]> git.sesse.net Git - ffmpeg/blobdiff - libavcodec/mpegvideo.c
convert only needed matrixes in convert_matrix() (mjpeg calls it for every frame)
[ffmpeg] / libavcodec / mpegvideo.c
index 180addb6f6953999308276e38fc7eb41c785ba55..1134774b6c66ebd6b4a719c183cce10ccc319ca5 100644 (file)
@@ -1,45 +1,50 @@
 /*
  * The simplest mpeg encoder (well, it was the simplest!)
- * Copyright (c) 2000,2001 Gerard Lantau.
+ * Copyright (c) 2000,2001 Fabrice Bellard.
  *
- * 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 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 program is distributed in the hope that it will be useful,
+ * 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 General Public License for more details.
+ * 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 General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ * 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., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ * 4MV & hq & b-frame encoding stuff by Michael Niedermayer <michaelni@gmx.at>
  */
-#include <stdlib.h>
-#include <stdio.h>
-#include <math.h>
-#include <string.h>
+#include <ctype.h>
 #include "avcodec.h"
 #include "dsputil.h"
 #include "mpegvideo.h"
+#include "simple_idct.h"
 
 #ifdef USE_FASTMEMCPY
 #include "fastmemcpy.h"
 #endif
 
+//#undef NDEBUG
+//#include <assert.h>
+
 static void encode_picture(MpegEncContext *s, int picture_number);
-static void rate_control_init(MpegEncContext *s);
-static int rate_estimate_qscale(MpegEncContext *s);
 static void dct_unquantize_mpeg1_c(MpegEncContext *s, 
                                    DCTELEM *block, int n, int qscale);
+static void dct_unquantize_mpeg2_c(MpegEncContext *s,
+                                   DCTELEM *block, int n, int qscale);
 static void dct_unquantize_h263_c(MpegEncContext *s, 
                                   DCTELEM *block, int n, int qscale);
 static void draw_edges_c(UINT8 *buf, int wrap, int width, int height, int w);
-static int dct_quantize_c(MpegEncContext *s, DCTELEM *block, int n, int qscale);
+static int dct_quantize_c(MpegEncContext *s, DCTELEM *block, int n, int qscale, int *overflow);
 
-int (*dct_quantize)(MpegEncContext *s, DCTELEM *block, int n, int qscale)= dct_quantize_c;
 void (*draw_edges)(UINT8 *buf, int wrap, int width, int height, int w)= draw_edges_c;
+static void emulated_edge_mc(MpegEncContext *s, UINT8 *src, int linesize, int block_w, int block_h, 
+                                    int src_x, int src_y, int w, int h);
 
 #define EDGE_WIDTH 16
 
@@ -48,6 +53,7 @@ void (*draw_edges)(UINT8 *buf, int wrap, int width, int height, int w)= draw_edg
 
 //#define DEBUG
 
+
 /* for jpeg fast DCT */
 #define CONST_BITS 14
 
@@ -67,39 +73,110 @@ static UINT8 h263_chroma_roundtab[16] = {
     0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2,
 };
 
-static UINT16 default_mv_penalty[MAX_FCODE][MAX_MV*2+1];
+static UINT16 default_mv_penalty[MAX_FCODE+1][MAX_MV*2+1];
 static UINT8 default_fcode_tab[MAX_MV*2+1];
 
 /* default motion estimation */
-int motion_estimation_method = ME_LOG;
+int motion_estimation_method = ME_EPZS;
 
-extern UINT8 zigzag_end[64];
+static void convert_matrix(MpegEncContext *s, int (*qmat)[64], uint16_t (*qmat16)[64], uint16_t (*qmat16_bias)[64],
+                           const UINT16 *quant_matrix, int bias, int qmin, int qmax)
+{
+    int qscale;
 
