1, 0);		/* data partitioned */
    if (vo_ver_id != 1){
        put_bits(&s->pb, 1, 0);		/* newpred */
        put_bits(&s->pb, 1, 0);		/* reduced res vop */
    }
    put_bits(&s->pb, 1, 0);		/* scalability */

    mpeg4_stuffing(&s->pb);
    put_bits(&s->pb, 16, 0);
    put_bits(&s->pb, 16, 0x1B2);	/* user_data */
    put_string(&s->pb, "ffmpeg"); //FIXME append some version ...

    s->no_rounding = 0;
}

/* write mpeg4 VOP header */
void mpeg4_encode_picture_header(MpegEncContext * s, int picture_number)
{
    if(s->pict_type==I_TYPE) mpeg4_encode_vol_header(s);

    if(get_bit_count(&s->pb)!=0) mpeg4_stuffing(&s->pb);
    put_bits(&s->pb, 16, 0);	        /* vop header */
    put_bits(&s->pb, 16, 0x1B6);	/* vop header */
    put_bits(&s->pb, 2, s->pict_type - 1);	/* pict type: I = 0 , P = 1 */
    /* XXX: time base + 1 not always correct */
    put_bits(&s->pb, 1, 1);
    put_bits(&s->pb, 1, 0);

    put_bits(&s->pb, 1, 1);	/* marker */
    put_bits(&s->pb, s->time_increment_bits, 1);	/* XXX: correct time increment */
    put_bits(&s->pb, 1, 1);	/* marker */
    put_bits(&s->pb, 1, 1);	/* vop coded */
    if (    s->pict_type == P_TYPE 
        || (s->pict_type == S_TYPE && s->vol_sprite_usage==GMC_SPRITE)) {
        s->no_rounding ^= 1;
	put_bits(&s->pb, 1, s->no_rounding);	/* rounding type */
    }
    put_bits(&s->pb, 3, 0);	/* intra dc VLC threshold */
    //FIXME sprite stuff

    put_bits(&s->pb, 5, s->qscale);

    if (s->pict_type != I_TYPE)
	put_bits(&s->pb, 3, s->f_code);	/* fcode_for */
    if (s->pict_type == B_TYPE)
	put_bits(&s->pb, 3, s->b_code);	/* fcode_back */
    //    printf("****frame %d\n", picture_number);
}

void h263_dc_scale(MpegEncContext * s)
{
#if 1
    const static UINT8 y_tab[32]={
    //  0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
        0, 8, 8, 8, 8,10,12,14,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,34,36,38,40,42,44,46
    };
    const static UINT8 c_tab[32]={
    //  0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
        0, 8, 8, 8, 8, 9, 9,10,10,11,11,12,12,13,13,14,14,15,15,16,16,17,17,18,18,19,20,21,22,23,24,25
    };
    s->y_dc_scale = y_tab[s->qscale];
    s->c_dc_scale = c_tab[s->qscale];
#else
    int quant;
    quant = s->qscale;
    /* luminance */
    if (quant < 5)
	s->y_dc_scale = 8;
    else if (quant > 4 && quant < 9)
	s->y_dc_scale = (2 * quant);
    else if (quant > 8 && quant < 25)
	s->y_dc_scale = (quant + 8);
    else
	s->y_dc_scale = (2 * quant - 16);
    /* chrominance */
    if (quant < 5)
	s->c_dc_scale = 8;
    else if (quant > 4 && quant < 25)
	s->c_dc_scale = ((quant + 13) / 2);
    else
	s->c_dc_scale = (quant - 6);
#endif
}

static inline int mpeg4_pred_dc(MpegEncContext * s, int n, UINT16 **dc_val_ptr, int *dir_ptr)
{
    int a, b, c, xy, wrap, pred, scale;
    UINT16 *dc_val;
    int dummy;

    /* find prediction */
    if (n < 4) {
	wrap = s->mb_width * 2 + 2;
	xy = 2 * s->mb_y + 1 + (n >> 1);
        xy *= wrap;
	xy += 2 * s->mb_x + 1 + (n & 1);
	dc_val = s->dc_val[0];
	scale = s->y_dc_scale;
    } else {
	wrap = s->mb_width + 2;
	xy = s->mb_y + 1;
	xy *= wrap;
	xy += s->mb_x + 1;
	dc_val = s->dc_val[n - 4 + 1];
	scale = s->c_dc_scale;
    }

    /* B C
     * A X 
     */
    a = dc_val[xy - 1];
    b = dc_val[xy - 1 - wrap];
    c = dc_val[xy - wrap];

    if (abs(a - b) < abs(b - c)) {
	pred = c;
        *dir_ptr = 1; /* top */
    } else {
	pred = a;
        *dir_ptr = 0; /* left */
    }
    /* we assume pred is positive */
#ifdef ARCH_X86
	asm volatile (
		"xorl %%edx, %%edx	\n\t"
		"mul %%ecx		\n\t"
		: "=d" (pred), "=a"(dummy)
		: "a" (pred + (scale >> 1)), "c" (inverse[scale])
	);
#else
    pred = (pred + (scale >> 1)) / scale;
#endif

    /* prepare address for prediction update */
    *dc_val_ptr = &dc_val[xy];

    return pred;
}

void mpeg4_pred_ac(MpegEncContext * s, INT16 *block, int n,
                   int dir)
{
    int x, y, wrap, i;
    INT16 *ac_val, *ac_val1;

    /* find prediction */
    if (n < 4) {
	x = 2 * s->mb_x + 1 + (n & 1);
	y = 2 * s->mb_y + 1 + (n >> 1);
	wrap = s->mb_width * 2 + 2;
	ac_val = s->ac_val[0][0];
    } else {
	x = s->mb_x + 1;
	y = s->mb_y + 1;
	wrap = s->mb_width + 2;
	ac_val = s->ac_val[n - 4 + 1][0];
    }
    ac_val += ((y) * wrap + (x)) * 16;
    ac_val1 = ac_val;
    if (s->ac_pred) {
        if (dir == 0) {
            /* left prediction */
            ac_val -= 16;
            for(i=1;i<8;i++) {
                block[block_permute_op(i*8)] += ac_val[i];
            }
        } else {
            /* top prediction */
            ac_val -= 16 * wrap;
            for(i=1;i<8;i++) {
                block[block_permute_op(i)] += ac_val[i + 8];
            }
        }
    }
    /* left copy */
    for(i=1;i<8;i++)
        ac_val1[i] = block[block_permute_op(i * 8)];
    /* top copy */
    for(i=1;i<8;i++)
        ac_val1[8 + i] = block[block_permute_op(i)];
}

static void mpeg4_inv_pred_ac(MpegEncContext * s, INT16 *block, int n,
                              int dir)
{
    int x, y, wrap, i;
    INT16 *ac_val;

    /* find prediction */
    if (n < 4) {
	x = 2 * s->mb_x + 1 + (n & 1);
	y = 2 * s->mb_y + 1 + (n >> 1);
	wrap = s->mb_width * 2 + 2;
	ac_val = s->ac_val[0][0];
    } else {
	x = s->mb_x + 1;
	y = s->mb_y + 1;
	wrap = s->mb_width + 2;
	ac_val = s->ac_val[n - 4 + 1][0];
    }
    ac_val += ((y) * wrap + (x)) * 16;
 
    if (dir == 0) {
        /* left prediction */
        ac_val -= 16;
        for(i=1;i<8;i++) {
            block[block_permute_op(i*8)] -= ac_val[i];
        }
    } else {
        /* top prediction */
        ac_val -= 16 * wrap;
        for(i=1;i<8;i++) {
            block[block_permute_op(i)] -= ac_val[i + 8];
        }
    }
}


static inline void mpeg4_encode_dc(MpegEncContext * s, int level, int n)
{
    int size, v;
    /* find number of bits */
    size = 0;
    v = abs(level);
    while (v) {
	v >>= 1;
	size++;
    }

    if (n < 4) {
	/* luminance */
	put_bits(&s->pb, DCtab_lum[size][1], DCtab_lum[size][0]);
    } else {
	/* chrominance */
	put_bits(&s->pb, DCtab_chrom[size][1], DCtab_chrom[size][0]);
    }

    /* encode remaining bits */
    if (size > 0) {
	if (level < 0)
	    level = (-level) ^ ((1 << size) - 1);
	put_bits(&s->pb, size, level);
	if (size > 8)
	    put_bits(&s->pb, 1, 1);
    }
}

static void mpeg4_encode_block(MpegEncContext * s, DCTELEM * block, int n, int intra_dc, UINT8 *scan_table)
{
    int level, run, last, i, j, last_index, last_non_zero, sign, slevel;
    int code;
    const RLTable *rl;

    if (s->mb_intra) {
	/* mpeg4 based DC predictor */
	mpeg4_encode_dc(s, intra_dc, n);
	i = 1;
        rl = &rl_intra;
    } else {
	i = 0;
        rl = &rl_inter;
    }

    /* AC coefs */
    last_index = s->block_last_index[n];
    last_non_zero = i - 1;
    for (; i <= last_index; i++) {
	j = scan_table[i];
	level = block[j];
	if (level) {
	    run = i - last_non_zero - 1;
	    last = (i == last_index);
	    sign = 0;
	    slevel = level;
	    if (level < 0) {
		sign = 1;
		level = -level;
	    }
            code = get_rl_index(rl, last, run, level);
            put_bits(&s->pb, rl->table_vlc[code][1], rl->table_vlc[code][0]);
            if (code == rl->n) {
                int level1, run1;
                level1 = level - rl->max_level[last][run];
                if (level1 < 1) 
                    goto esc2;
                code = get_rl_index(rl, last, run, level1);
                if (code == rl->n) {
                esc2:
                    put_bits(&s->pb, 1, 1);
                    if (level > MAX_LEVEL)
                        goto esc3;
                    run1 = run - rl->max_run[last][level] - 1;
                    if (run1 < 0)
                        goto esc3;
                    code = get_rl_index(rl, last, run1, level);
                    if (code == rl->n) {
                    esc3:
                        /* third escape */
                        put_bits(&s->pb, 1, 1);
                        put_bits(&s->pb, 1, last);
                        put_bits(&s->pb, 6, run);
                        put_bits(&s->pb, 1, 1);
                        put_bits(&s->pb, 12, slevel & 0xfff);
                        put_bits(&s->pb, 1, 1);
                    } else {
                        /* second escape */
                        put_bits(&s->pb, 1, 0);
                        put_bits(&s->pb, rl->table_vlc[code][1], rl->table_vlc[code][0]);
                        put_bits(&s->pb, 1, sign);
                    }
                } else {
                    /* first escape */
                    put_bits(&s->pb, 1, 0);
                    put_bits(&s->pb, rl->table_vlc[code][1], rl->table_vlc[code][0]);
                    put_bits(&s->pb, 1, sign);
                }
            } else {
                put_bits(&s->pb, 1, sign);
            }
	    last_non_zero = i;
	}
    }
}



/***********************************************/
/* decoding */

static VLC intra_MCBPC_vlc;
static VLC inter_MCBPC_vlc;
static VLC cbpy_vlc;
static VLC mv_vlc;
static VLC dc_lum, dc_chrom;
static VLC sprite_trajectory;
static VLC mb_type_b_vlc;

void init_rl(RLTable *rl)
{
    INT8 max_level[MAX_RUN+1], max_run[MAX_LEVEL+1];
    UINT8 index_run[MAX_RUN+1];
    int last, run, level, start, end, i;

    /* compute max_level[], max_run[] and index_run[] */
    for(last=0;last<2;last++) {
        if (last == 0) {
            start = 0;
            end = rl->last;
        } else {
            start = rl->last;
            end = rl->n;
        }

        memset(max_level, 0, MAX_RUN + 1);
        memset(max_run, 0, MAX_LEVEL + 1);
        memset(index_run, rl->n, MAX_RUN + 1);
        for(i=start;i<end;i++) {
            run = rl->table_run[i];
            level = rl->table_level[i];
            if (index_run[run] == rl->n)
                index_run[run] = i;
            if (level > max_level[run])
                max_level[run] = level;
            if (run > max_run[level])
                max_run[level] = run;
        }
        rl->max_level[last] = malloc(MAX_RUN + 1);
        memcpy(rl->max_level[last], max_level, MAX_RUN + 1);
        rl->max_run[last] = malloc(MAX_LEVEL + 1);
        memcpy(rl->max_run[last], max_run, MAX_LEVEL + 1);
        rl->index_run[last] = malloc(MAX_RUN + 1);
        memcpy(rl->index_run[last], index_run, MAX_RUN + 1);
    }
}

void init_vlc_rl(RLTable *rl)
{
    init_vlc(&rl->vlc, 9, rl->n + 1, 
             &rl->table_vlc[0][1], 4, 2,
             &rl->table_vlc[0][0], 4, 2);
}

/* init vlcs */

/* XXX: find a better solution to handle static init */
void h263_decode_init_vlc(MpegEncContext *s)
{
    static int done = 0;

    if (!done) {
        done = 1;

        init_vlc(&intra_MCBPC_vlc, 6, 8, 
                 intra_MCBPC_bits, 1, 1,
                 intra_MCBPC_code, 1, 1);
        init_vlc(&inter_MCBPC_vlc, 9, 25, 
                 inter_MCBPC_bits, 1, 1,
                 inter_MCBPC_code, 1, 1);
        init_vlc(&cbpy_vlc, 6, 16,
                 &cbpy_tab[0][1], 2, 1,
                 &cbpy_tab[0][0], 2, 1);
        init_vlc(&mv_vlc, 9, 33,
                 &mvtab[0][1], 2, 1,
                 &mvtab[0][0], 2, 1);
        init_rl(&rl_inter);
        init_rl(&rl_intra);
        init_rl(&rl_intra_aic);
        init_vlc_rl(&rl_inter);
        init_vlc_rl(&rl_intra);
        init_vlc_rl(&rl_intra_aic);
        init_vlc(&dc_lum, 9, 13,
                 &DCtab_lum[0][1], 2, 1,
                 &DCtab_lum[0][0], 2, 1);
        init_vlc(&dc_chrom, 9, 13,
                 &DCtab_chrom[0][1], 2, 1,
                 &DCtab_chrom[0][0], 2, 1);
        init_vlc(&sprite_trajectory, 9, 15,
                 &sprite_trajectory_tab[0][1], 4, 2,
                 &sprite_trajectory_tab[0][0], 4, 2);
        init_vlc(&mb_type_b_vlc, 4, 4,
                 &mb_type_b_tab[0][1], 2, 1,
                 &mb_type_b_tab[0][0], 2, 1);
    }
}

int h263_decode_gob_header(MpegEncContext *s)
{
    unsigned int val, gfid;
    
    /* Check for GOB Start Code */
    val = show_bits(&s->gb, 16);
    if (val == 0) {
        /* We have a GBSC probably with GSTUFF */
        skip_bits(&s->gb, 16); /* Drop the zeros */
        while (get_bits1(&s->gb) == 0); /* Seek the '1' bit */
#ifdef DEBUG
        fprintf(stderr,"\nGOB Start Code at MB %d\n", (s->mb_y * s->mb_width) + s->mb_x);
#endif
        s->gob_number = get_bits(&s->gb, 5); /* GN */
        gfid = get_bits(&s->gb, 2); /* GFID */
        s->qscale = get_bits(&s->gb, 5); /* GQUANT */
#ifdef DEBUG
        fprintf(stderr, "\nGN: %u GFID: %u Quant: %u\n", s->gob_number, gfid, s->qscale);
#endif
        return 1;
    }
    return 0;
            