-static void convert_matrix(int *qmat, UINT16 *qmat16, const UINT16 *quant_matrix, int qscale)
+    for(qscale=qmin; qscale<=qmax; qscale++){
+        int i;
+        if (s->fdct == ff_jpeg_fdct_islow) {
+            for(i=0;i<64;i++) {
+                const int j= s->idct_permutation[i];
+                /* 16 <= qscale * quant_matrix[i] <= 7905 */
+                /* 19952         <= aanscales[i] * qscale * quant_matrix[i]           <= 249205026 */
+                /* (1<<36)/19952 >= (1<<36)/(aanscales[i] * qscale * quant_matrix[i]) >= (1<<36)/249205026 */
+                /* 3444240       >= (1<<36)/(aanscales[i] * qscale * quant_matrix[i]) >= 275 */
+                
+                qmat[qscale][j] = (int)((UINT64_C(1) << QMAT_SHIFT) / 
+                                (qscale * quant_matrix[j]));
+            }
+        } else if (s->fdct == fdct_ifast) {
+            for(i=0;i<64;i++) {
+                const int j= s->idct_permutation[i];
+                /* 16 <= qscale * quant_matrix[i] <= 7905 */
+                /* 19952         <= aanscales[i] * qscale * quant_matrix[i]           <= 249205026 */
+                /* (1<<36)/19952 >= (1<<36)/(aanscales[i] * qscale * quant_matrix[i]) >= (1<<36)/249205026 */
+                /* 3444240       >= (1<<36)/(aanscales[i] * qscale * quant_matrix[i]) >= 275 */
+                
+                qmat[qscale][j] = (int)((UINT64_C(1) << (QMAT_SHIFT + 14)) / 
+                                (aanscales[i] * qscale * quant_matrix[j]));
+            }
+        } else {
+            for(i=0;i<64;i++) {
+                const int j= s->idct_permutation[i];
+                /* We can safely suppose that 16 <= quant_matrix[i] <= 255
+                   So 16           <= qscale * quant_matrix[i]             <= 7905
+                   so (1<<19) / 16 >= (1<<19) / (qscale * quant_matrix[i]) >= (1<<19) / 7905
+                   so 32768        >= (1<<19) / (qscale * quant_matrix[i]) >= 67
+                */
+                qmat  [qscale][i] = (1 << QMAT_SHIFT_MMX) / (qscale * quant_matrix[i]);
+                qmat16[qscale][i] = (1 << QMAT_SHIFT_MMX) / (qscale * quant_matrix[j]);
+
+                if(qmat16[qscale][i]==0 || qmat16[qscale][i]==128*256) qmat16[qscale][i]=128*256-1;
+                qmat16_bias[qscale][i]= ROUNDED_DIV(bias<<(16-QUANT_BIAS_SHIFT), qmat16[qscale][i]);
+            }
+        }
+    }
+}
+// move into common.c perhaps 
+#define CHECKED_ALLOCZ(p, size)\
+{\
+    p= av_mallocz(size);\
+    if(p==NULL){\
+        perror("malloc");\
+        goto fail;\
+    }\
+}
+/*
+static void build_end(void)
 {
+    int lastIndex;
+    int lastIndexAfterPerm=0;
+    for(lastIndex=0; lastIndex<64; lastIndex++)
+    {
+        if(ff_zigzag_direct[lastIndex] > lastIndexAfterPerm) 
+            lastIndexAfterPerm= ff_zigzag_direct[lastIndex];
+        zigzag_end[lastIndex]= lastIndexAfterPerm + 1;
+    }
+}
+*/
+void ff_init_scantable(MpegEncContext *s, ScanTable *st, const UINT8 *src_scantable){
     int i;
+    int end;
 