}

int h263_decode_mb(MpegEncContext *s,
                   DCTELEM block[6][64])
{
    int cbpc, cbpy, i, cbp, pred_x, pred_y, mx, my, dquant;
    INT16 *mot_val;
    static INT8 quant_tab[4] = { -1, -2, 1, 2 };

    if (s->pict_type == P_TYPE || s->pict_type==S_TYPE) {
        if (get_bits1(&s->gb)) {
            /* skip mb */
            s->mb_intra = 0;
            for(i=0;i<6;i++)
                s->block_last_index[i] = -1;
            s->mv_dir = MV_DIR_FORWARD;
            s->mv_type = MV_TYPE_16X16;
            if(s->pict_type==S_TYPE && s->vol_sprite_usage==GMC_SPRITE){
                const int a= s->sprite_warping_accuracy;
//                int l = (1 << (s->f_code - 1)) * 32;

                s->mcsel=1;
                s->mv[0][0][0] = RSHIFT(s->sprite_offset[0][0], a-s->quarter_sample);
                s->mv[0][0][1] = RSHIFT(s->sprite_offset[0][1], a-s->quarter_sample);
/*                if (s->mv[0][0][0] < -l) s->mv[0][0][0]= -l;
                else if (s->mv[0][0][0] >= l) s->mv[0][0][0]= l-1;
                if (s->mv[0][0][1] < -l) s->mv[0][0][1]= -l;
                else if (s->mv[0][0][1] >= l) s->mv[0][0][1]= l-1;*/

                s->mb_skiped = 0;
            }else{
                s->mcsel=0;
                s->mv[0][0][0] = 0;
                s->mv[0][0][1] = 0;
                s->mb_skiped = 1;
            }
            return 0;
        }
        cbpc = get_vlc(&s->gb, &inter_MCBPC_vlc);
        //fprintf(stderr, "\tCBPC: %d", cbpc);
        if (cbpc < 0)
            return -1;
        if (cbpc > 20)
            cbpc+=3;
        else if (cbpc == 20)
            fprintf(stderr, "Stuffing !");
        
        dquant = cbpc & 8;
        s->mb_intra = ((cbpc & 4) != 0);
        if (s->mb_intra) goto intra;
        
        if(s->pict_type==S_TYPE && s->vol_sprite_usage==GMC_SPRITE && (cbpc & 16) == 0)
            s->mcsel= get_bits1(&s->gb);
        else s->mcsel= 0;
        cbpy = get_vlc(&s->gb, &cbpy_vlc);
        cbp = (cbpc & 3) | ((cbpy ^ 0xf) << 2);
        if (dquant) {
            s->qscale += quant_tab[get_bits(&s->gb, 2)];
            if (s->qscale < 1)
                s->qscale = 1;
            else if (s->qscale > 31)
                s->qscale = 31;
        }
        s->mv_dir = MV_DIR_FORWARD;
        if ((cbpc & 16) == 0) {
            /* 16x16 motion prediction */
            s->mv_type = MV_TYPE_16X16;
            h263_pred_motion(s, 0, &pred_x, &pred_y);
            if (s->umvplus_dec)
               mx = h263p_decode_umotion(s, pred_x);
            else if(!s->mcsel)
               mx = h263_decode_motion(s, pred_x, s->f_code);
            else {
               const int a= s->sprite_warping_accuracy;
//        int l = (1 << (s->f_code - 1)) * 32;
               mx= RSHIFT(s->sprite_offset[0][0], a-s->quarter_sample);
//        if (mx < -l) mx= -l;
//        else if (mx >= l) mx= l-1;
            }
            if (mx >= 0xffff)
                return -1;
            
            if (s->umvplus_dec)
               my = h263p_decode_umotion(s, pred_y);
            else if(!s->mcsel)
               my = h263_decode_motion(s, pred_y, s->f_code);
            else{
               const int a= s->sprite_warping_accuracy;
//       int l = (1 << (s->f_code - 1)) * 32;
               my= RSHIFT(s->sprite_offset[0][1], a-s->quarter_sample);
//       if (my < -l) my= -l;
//       else if (my >= l) my= l-1;
            }
            if (my >= 0xffff)
                return -1;
            s->mv[0][0][0] = mx;
            s->mv[0][0][1] = my;
            /*fprintf(stderr, "\n MB %d", (s->mb_y * s->mb_width) + s->mb_x);
            fprintf(stderr, "\n\tmvx: %d\t\tpredx: %d", mx, pred_x);
            fprintf(stderr, "\n\tmvy: %d\t\tpredy: %d", my, pred_y);*/
            if (s->umvplus_dec && (mx - pred_x) == 1 && (my - pred_y) == 1)
               skip_bits1(&s->gb); /* Bit stuffing to prevent PSC */
                           
        } else {
            s->mv_type = MV_TYPE_8X8;
            for(i=0;i<4;i++) {
                mot_val = h263_pred_motion(s, i, &pred_x, &pred_y);
                if (s->umvplus_dec)
                  mx = h263p_decode_umotion(s, pred_x);
                else
                  mx = h263_decode_motion(s, pred_x, s->f_code);
                if (mx >= 0xffff)
                    return -1;
                
                if (s->umvplus_dec)
                  my = h263p_decode_umotion(s, pred_y);
                else    
                  my = h263_decode_motion(s, pred_y, s->f_code);
                if (my >= 0xffff)
                    return -1;
                s->mv[0][i][0] = mx;
                s->mv[0][i][1] = my;
                if (s->umvplus_dec && (mx - pred_x) == 1 && (my - pred_y) == 1)
                  skip_bits1(&s->gb); /* Bit stuffing to prevent PSC */
                mot_val[0] = mx;
                mot_val[1] = my;
            }
        }
    } else if(s->pict_type==B_TYPE) {
        int modb1; // first bit of modb
        int modb2; // second bit of modb
        int mb_type;
        int time_pp;
        int time_pb;
        int xy;

        s->mb_intra = 0; //B-frames never contain intra blocks
        s->mcsel=0;      //     ...               true gmc blocks

        if(s->mb_x==0){
            s->last_mv[0][0][0]= 
            s->last_mv[0][0][1]= 
            s->last_mv[1][0][0]= 
            s->last_mv[1][0][1]= 0;
        }

        /* if we skipped it in the future P Frame than skip it now too */
        s->mb_skiped= s->mbskip_table[s->mb_y * s->mb_width + s->mb_x]; // Note, skiptab=0 if last was GMC

        if(s->mb_skiped){
                /* skip mb */
            for(i=0;i<6;i++)
                s->block_last_index[i] = -1;

            s->mv_dir = MV_DIR_FORWARD;
            s->mv_type = MV_TYPE_16X16;
            s->mv[0][0][0] = 0;
            s->mv[0][0][1] = 0;
            s->mv[1][0][0] = 0;
            s->mv[1][0][1] = 0;
            s->last_mv[0][0][0]=
            s->last_mv[0][0][1]= 
            s->last_mv[1][0][0]= 
            s->last_mv[1][0][1]= 0;
            s->mb_skiped = 1;
            return 0;
        }

        modb1= get_bits1(&s->gb);
        if(modb1==0){
            modb2= get_bits1(&s->gb);
            mb_type= get_vlc(&s->gb, &mb_type_b_vlc);
            if(modb2==0) cbp= get_bits(&s->gb, 6);
            else cbp=0;
            if (mb_type && cbp) {
                if(get_bits1(&s->gb)){
                    s->qscale +=get_bits1(&s->gb)*4 - 2;
                    if (s->qscale < 1)
                        s->qscale = 1;
                    else if (s->qscale > 31)
                        s->qscale = 31;
                }
            }
        }else{
            mb_type=4; //like 0 but no vectors coded
            cbp=0;
        }
        s->mv_type = MV_TYPE_16X16; // we'll switch to 8x8 only if the last P frame had 8x8 for this MB and mb_type=0 here
        mx=my=0; //for case 4, we could put this to the mb_type=4 but than gcc compains about uninitalized mx/my
        switch(mb_type)
        {
        case 0: 
            mx = h263_decode_motion(s, 0, 1);
            my = h263_decode_motion(s, 0, 1);
        case 4: 
            s->mv_dir = MV_DIR_FORWARD | MV_DIR_BACKWARD | MV_DIRECT;
            xy= (2*s->mb_width + 2)*(2*s->mb_y + 1) + 2*s->mb_x + 1;
            time_pp= s->last_non_b_time[0] - s->last_non_b_time[1];
            time_pb= s->time - s->last_non_b_time[1];
//if(time_pp>3000 )printf("%d %d  ", time_pp, time_pb);
            //FIXME 4MV
            //FIXME avoid divides
            s->mv[0][0][0] = s->motion_val[xy][0]*time_pb/time_pp + mx;
            s->mv[0][0][1] = s->motion_val[xy][1]*time_pb/time_pp + my;
            s->mv[1][0][0] = mx ? s->mv[0][0][0] - s->motion_val[xy][0]
                                : s->motion_val[xy][0]*(time_pb - time_pp)/time_pp + mx;
            s->mv[1][0][1] = my ? s->mv[0][0][1] - s->motion_val[xy][1] 
                                : s->motion_val[xy][1]*(time_pb - time_pp)/time_pp + my;
/*            s->mv[0][0][0] = 
            s->mv[0][0][1] = 
            s->mv[1][0][0] = 
            s->mv[1][0][1] = 1000;*/
            break;
        case 1: 
            s->mv_dir = MV_DIR_FORWARD | MV_DIR_BACKWARD;
            mx = h263_decode_motion(s, s->last_mv[0][0][0], s->f_code);
            my = h263_decode_motion(s, s->last_mv[0][0][1], s->f_code);
            s->last_mv[0][0][0]= s->mv[0][0][0] = mx;
            s->last_mv[0][0][1]= s->mv[0][0][1] = my;

            mx = h263_decode_motion(s, s->last_mv[1][0][0], s->b_code);
            my = h263_decode_motion(s, s->last_mv[1][0][1], s->b_code);
            s->last_mv[1][0][0]= s->mv[1][0][0] = mx;
            s->last_mv[1][0][1]= s->mv[1][0][1] = my;
            break;
        case 2: 
            s->mv_dir = MV_DIR_BACKWARD;
            mx = h263_decode_motion(s, s->last_mv[1][0][0], s->b_code);
            my = h263_decode_motion(s, s->last_mv[1][0][1], s->b_code);
            s->last_mv[1][0][0]= s->mv[1][0][0] = mx;
            s->last_mv[1][0][1]= s->mv[1][0][1] = my;
            break;
        case 3:
            s->mv_dir = MV_DIR_FORWARD;
            mx = h263_decode_motion(s, s->last_mv[0][0][0], s->f_code);
            my = h263_decode_motion(s, s->last_mv[0][0][1], s->f_code);
            s->last_mv[0][0][0]= s->mv[0][0][0] = mx;
            s->last_mv[0][0][1]= s->mv[0][0][1] = my;
            break;
        default: return -1;
        }
    } else { /* I-Frame */
        cbpc = get_vlc(&s->gb, &intra_MCBPC_vlc);
        if (cbpc < 0)
            return -1;
        dquant = cbpc & 4;
        s->mb_intra = 1;
intra:
        s->ac_pred = 0;
        if (s->h263_pred || s->h263_aic) {
            s->ac_pred = get_bits1(&s->gb);
            if (s->ac_pred && s->h263_aic)
                s->h263_aic_dir = get_bits1(&s->gb);
        }
        if (s->h263_aic) {
            s->y_dc_scale = 2 * s->qscale;
            s->c_dc_scale = 2 * s->qscale;
        }
        cbpy = get_vlc(&s->gb, &cbpy_vlc);
        cbp = (cbpc & 3) | (cbpy << 2);
        if (dquant) {
            s->qscale += quant_tab[get_bits(&s->gb, 2)];
            if (s->qscale < 1)
                s->qscale = 1;
            else if (s->qscale > 31)
                s->qscale = 31;
        }
    }

    /* decode each block */
    if (s->h263_pred) {
	for (i = 0; i < 6; i++) {
	    if (mpeg4_decode_block(s, block[i], i, (cbp >> (5 - i)) & 1) < 0)
                return -1;
	}
    } else {
	for (i = 0; i < 6; i++) {
	    if (h263_decode_block(s, block[i], i, (cbp >> (5 - i)) & 1) < 0)
                return -1;
	}
    }
    return 0;
}

static int h263_decode_motion(MpegEncContext * s, int pred, int f_code)
{
    int code, val, sign, shift, l, m;

    code = get_vlc(&s->gb, &mv_vlc);
    if (code < 0)
        return 0xffff;

    if (code == 0)
        return pred;
    sign = get_bits1(&s->gb);
    shift = f_code - 1;
    val = (code - 1) << shift;
    if (shift > 0)
        val |= get_bits(&s->gb, shift);
    val++;
    if (sign)
        val = -val;
    val += pred;
    
    /* modulo decoding */
    if (!s->h263_long_vectors) {
        l = (1 << (f_code - 1)) * 32;
        m = 2 * l;
        if (val < -l) {
            val += m;
        } else if (val >= l) {
            val -= m;
        }
    } else {
        /* horrible h263 long vector mode */
        if (pred < -31 && val < -63)
            val += 64;
        if (pred > 32 && val > 63)
            val -= 64;
        
    }
    return val;
}

/* Decodes RVLC of H.263+ UMV */
static int h263p_decode_umotion(MpegEncContext * s, int pred)
{
   int code = 0, sign;
   
   if (get_bits1(&s->gb)) /* Motion difference = 0 */
      return pred;
   
   code = 2 + get_bits1(&s->gb);
   
   while (get_bits1(&s->gb))
   {
      code <<= 1;
      code += get_bits1(&s->gb);
   }
   sign = code & 1;
   code >>= 1;
   
   code = (sign) ? (pred - code) : (pred + code);
#ifdef DEBUG
   fprintf(stderr,"H.263+ UMV Motion = %d\n", code);
#endif
   return code;   

}

static int h263_decode_block(MpegEncContext * s, DCTELEM * block,
                             int n, int coded)
{
    int code, level, i, j, last, run;
    RLTable *rl = &rl_inter;
    const UINT8 *scan_table;

    scan_table = zigzag_direct;
    if (s->h263_aic && s->mb_intra) {
        rl = &rl_intra_aic;
        i = 0;
        if (s->ac_pred) {
            if (s->h263_aic_dir) 
                scan_table = ff_alternate_vertical_scan; /* left */
            else
                scan_table = ff_alternate_horizontal_scan; /* top */
        }
    } else if (s->mb_intra) {
        /* DC coef */
        if (s->h263_rv10 && s->rv10_version == 3 && s->pict_type == I_TYPE) {
            int component, diff;
            component = (n <= 3 ? 0 : n - 4 + 1);
            level = s->last_dc[component];
            if (s->rv10_first_dc_coded[component]) {
                diff = rv_decode_dc(s, n);
                if (diff == 0xffff)
                    return -1;
                level += diff;
                level = level & 0xff; /* handle wrap round */
                s->last_dc[component] = level;
            } else {
                s->rv10_first_dc_coded[component] = 1;
            }
        } else {
            level = get_bits(&s->gb, 8);
            if (level == 255)
                level = 128;
        }
        block[0] = level;
        i = 1;
    } else {
        i = 0;
    }
    if (!coded) {
        if (s->mb_intra && s->h263_aic)
            goto not_coded;
        s->block_last_index[n] = i - 1;
        return 0;
    }

    for(;;) {
        code = get_vlc(&s->gb, &rl->vlc);
        if (code < 0)
            return -1;
        if (code == rl->n) {
            /* escape */
            last = get_bits1(&s->gb);
            run = get_bits(&s->gb, 6);
            level = (INT8)get_bits(&s->gb, 8);
            if (s->h263_rv10 && level == -128) {
                /* XXX: should patch encoder too */
                level = get_bits(&s->gb, 12);
                level = (level << 20) >> 20;
            }
        } else {
            run = rl->table_run[code];
            level = rl->table_level[code];
            last = code >= rl->last;
            if (get_bits1(&s->gb))
                level = -level;
        }
        i += run;
        if (i >= 64)
            return -1;
        j = scan_table[i];
        block[j] = level;
        if (last)
            break;
        i++;
    }
not_coded:    
    if (s->mb_intra && s->h263_aic) {
        h263_pred_acdc(s, block, n);
        i = 63;
    }
    s->block_last_index[n] = i;
    return 0;
}

static int mpeg4_decode_dc(MpegEncContext * s, int n, int *dir_ptr)
{
    int level, pred, code;
    UINT16 *dc_val;

    if (n < 4) 
        code = get_vlc(&s->gb, &dc_lum);
    else 
        code = get_vlc(&s->gb, &dc_chrom);
    if (code < 0)
        return -1;
    if (code == 0) {
        level = 0;
    } else {
        level = get_bits(&s->gb, code);
        if ((level >> (code - 1)) == 0) /* if MSB not set it is negative*/
            level = - (level ^ ((1 << code) - 1));
        if (code > 8)
            skip_bits1(&s->gb); /* marker */
    }

    pred = mpeg4_pred_dc(s, n, &dc_val, dir_ptr);
    level += pred;
    if (level < 0)
        level = 0;
    if (n < 4) {
        *dc_val = level * s->y_dc_scale;
    } else {
        *dc_val = level * s->c_dc_scale;
    }
    return level;
}