-    if (av_fdct == jpeg_fdct_ifast) {
-        for(i=0;i<64;i++) {
-            /* 16 <= qscale * quant_matrix[i] <= 7905 */
-            /* 19952         <= aanscales[i] * qscale * quant_matrix[i]           <= 249205026 */
-            /* (1<<36)/19952 >= (1<<36)/(aanscales[i] * qscale * quant_matrix[i]) >= (1<<36)/249205026 */
-            /* 3444240       >= (1<<36)/(aanscales[i] * qscale * quant_matrix[i]) >= 275 */
-            
-            qmat[block_permute_op(i)] = (int)((UINT64_C(1) << (QMAT_SHIFT + 11)) / 
-                            (aanscales[i] * qscale * quant_matrix[block_permute_op(i)]));
-        }
-    } else {
-        for(i=0;i<64;i++) {
-            /* We can safely suppose that 16 <= quant_matrix[i] <= 255
-               So 16           <= qscale * quant_matrix[i]             <= 7905
-               so (1<<19) / 16 >= (1<<19) / (qscale * quant_matrix[i]) >= (1<<19) / 7905
-               so 32768        >= (1<<19) / (qscale * quant_matrix[i]) >= 67
-            */
-            qmat[i]   = (1 << QMAT_SHIFT_MMX) / (qscale * quant_matrix[i]);
-            qmat16[i] = (1 << QMAT_SHIFT_MMX) / (qscale * quant_matrix[block_permute_op(i)]);
-        }
+    for(i=0; i<64; i++){
+        int j;
+        j = src_scantable[i];
+        st->permutated[i] = s->idct_permutation[j];
     }
+    
+    end=-1;
+    for(i=0; i<64; i++){
+        int j;
+        j = st->permutated[i];
+        if(j>end) end=j;
+        st->raster_end[i]= end;
+    }
+}
+
+/* XXX: those functions should be suppressed ASAP when all IDCTs are
+   converted */
+static void ff_jref_idct_put(UINT8 *dest, int line_size, DCTELEM *block)
+{
+    j_rev_dct (block);
+    put_pixels_clamped(block, dest, line_size);
+}
+static void ff_jref_idct_add(UINT8 *dest, int line_size, DCTELEM *block)
+{
+    j_rev_dct (block);
+    add_pixels_clamped(block, dest, line_size);
 }
 
 /* init common structure for both encoder and decoder */
@@ -108,80 +185,164 @@ int MPV_common_init(MpegEncContext *s)
     int c_size, i;
     UINT8 *pict;
 
-    if (s->out_format == FMT_H263) 
-        s->dct_unquantize = dct_unquantize_h263_c;
+    s->dct_unquantize_h263 = dct_unquantize_h263_c;
+    s->dct_unquantize_mpeg1 = dct_unquantize_mpeg1_c;
+    s->dct_unquantize_mpeg2 = dct_unquantize_mpeg2_c;
+    s->dct_quantize= dct_quantize_c;
+
+    if(s->avctx->dct_algo==FF_DCT_FASTINT)
+        s->fdct = fdct_ifast;
     else
-        s->dct_unquantize = dct_unquantize_mpeg1_c;
+        s->fdct = ff_jpeg_fdct_islow; //slow/accurate/default
+
+    if(s->avctx->idct_algo==FF_IDCT_INT){
+        s->idct_put= ff_jref_idct_put;
+        s->idct_add= ff_jref_idct_add;
+        for(i=0; i<64; i++)
+            s->idct_permutation[i]= (i & 0x38) | ((i & 6) >> 1) | ((i & 1) << 2);
+    }else{ //accurate/default
+        s->idct_put= simple_idct_put;
+        s->idct_add= simple_idct_add;
+        for(i=0; i<64; i++)
+            s->idct_permutation[i]= i;
+    }
         
 #ifdef HAVE_MMX
     MPV_common_init_mmx(s);
 #endif
+#ifdef ARCH_ALPHA
+    MPV_common_init_axp(s);
+#endif
+#ifdef HAVE_MLIB
+    MPV_common_init_mlib(s);
+#endif
+    
+
+    /* load & permutate scantables
+       note: only wmv uses differnt ones 
+    */
+    ff_init_scantable(s, &s->inter_scantable  , ff_zigzag_direct);
+    ff_init_scantable(s, &s->intra_scantable  , ff_zigzag_direct);
+    ff_init_scantable(s, &s->intra_h_scantable, ff_alternate_horizontal_scan);
+    ff_init_scantable(s, &s->intra_v_scantable, ff_alternate_vertical_scan);
+
     s->mb_width = (s->width + 15) / 16;
     s->mb_height = (s->height + 15) / 16;
+    
+    /* set default edge pos, will be overriden in decode_header if needed */
+    s->h_edge_pos= s->mb_width*16;
+    s->v_edge_pos= s->mb_height*16;
+    
+    /* convert fourcc to upper case */
+    s->avctx->fourcc=   toupper( s->avctx->fourcc     &0xFF)          
+                     + (toupper((s->avctx->fourcc>>8 )&0xFF)<<8 )
+                     + (toupper((s->avctx->fourcc>>16)&0xFF)<<16) 
+                     + (toupper((s->avctx->fourcc>>24)&0xFF)<<24);
+
     s->mb_num = s->mb_width * s->mb_height;
-    s->linesize = s->mb_width * 16 + 2 * EDGE_WIDTH;
+    if(!(s->flags&CODEC_FLAG_DR1)){
+      s->linesize   = s->mb_width * 16 + 2 * EDGE_WIDTH;
+      s->uvlinesize = s->mb_width * 8  +     EDGE_WIDTH;
 