static int mpeg4_decode_block(MpegEncContext * s, DCTELEM * block,
                              int n, int coded)
{
    int code, level, i, j, last, run;
    int dc_pred_dir;
    RLTable *rl;
    const UINT8 *scan_table;

    if (s->mb_intra) {
	/* DC coef */
        level = mpeg4_decode_dc(s, n, &dc_pred_dir);
        if (level < 0)
            return -1;
        block[0] = level;
	i = 1;
        if (!coded) 
            goto not_coded;
        rl = &rl_intra;
        if (s->ac_pred) {
            if (dc_pred_dir == 0) 
                scan_table = ff_alternate_vertical_scan; /* left */
            else
                scan_table = ff_alternate_horizontal_scan; /* top */
        } else {
            scan_table = zigzag_direct;
        }
    } else {
	i = 0;
        if (!coded) {
            s->block_last_index[n] = i - 1;
            return 0;
        }
        rl = &rl_inter;
        scan_table = zigzag_direct;
    }

    for(;;) {
        code = get_vlc(&s->gb, &rl->vlc);
        if (code < 0)
            return -1;
        if (code == rl->n) {
            /* escape */
            if (get_bits1(&s->gb) != 0) {
                if (get_bits1(&s->gb) != 0) {
                    /* third escape */
                    last = get_bits1(&s->gb);
                    run = get_bits(&s->gb, 6);
                    get_bits1(&s->gb); /* marker */
                    level = get_bits(&s->gb, 12);
                    level = (level << 20) >> 20; /* sign extend */
                    skip_bits1(&s->gb); /* marker */
                } else {
                    /* second escape */
                    code = get_vlc(&s->gb, &rl->vlc);
                    if (code < 0 || code >= rl->n)
                        return -1;
                    run = rl->table_run[code];
                    level = rl->table_level[code];
                    last = code >= rl->last;
                    run += rl->max_run[last][level] + 1;
                    if (get_bits1(&s->gb))
                        level = -level;
                }
            } else {
                /* first escape */
                code = get_vlc(&s->gb, &rl->vlc);
                if (code < 0 || code >= rl->n)
                    return -1;
                run = rl->table_run[code];
                level = rl->table_level[code];
                last = code >= rl->last;
                level += rl->max_level[last][run];
                if (get_bits1(&s->gb))
                    level = -level;
            }
        } else {
            run = rl->table_run[code];
            level = rl->table_level[code];
            last = code >= rl->last;
            if (get_bits1(&s->gb))
                level = -level;
        }
        i += run;
        if (i >= 64)
            return -1;
	j = scan_table[i];
        block[j] = level;
        i++;
        if (last)
            break;
    }
 not_coded:
    if (s->mb_intra) {
        mpeg4_pred_ac(s, block, n, dc_pred_dir);
        if (s->ac_pred) {
            i = 64; /* XXX: not optimal */
        }
    }
    s->block_last_index[n] = i - 1;
    return 0;
}

/* most is hardcoded. should extend to handle all h263 streams */
int h263_decode_picture_header(MpegEncContext *s)
{
    int format, width, height;

    /* picture header */
    if (get_bits(&s->gb, 22) != 0x20)
        return -1;
    s->picture_number = get_bits(&s->gb, 8); /* picture timestamp */
    
    if (get_bits1(&s->gb) != 1)
        return -1;	/* marker */
    if (get_bits1(&s->gb) != 0)
        return -1;	/* h263 id */
    skip_bits1(&s->gb);	/* split screen off */
    skip_bits1(&s->gb);	/* camera  off */
    skip_bits1(&s->gb);	/* freeze picture release off */

    /* Reset GOB number */
    s->gob_number = 0;
        
    format = get_bits(&s->gb, 3);

    if (format != 7 && format != 6) {
        s->h263_plus = 0;
        /* H.263v1 */
        width = h263_format[format][0];
        height = h263_format[format][1];
        if (!width)
            return -1;
        
        s->width = width;
        s->height = height;
        s->pict_type = I_TYPE + get_bits1(&s->gb);

        s->unrestricted_mv = get_bits1(&s->gb); 
        s->h263_long_vectors = s->unrestricted_mv;

        if (get_bits1(&s->gb) != 0)
            return -1;	/* SAC: off */
        if (get_bits1(&s->gb) != 0) {
            s->mv_type = MV_TYPE_8X8; /* Advanced prediction mode */
        }   
        
        if (get_bits1(&s->gb) != 0)
            return -1;	/* not PB frame */

        s->qscale = get_bits(&s->gb, 5);
        skip_bits1(&s->gb);	/* Continuous Presence Multipoint mode: off */
    } else {
        int ufep;
        
        /* H.263v2 */
        s->h263_plus = 1;
        ufep = get_bits(&s->gb, 3); /* Update Full Extended PTYPE */
        
        if (ufep == 1) {
            /* OPPTYPE */       
            format = get_bits(&s->gb, 3);
            skip_bits(&s->gb,1); /* Custom PCF */
            s->umvplus_dec = get_bits(&s->gb, 1); /* Unrestricted Motion Vector */
            skip_bits1(&s->gb); /* Syntax-based Arithmetic Coding (SAC) */
            if (get_bits1(&s->gb) != 0) {
                s->mv_type = MV_TYPE_8X8; /* Advanced prediction mode */
            }
            if (get_bits1(&s->gb) != 0) { /* Advanced Intra Coding (AIC) */
                s->h263_aic = 1;
            }
            skip_bits(&s->gb, 7);
            skip_bits(&s->gb, 3); /* Reserved */
        } else if (ufep != 0)
            return -1;
            
        /* MPPTYPE */
        s->pict_type = get_bits(&s->gb, 3) + 1;
        if (s->pict_type != I_TYPE &&
            s->pict_type != P_TYPE)
            return -1;
        skip_bits(&s->gb, 2);
        s->no_rounding = get_bits1(&s->gb);
        //fprintf(stderr, "\nRTYPE: %d", s->no_rounding);
        skip_bits(&s->gb, 4);
        
        /* Get the picture dimensions */
        if (ufep) {
            if (format == 6) {
                /* Custom Picture Format (CPFMT) */
                skip_bits(&s->gb, 4); /* aspect ratio */
                width = (get_bits(&s->gb, 9) + 1) * 4;
                skip_bits1(&s->gb);
                height = get_bits(&s->gb, 9) * 4;
#ifdef DEBUG 
                fprintf(stderr,"\nH.263+ Custom picture: %dx%d\n",width,height);
#endif            
            }
            else {
                width = h263_format[format][0];
                height = h263_format[format][1];
            }
            if ((width == 0) || (height == 0))
                return -1;
            s->width = width;
            s->height = height;
            if (s->umvplus_dec) {
                skip_bits1(&s->gb); /* Unlimited Unrestricted Motion Vectors Indicator (UUI) */
            }
        }
            
        s->qscale = get_bits(&s->gb, 5);
    }
    /* PEI */
    while (get_bits1(&s->gb) != 0) {
        skip_bits(&s->gb, 8);
    }
    s->f_code = 1;
    return 0;
}

static void mpeg4_decode_sprite_trajectory(MpegEncContext * s)
{
    int i;
    int a= 2<<s->sprite_warping_accuracy;
    int rho= 3-s->sprite_warping_accuracy;
    int r=16/a;
    const int vop_ref[4][2]= {{0,0}, {s->width,0}, {0, s->height}, {s->width, s->height}}; // only true for rectangle shapes
    int d[4][2]={{0,0}, {0,0}, {0,0}, {0,0}};
    int sprite_ref[4][2];
    int virtual_ref[2][2];
    int w2, h2;
    int alpha=0, beta=0;
    int w= s->width;
    int h= s->height;
//printf("SP %d\n", s->sprite_warping_accuracy);
    for(i=0; i<s->num_sprite_warping_points; i++){
        int length;
        int x=0, y=0;

        length= get_vlc(&s->gb, &sprite_trajectory);
        if(length){
            x= get_bits(&s->gb, length);
//printf("lx %d %d\n", length, x);
            if ((x >> (length - 1)) == 0) /* if MSB not set it is negative*/
                x = - (x ^ ((1 << length) - 1));
        }
        if(!(s->divx_version==500 && s->divx_build==413)) skip_bits1(&s->gb); /* marker bit */
        
        length= get_vlc(&s->gb, &sprite_trajectory);
        if(length){
            y=get_bits(&s->gb, length);
//printf("ly %d %d\n", length, y);
            if ((y >> (length - 1)) == 0) /* if MSB not set it is negative*/
                y = - (y ^ ((1 << length) - 1));
        }
        skip_bits1(&s->gb); /* marker bit */
//printf("%d %d %d %d\n", x, y, i, s->sprite_warping_accuracy);
//if(i>0 && (x!=0 || y!=0)) printf("AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA\n");
//x=y=0;
        d[i][0]= x;
        d[i][1]= y;
    }

    while((1<<alpha)<w) alpha++;
    while((1<<beta )<h) beta++; // there seems to be a typo in the mpeg4 std for the definition of w' and h'
    w2= 1<<alpha;
    h2= 1<<beta;