-    for(i=0;i<3;i++) {
+      for(i=0;i<3;i++) {
         int w, h, shift, pict_start;
 
         w = s->linesize;
         h = s->mb_height * 16 + 2 * EDGE_WIDTH;
         shift = (i == 0) ? 0 : 1;
-        c_size = (w >> shift) * (h >> shift);
-        pict_start = (w >> shift) * (EDGE_WIDTH >> shift) + (EDGE_WIDTH >> shift);
+        c_size = (s->linesize>>shift) * (h >> shift);
+        pict_start = (s->linesize>>shift) * (EDGE_WIDTH >> shift) + (EDGE_WIDTH >> shift);
 
-        pict = av_mallocz(c_size);
-        if (pict == NULL)
-            goto fail;
+        CHECKED_ALLOCZ(pict, c_size)
         s->last_picture_base[i] = pict;
         s->last_picture[i] = pict + pict_start;
+        if(i>0) memset(s->last_picture_base[i], 128, c_size);
     
-        pict = av_mallocz(c_size);
-        if (pict == NULL)
-            goto fail;
+        CHECKED_ALLOCZ(pict, c_size)
         s->next_picture_base[i] = pict;
         s->next_picture[i] = pict + pict_start;
-
-        if (s->has_b_frames) {
-            pict = av_mallocz(c_size);
-            if (pict == NULL) 
-                goto fail;
+        if(i>0) memset(s->next_picture_base[i], 128, c_size);
+        
+        if (s->has_b_frames || s->codec_id==CODEC_ID_MPEG4) {
+        /* Note the MPEG4 stuff is here cuz of buggy encoders which dont set the low_delay flag but 
+           do low-delay encoding, so we cant allways distinguish b-frame containing streams from low_delay streams */
+            CHECKED_ALLOCZ(pict, c_size)
             s->aux_picture_base[i] = pict;
             s->aux_picture[i] = pict + pict_start;
+            if(i>0) memset(s->aux_picture_base[i], 128, c_size);
         }
+      }
+      s->ip_buffer_count= 2;
     }
     
+    CHECKED_ALLOCZ(s->edge_emu_buffer, (s->width+64)*2*17*2); //(width + edge + align)*interlaced*MBsize*tolerance
+    
     if (s->encoding) {
-        /* Allocate MB type table */
-        s->mb_type = av_mallocz(s->mb_num * sizeof(char));
-        if (s->mb_type == NULL) {
-            perror("malloc");
-            goto fail;
-        }
+        int j;
+        int mv_table_size= (s->mb_width+2)*(s->mb_height+2);
+        
+        CHECKED_ALLOCZ(s->mb_var   , s->mb_num * sizeof(INT16))
+        CHECKED_ALLOCZ(s->mc_mb_var, s->mb_num * sizeof(INT16))
+        CHECKED_ALLOCZ(s->mb_mean  , s->mb_num * sizeof(INT8))
+
+        /* Allocate MV tables */
+        CHECKED_ALLOCZ(s->p_mv_table            , mv_table_size * 2 * sizeof(INT16))
+        CHECKED_ALLOCZ(s->b_forw_mv_table       , mv_table_size * 2 * sizeof(INT16))
+        CHECKED_ALLOCZ(s->b_back_mv_table       , mv_table_size * 2 * sizeof(INT16))
+        CHECKED_ALLOCZ(s->b_bidir_forw_mv_table , mv_table_size * 2 * sizeof(INT16))
+        CHECKED_ALLOCZ(s->b_bidir_back_mv_table , mv_table_size * 2 * sizeof(INT16))
+        CHECKED_ALLOCZ(s->b_direct_forw_mv_table, mv_table_size * 2 * sizeof(INT16))
+        CHECKED_ALLOCZ(s->b_direct_back_mv_table, mv_table_size * 2 * sizeof(INT16))
+        CHECKED_ALLOCZ(s->b_direct_mv_table     , mv_table_size * 2 * sizeof(INT16))
+
+        CHECKED_ALLOCZ(s->me_scratchpad,  s->linesize*16*3*sizeof(uint8_t))
         