// Note, the 4th point isnt used for GMC
    if(s->divx_version==500 && s->divx_build==413){
        sprite_ref[0][0]= a*vop_ref[0][0] + d[0][0];
        sprite_ref[0][1]= a*vop_ref[0][1] + d[0][1];
        sprite_ref[1][0]= a*vop_ref[1][0] + d[0][0] + d[1][0];
        sprite_ref[1][1]= a*vop_ref[1][1] + d[0][1] + d[1][1];
        sprite_ref[2][0]= a*vop_ref[2][0] + d[0][0] + d[2][0];
        sprite_ref[2][1]= a*vop_ref[2][1] + d[0][1] + d[2][1];
    } else {
        sprite_ref[0][0]= (a>>1)*(2*vop_ref[0][0] + d[0][0]);
        sprite_ref[0][1]= (a>>1)*(2*vop_ref[0][1] + d[0][1]);
        sprite_ref[1][0]= (a>>1)*(2*vop_ref[1][0] + d[0][0] + d[1][0]);
        sprite_ref[1][1]= (a>>1)*(2*vop_ref[1][1] + d[0][1] + d[1][1]);
        sprite_ref[2][0]= (a>>1)*(2*vop_ref[2][0] + d[0][0] + d[2][0]);
        sprite_ref[2][1]= (a>>1)*(2*vop_ref[2][1] + d[0][1] + d[2][1]);
    }
/*    sprite_ref[3][0]= (a>>1)*(2*vop_ref[3][0] + d[0][0] + d[1][0] + d[2][0] + d[3][0]);
    sprite_ref[3][1]= (a>>1)*(2*vop_ref[3][1] + d[0][1] + d[1][1] + d[2][1] + d[3][1]); */
    
// this is mostly identical to the mpeg4 std (and is totally unreadable because of that ...)
// perhaps it should be reordered to be more readable ...
// the idea behind this virtual_ref mess is to be able to use shifts later per pixel instead of divides
// so the distance between points is converted from w&h based to w2&h2 based which are of the 2^x form
    virtual_ref[0][0]= 16*(vop_ref[0][0] + w2) 
        + RDIV(((w - w2)*(r*sprite_ref[0][0] - 16*vop_ref[0][0]) + w2*(r*sprite_ref[1][0] - 16*vop_ref[1][0])),w);
    virtual_ref[0][1]= 16*vop_ref[0][1] 
        + RDIV(((w - w2)*(r*sprite_ref[0][1] - 16*vop_ref[0][1]) + w2*(r*sprite_ref[1][1] - 16*vop_ref[1][1])),w);
    virtual_ref[1][0]= 16*vop_ref[0][0] 
        + RDIV(((h - h2)*(r*sprite_ref[0][0] - 16*vop_ref[0][0]) + h2*(r*sprite_ref[2][0] - 16*vop_ref[2][0])),h);
    virtual_ref[1][1]= 16*(vop_ref[0][1] + h2) 
        + RDIV(((h - h2)*(r*sprite_ref[0][1] - 16*vop_ref[0][1]) + h2*(r*sprite_ref[2][1] - 16*vop_ref[2][1])),h);

    switch(s->num_sprite_warping_points)
    {
        case 0:
            s->sprite_offset[0][0]= 0;
            s->sprite_offset[0][1]= 0;
            s->sprite_offset[1][0]= 0;
            s->sprite_offset[1][1]= 0;
            s->sprite_delta[0][0][0]= a;
            s->sprite_delta[0][0][1]= 0;
            s->sprite_delta[0][1][0]= 0;
            s->sprite_delta[0][1][1]= a;
            s->sprite_delta[1][0][0]= a;
            s->sprite_delta[1][0][1]= 0;
            s->sprite_delta[1][1][0]= 0;
            s->sprite_delta[1][1][1]= a;
            s->sprite_shift[0][0]= 0;
            s->sprite_shift[0][1]= 0;
            s->sprite_shift[1][0]= 0;
            s->sprite_shift[1][1]= 0;
            break;
        case 1: //GMC only
            s->sprite_offset[0][0]= sprite_ref[0][0] - a*vop_ref[0][0];
            s->sprite_offset[0][1]= sprite_ref[0][1] - a*vop_ref[0][1];
            s->sprite_offset[1][0]= ((sprite_ref[0][0]>>1)|(sprite_ref[0][0]&1)) - a*(vop_ref[0][0]/2);
            s->sprite_offset[1][1]= ((sprite_ref[0][1]>>1)|(sprite_ref[0][1]&1)) - a*(vop_ref[0][1]/2);
            s->sprite_delta[0][0][0]= a;
            s->sprite_delta[0][0][1]= 0;
            s->sprite_delta[0][1][0]= 0;
            s->sprite_delta[0][1][1]= a;
            s->sprite_delta[1][0][0]= a;
            s->sprite_delta[1][0][1]= 0;
            s->sprite_delta[1][1][0]= 0;
            s->sprite_delta[1][1][1]= a;
            s->sprite_shift[0][0]= 0;
            s->sprite_shift[0][1]= 0;
            s->sprite_shift[1][0]= 0;
            s->sprite_shift[1][1]= 0;
            break;
        case 2:
        case 3: //FIXME
            s->sprite_offset[0][0]= (sprite_ref[0][0]<<(alpha+rho))
                                                  + ((-r*sprite_ref[0][0] + virtual_ref[0][0])*(-vop_ref[0][0])
                                                    +( r*sprite_ref[0][1] - virtual_ref[0][1])*(-vop_ref[0][1]));
            s->sprite_offset[0][1]= (sprite_ref[0][1]<<(alpha+rho))
                                                  + ((-r*sprite_ref[0][1] + virtual_ref[0][1])*(-vop_ref[0][0])
                                                    +(-r*sprite_ref[0][0] + virtual_ref[0][0])*(-vop_ref[0][1]));
            s->sprite_offset[1][0]= ((-r*sprite_ref[0][0] + virtual_ref[0][0])*(-2*vop_ref[0][0] + 1)
                                 +( r*sprite_ref[0][1] - virtual_ref[0][1])*(-2*vop_ref[0][1] + 1)
                                 +2*w2*r*sprite_ref[0][0] - 16*w2);
            s->sprite_offset[1][1]= ((-r*sprite_ref[0][1] + virtual_ref[0][1])*(-2*vop_ref[0][0] + 1) 
                                 +(-r*sprite_ref[0][0] + virtual_ref[0][0])*(-2*vop_ref[0][1] + 1)
                                 +2*w2*r*sprite_ref[0][1] - 16*w2);
            s->sprite_delta[0][0][0]=   (-r*sprite_ref[0][0] + virtual_ref[0][0]);
            s->sprite_delta[0][0][1]=   ( r*sprite_ref[0][1] - virtual_ref[0][1]);
            s->sprite_delta[0][1][0]=   (-r*sprite_ref[0][1] + virtual_ref[0][1]);
            s->sprite_delta[0][1][1]=   (-r*sprite_ref[0][0] + virtual_ref[0][0]);
            s->sprite_delta[1][0][0]= 4*(-r*sprite_ref[0][0] + virtual_ref[0][0]);
            s->sprite_delta[1][0][1]= 4*( r*sprite_ref[0][1] - virtual_ref[0][1]);
            s->sprite_delta[1][1][0]= 4*(-r*sprite_ref[0][1] + virtual_ref[0][1]);
            s->sprite_delta[1][1][1]= 4*(-r*sprite_ref[0][0] + virtual_ref[0][0]);
            s->sprite_shift[0][0]= alpha+rho;
            s->sprite_shift[0][1]= alpha+rho;
            s->sprite_shift[1][0]= alpha+rho+2;
            s->sprite_shift[1][1]= alpha+rho+2;
            break;
//        case 3:
            break;
    }
/*printf("%d %d\n", s->sprite_delta[0][0][0], a<<s->sprite_shift[0][0]);
printf("%d %d\n", s->sprite_delta[0][0][1], 0);
printf("%d %d\n", s->sprite_delta[0][1][0], 0);
printf("%d %d\n", s->sprite_delta[0][1][1], a<<s->sprite_shift[0][1]);
printf("%d %d\n", s->sprite_delta[1][0][0], a<<s->sprite_shift[1][0]);
printf("%d %d\n", s->sprite_delta[1][0][1], 0);
printf("%d %d\n", s->sprite_delta[1][1][0], 0);
printf("%d %d\n", s->sprite_delta[1][1][1], a<<s->sprite_shift[1][1]);*/
    /* try to simplify the situation */ 
    if(   s->sprite_delta[0][0][0] == a<<s->sprite_shift[0][0]
       && s->sprite_delta[0][0][1] == 0
       && s->sprite_delta[0][1][0] == 0
       && s->sprite_delta[0][1][1] == a<<s->sprite_shift[0][1]
       && s->sprite_delta[1][0][0] == a<<s->sprite_shift[1][0]
       && s->sprite_delta[1][0][1] == 0
       && s->sprite_delta[1][1][0] == 0
       && s->sprite_delta[1][1][1] == a<<s->sprite_shift[1][1])
    {
        s->sprite_offset[0][0]>>=s->sprite_shift[0][0];
        s->sprite_offset[0][1]>>=s->sprite_shift[0][1];
        s->sprite_offset[1][0]>>=s->sprite_shift[1][0];
        s->sprite_offset[1][1]>>=s->sprite_shift[1][1];
        s->sprite_delta[0][0][0]= a;
        s->sprite_delta[0][0][1]= 0;
        s->sprite_delta[0][1][0]= 0;
        s->sprite_delta[0][1][1]= a;
        s->sprite_delta[1][0][0]= a;
        s->sprite_delta[1][0][1]= 0;
        s->sprite_delta[1][1][0]= 0;
        s->sprite_delta[1][1][1]= a;
        s->sprite_shift[0][0]= 0;
        s->sprite_shift[0][1]= 0;
        s->sprite_shift[1][0]= 0;
        s->sprite_shift[1][1]= 0;
        s->real_sprite_warping_points=1;
    }
    else
        s->real_sprite_warping_points= s->num_sprite_warping_points;