-        s->mb_var = av_mallocz(s->mb_num * sizeof(INT16));
-        if (s->mb_var == NULL) {
-            perror("malloc");
-            goto fail;
+        CHECKED_ALLOCZ(s->me_map      , ME_MAP_SIZE*sizeof(uint32_t))
+        CHECKED_ALLOCZ(s->me_score_map, ME_MAP_SIZE*sizeof(uint16_t))
+
+        if(s->max_b_frames){
+            for(j=0; j<REORDER_BUFFER_SIZE; j++){
+                int i;
+                for(i=0;i<3;i++) {
+                    int w, h, shift;
+
+                    w = s->linesize;
+                    h = s->mb_height * 16;
+                    shift = (i == 0) ? 0 : 1;
+                    c_size = (w >> shift) * (h >> shift);
+
+                    CHECKED_ALLOCZ(pict, c_size);
+                    s->picture_buffer[j][i] = pict;
+                }
+            }
         }
-        /* Allocate MV table */
-        /* By now we just have one MV per MB */
-        s->mv_table[0] = av_mallocz(s->mb_num * sizeof(INT16));
-        s->mv_table[1] = av_mallocz(s->mb_num * sizeof(INT16));
-        if (s->mv_table[1] == NULL || s->mv_table[0] == NULL) {
-            perror("malloc");
-            goto fail;
+
+        if(s->codec_id==CODEC_ID_MPEG4){
+            CHECKED_ALLOCZ(s->tex_pb_buffer, PB_BUFFER_SIZE);
+            CHECKED_ALLOCZ(   s->pb2_buffer, PB_BUFFER_SIZE);
+        }
+        
+        if(s->msmpeg4_version){
+            CHECKED_ALLOCZ(s->ac_stats, 2*2*(MAX_LEVEL+1)*(MAX_RUN+1)*2*sizeof(int));
         }
+        CHECKED_ALLOCZ(s->avctx->stats_out, 256);
     }
     
-    if (s->out_format == FMT_H263) {
+    if (s->out_format == FMT_H263 || s->encoding) {
         int size;
+        /* Allocate MB type table */
+        CHECKED_ALLOCZ(s->mb_type  , s->mb_num * sizeof(UINT8))
+
         /* MV prediction */
         size = (2 * s->mb_width + 2) * (2 * s->mb_height + 2);
-        s->motion_val = malloc(size * 2 * sizeof(INT16));
-        if (s->motion_val == NULL)
-            goto fail;
-        memset(s->motion_val, 0, size * 2 * sizeof(INT16));
+        CHECKED_ALLOCZ(s->motion_val, size * 2 * sizeof(INT16));
+    }
+
+    if(s->codec_id==CODEC_ID_MPEG4){
+        /* 4mv and interlaced direct mode decoding tables */
+        CHECKED_ALLOCZ(s->co_located_type_table, s->mb_num * sizeof(UINT8))
+        CHECKED_ALLOCZ(s->field_mv_table, s->mb_num*2*2 * sizeof(INT16))
+        CHECKED_ALLOCZ(s->field_select_table, s->mb_num*2* sizeof(INT8))
     }
 
     if (s->h263_pred || s->h263_plus) {
@@ -192,41 +353,40 @@ int MPV_common_init(MpegEncContext *s)
         y_size = (2 * s->mb_width + 2) * (2 * s->mb_height + 2);
         c_size = (s->mb_width + 2) * (s->mb_height + 2);
         size = y_size + 2 * c_size;
-        s->dc_val[0] = malloc(size * sizeof(INT16));
-        if (s->dc_val[0] == NULL)
-            goto fail;
+        CHECKED_ALLOCZ(s->dc_val[0], size * sizeof(INT16));
         s->dc_val[1] = s->dc_val[0] + y_size;
         s->dc_val[2] = s->dc_val[1] + c_size;
         for(i=0;i<size;i++)
             s->dc_val[0][i] = 1024;
 