}

/* decode mpeg4 VOP header */
int mpeg4_decode_picture_header(MpegEncContext * s)
{
    int time_incr, startcode, state, v;

 redo:
    /* search next start code */
    align_get_bits(&s->gb);
    state = 0xff;
    for(;;) {
        v = get_bits(&s->gb, 8);
        if (state == 0x000001) {
            state = ((state << 8) | v) & 0xffffff;
            startcode = state;
            break;
        }
        state = ((state << 8) | v) & 0xffffff;
        /* XXX: really detect end of frame */
        if (state == 0){
            printf("illegal zero code found\n");
            return -1;
        }
    }
//printf("startcode %X %d\n", startcode, get_bits_count(&s->gb));
    if (startcode == 0x120) { // Video Object Layer
        int width, height, vo_ver_id;

        /* vol header */
        skip_bits(&s->gb, 1); /* random access */
        skip_bits(&s->gb, 8); /* vo_type */
        if (get_bits1(&s->gb) != 0) { /* is_ol_id */
            vo_ver_id = get_bits(&s->gb, 4); /* vo_ver_id */
            skip_bits(&s->gb, 3); /* vo_priority */
        } else {
            vo_ver_id = 1;
        }
        
        s->aspect_ratio_info= get_bits(&s->gb, 4);
	if(s->aspect_ratio_info == EXTENDET_PAR){
            skip_bits(&s->gb, 8); //par_width
            skip_bits(&s->gb, 8); // par_height
        }
        if(get_bits1(&s->gb)){ /* vol control parameter */
            printf("vol control parameter not supported\n");
            return -1;   
        }
        s->shape = get_bits(&s->gb, 2); /* vol shape */
        if(s->shape != RECT_SHAPE) printf("only rectangular vol supported\n");
        if(s->shape == GRAY_SHAPE && vo_ver_id != 1){
            printf("Gray shape not supported\n");
            skip_bits(&s->gb, 4);  //video_object_layer_shape_extension
        }

        skip_bits1(&s->gb);   /* marker */
        
        s->time_increment_resolution = get_bits(&s->gb, 16);
        s->time_increment_bits = av_log2(s->time_increment_resolution - 1) + 1;
        if (s->time_increment_bits < 1)
            s->time_increment_bits = 1;
        skip_bits1(&s->gb);   /* marker */

        if (get_bits1(&s->gb) != 0) {   /* fixed_vop_rate  */
            skip_bits(&s->gb, s->time_increment_bits);
        }

        if (s->shape != BIN_ONLY_SHAPE) {
            if (s->shape == RECT_SHAPE) {
                skip_bits1(&s->gb);   /* marker */
                width = get_bits(&s->gb, 13);
                skip_bits1(&s->gb);   /* marker */
                height = get_bits(&s->gb, 13);
                skip_bits1(&s->gb);   /* marker */
            }
            
            if(get_bits1(&s->gb)) printf("interlaced not supported\n");   /* interlaced */
            if(!get_bits1(&s->gb)) printf("OBMC not supported\n");   /* OBMC Disable */
            if (vo_ver_id == 1) {
                s->vol_sprite_usage = get_bits1(&s->gb); /* vol_sprite_usage */
            } else {
                s->vol_sprite_usage = get_bits(&s->gb, 2); /* vol_sprite_usage */
            }
            if(s->vol_sprite_usage==STATIC_SPRITE) printf("Static Sprites not supported\n");
            if(s->vol_sprite_usage==STATIC_SPRITE || s->vol_sprite_usage==GMC_SPRITE){
                if(s->vol_sprite_usage==STATIC_SPRITE){
                    s->sprite_width = get_bits(&s->gb, 13);
                    skip_bits1(&s->gb); /* marker */
                    s->sprite_height= get_bits(&s->gb, 13);
                    skip_bits1(&s->gb); /* marker */
                    s->sprite_left  = get_bits(&s->gb, 13);
                    skip_bits1(&s->gb); /* marker */
                    s->sprite_top   = get_bits(&s->gb, 13);
                    skip_bits1(&s->gb); /* marker */
                }
                s->num_sprite_warping_points= get_bits(&s->gb, 6);
                s->sprite_warping_accuracy = get_bits(&s->gb, 2);
                s->sprite_brightness_change= get_bits1(&s->gb);
                if(s->vol_sprite_usage==STATIC_SPRITE)
                    s->low_latency_sprite= get_bits1(&s->gb);            
            }
            // FIXME sadct disable bit if verid!=1 && shape not rect
            
            if (get_bits1(&s->gb) == 1) {   /* not_8_bit */
                s->quant_precision = get_bits(&s->gb, 4); /* quant_precision */
                if(get_bits(&s->gb, 4)!=8) printf("N-bit not supported\n"); /* bits_per_pixel */
            } else {
                s->quant_precision = 5;
            }
            
            // FIXME a bunch of grayscale shape things
            if(get_bits1(&s->gb)) printf("Quant-Type not supported\n");  /* vol_quant_type */ //FIXME
            if(vo_ver_id != 1)
                 s->quarter_sample= get_bits1(&s->gb);
            else s->quarter_sample=0;

            if(!get_bits1(&s->gb)) printf("Complexity estimation not supported\n");
#if 0
            if(get_bits1(&s->gb)) printf("resync disable\n");
#else
            skip_bits1(&s->gb);   /* resync_marker_disabled */
#endif
            s->data_partioning= get_bits1(&s->gb);
            if(s->data_partioning){
                printf("data partitioning not supported\n");
                skip_bits1(&s->gb); // reversible vlc
            }
            
            if(vo_ver_id != 1) {
                s->new_pred= get_bits1(&s->gb);
                if(s->new_pred){
                    printf("new pred not supported\n");
                    skip_bits(&s->gb, 2); /* requested upstream message type */
                    skip_bits1(&s->gb); /* newpred segment type */
                }
                s->reduced_res_vop= get_bits1(&s->gb);
                if(s->reduced_res_vop) printf("reduced resolution VOP not supported\n");
            }
            else{
                s->new_pred=0;
                s->reduced_res_vop= 0;
            }

            s->scalability= get_bits1(&s->gb);
            if (s->scalability) {
                printf("bad scalability!!!\n");
                return -1;
            }
        }
//printf("end Data %X %d\n", show_bits(&s->gb, 32), get_bits_count(&s->gb)&0x7);
        goto redo;
    } else if (startcode == 0x1b2) { //userdata
        char buf[256];
        int i;
        int e;
        int ver, build;

//printf("user Data %X\n", show_bits(&s->gb, 32));
        buf[0]= show_bits(&s->gb, 8);
        for(i=1; i<256; i++){
            buf[i]= show_bits(&s->gb, 16)&0xFF;
            if(buf[i]==0) break;
            skip_bits(&s->gb, 8);
        }
        buf[255]=0;
        e=sscanf(buf, "DivX%dBuild%d", &ver, &build);
        if(e==2){
            s->divx_version= ver;
            s->divx_build= build;
            if(s->picture_number==0){
                printf("This file was encoded with DivX%d Build%d\n", ver, build);
                if(ver==500 && build==413){ //most likely all version are indeed totally buggy but i dunno for sure ...
                    printf("WARNING: this version of DivX is not MPEG4 compatible, trying to workaround these bugs...\n");
                }else{
                    printf("hmm, i havnt seen that version of divx yet, lets assume they fixed these bugs ...\n"
                           "using mpeg4 decoder, if it fails contact the developers (of ffmpeg)\n");
                }
            }
        }
//printf("User Data: %s\n", buf);
        goto redo;
    } else if (startcode != 0x1b6) { //VOP
        goto redo;
    }

    s->pict_type = get_bits(&s->gb, 2) + 1;	/* pict type: I = 0 , P = 1 */
//printf("pic: %d\n", s->pict_type); 
    time_incr=0;
    while (get_bits1(&s->gb) != 0) 
        time_incr++;

    check_marker(&s->gb, "before time_increment");
    s->time_increment= get_bits(&s->gb, s->time_increment_bits);
    if(s->pict_type!=B_TYPE){
        s->time_base+= time_incr;
        s->last_non_b_time[1]= s->last_non_b_time[0];
        s->last_non_b_time[0]= s->time_base*s->time_increment_resolution + s->time_increment;
    }else{
        s->time= (s->last_non_b_time[1]/s->time_increment_resolution + time_incr)*s->time_increment_resolution;
        s->time+= s->time_increment;
    }

    if(check_marker(&s->gb, "before vop_coded")==0 && s->picture_number==0){
        printf("hmm, seems the headers arnt complete, trying to guess time_increment_bits\n");
        for(s->time_increment_bits++ ;s->time_increment_bits<16; s->time_increment_bits++){
            if(get_bits1(&s->gb)) break;
        }
        printf("my guess is %d bits ;)\n",s->time_increment_bits);
    }
    /* vop coded */
    if (get_bits1(&s->gb) != 1)
        goto redo;
//printf("time %d %d %d || %d %d %d\n", s->time_increment_bits, s->time_increment, s->time_base,
//s->time, s->last_non_b_time[0], s->last_non_b_time[1]);  
    if (s->shape != BIN_ONLY_SHAPE && ( s->pict_type == P_TYPE
                          || (s->pict_type == S_TYPE && s->vol_sprite_usage==GMC_SPRITE))) {
        /* rounding type for motion estimation */
	s->no_rounding = get_bits1(&s->gb);
    } else {
	s->no_rounding = 0;
    }
//FIXME reduced res stuff

     if (s->shape != RECT_SHAPE) {
         if (s->vol_sprite_usage != 1 || s->pict_type != I_TYPE) {
             int width, height, hor_spat_ref, ver_spat_ref;
 
             width = get_bits(&s->gb, 13);
             skip_bits1(&s->gb);   /* marker */
             height = get_bits(&s->gb, 13);
             skip_bits1(&s->gb);   /* marker */
             hor_spat_ref = get_bits(&s->gb, 13); /* hor_spat_ref */
             skip_bits1(&s->gb);   /* marker */
             ver_spat_ref = get_bits(&s->gb, 13); /* ver_spat_ref */
         }
         skip_bits1(&s->gb); /* change_CR_disable */
 
         if (get_bits1(&s->gb) != 0) {
             skip_bits(&s->gb, 8); /* constant_alpha_value */
         }
     }
//FIXME complexity estimation stuff
     
     if (s->shape != BIN_ONLY_SHAPE) {
         skip_bits(&s->gb, 3); /* intra dc VLC threshold */
         //FIXME interlaced specific bits
     }

     if(s->pict_type == S_TYPE && (s->vol_sprite_usage==STATIC_SPRITE || s->vol_sprite_usage==GMC_SPRITE)){
         if(s->num_sprite_warping_points){
             mpeg4_decode_sprite_trajectory(s);
         }
         if(s->sprite_brightness_change) printf("sprite_brightness_change not supported\n");
         if(s->vol_sprite_usage==STATIC_SPRITE) printf("static sprite not supported\n");
     }

     if (s->shape != BIN_ONLY_SHAPE) {
         /* note: we do not use quant_precision to avoid problem if no
            MPEG4 vol header as it is found on some old opendivx
            movies */
         s->qscale = get_bits(&s->gb, 5);
         if(s->qscale==0){
             printf("Error, header damaged or not MPEG4 header (qscale=0)\n");
             return -1; // makes no sense to continue, as there is nothing left from the image then
         }
  
         if (s->pict_type != I_TYPE) {
             s->f_code = get_bits(&s->gb, 3);	/* fcode_for */
             if(s->f_code==0){
                 printf("Error, header damaged or not MPEG4 header (f_code=0)\n");
                 return -1; // makes no sense to continue, as the MV decoding will break very quickly
             }
         }
         if (s->pict_type == B_TYPE) {
             s->b_code = get_bits(&s->gb, 3);
//printf("b-code %d\n", s->b_code);
         }
//printf("quant:%d fcode:%d\n", s->qscale, s->f_code);
         if(!s->scalability){
             if (s->shape!=RECT_SHAPE && s->pict_type!=I_TYPE) {
                 skip_bits1(&s->gb); // vop shape coding type
             }
         }
     }
     s->picture_number++; // better than pic number==0 allways ;)
     return 0;
}

/* don't understand why they choose a different header ! */
int intel_h263_decode_picture_header(MpegEncContext *s)
{
    int format;

    /* picture header */
    if (get_bits(&s->gb, 22) != 0x20)
        return -1;
    skip_bits(&s->gb, 8); /* picture timestamp */

    if (get_bits1(&s->gb) != 1)
        return -1;	/* marker */
    if (get_bits1(&s->gb) != 0)
        return -1;	/* h263 id */
    skip_bits1(&s->gb);	/* split screen off */
    skip_bits1(&s->gb);	/* camera  off */
    skip_bits1(&s->gb);	/* freeze picture release off */

    format = get_bits(&s->gb, 3);
    if (format != 7)
        return -1;

    s->h263_plus = 0;

    s->pict_type = I_TYPE + get_bits1(&s->gb);
    
    s->unrestricted_mv = get_bits1(&s->gb); 
    s->h263_long_vectors = s->unrestricted_mv;

    if (get_bits1(&s->gb) != 0)
        return -1;	/* SAC: off */
    if (get_bits1(&s->gb) != 0)
        return -1;	/* advanced prediction mode: off */
    if (get_bits1(&s->gb) != 0)
        return -1;	/* not PB frame */

    /* skip unknown header garbage */
    skip_bits(&s->gb, 41);

    s->qscale = get_bits(&s->gb, 5);
    skip_bits1(&s->gb);	/* Continuous Presence Multipoint mode: off */

    /* PEI */
    while (get_bits1(&s->gb) != 0) {
        skip_bits(&s->gb, 8);
    }
    s->f_code = 1;
    return 0;
}