         /* ac values */
-        s->ac_val[0] = av_mallocz(size * sizeof(INT16) * 16);
-        if (s->ac_val[0] == NULL)
-            goto fail;
+        CHECKED_ALLOCZ(s->ac_val[0], size * sizeof(INT16) * 16);
         s->ac_val[1] = s->ac_val[0] + y_size;
         s->ac_val[2] = s->ac_val[1] + c_size;
         
         /* cbp values */
-        s->coded_block = av_mallocz(y_size);
-        if (!s->coded_block)
-            goto fail;
-
-        /* which mb is a intra block */
-        s->mbintra_table = av_mallocz(s->mb_num);
-        if (!s->mbintra_table)
-            goto fail;
-        memset(s->mbintra_table, 1, s->mb_num);
+        CHECKED_ALLOCZ(s->coded_block, y_size);
+        
+        /* divx501 bitstream reorder buffer */
+        CHECKED_ALLOCZ(s->bitstream_buffer, BITSTREAM_BUFFER_SIZE);
+        
+        /* cbp, ac_pred, pred_dir */
+        CHECKED_ALLOCZ(s->cbp_table  , s->mb_num * sizeof(UINT8))
+        CHECKED_ALLOCZ(s->pred_dir_table, s->mb_num * sizeof(UINT8))
     }
+    CHECKED_ALLOCZ(s->qscale_table  , s->mb_num * sizeof(UINT8))
+    
+    /* which mb is a intra block */
+    CHECKED_ALLOCZ(s->mbintra_table, s->mb_num);
+    memset(s->mbintra_table, 1, s->mb_num);
+    
     /* default structure is frame */
     s->picture_structure = PICT_FRAME;
-
+    
     /* init macroblock skip table */
-    if (!s->encoding) {
-        s->mbskip_table = av_mallocz(s->mb_num);
-        if (!s->mbskip_table)
-            goto fail;
-    }
+    CHECKED_ALLOCZ(s->mbskip_table, s->mb_num);
+    
+    s->block= s->blocks[0];
 
     s->context_initialized = 1;
     return 0;
@@ -235,39 +395,66 @@ int MPV_common_init(MpegEncContext *s)
     return -1;
 }
 
+
+//extern int sads;
+
 /* init common structure for both encoder and decoder */
 void MPV_common_end(MpegEncContext *s)
 {
     int i;
 
-    if (s->mb_type)
-        free(s->mb_type);
-    if (s->mb_var)
-        free(s->mb_var);
-    if (s->mv_table[0])
-        free(s->mv_table[0]);
-    if (s->mv_table[1])
-        free(s->mv_table[1]);
-    if (s->motion_val)
-        free(s->motion_val);
-    if (s->dc_val[0])
-        free(s->dc_val[0]);
-    if (s->ac_val[0])
-        free(s->ac_val[0]);
-    if (s->coded_block)
-        free(s->coded_block);
-    if (s->mbintra_table)
-        free(s->mbintra_table);
-
-    if (s->mbskip_table)
-        free(s->mbskip_table);
+    av_freep(&s->mb_type);
+    av_freep(&s->mb_var);
+    av_freep(&s->mc_mb_var);
+    av_freep(&s->mb_mean);
+    av_freep(&s->p_mv_table);
+    av_freep(&s->b_forw_mv_table);
+    av_freep(&s->b_back_mv_table);
+    av_freep(&s->b_bidir_forw_mv_table);
+    av_freep(&s->b_bidir_back_mv_table);
+    av_freep(&s->b_direct_forw_mv_table);
+    av_freep(&s->b_direct_back_mv_table);
+    av_freep(&s->b_direct_mv_table);
+    av_freep(&s->motion_val);
+    av_freep(&s->dc_val[0]);
+    av_freep(&s->ac_val[0]);
+    av_freep(&s->coded_block);
+    av_freep(&s->mbintra_table);
+    av_freep(&s->cbp_table);
+    av_freep(&s->pred_dir_table);
+    av_freep(&s->qscale_table);
+    av_freep(&s->me_scratchpad);
+    av_freep(&s->me_map);
+    av_freep(&s->me_score_map);
+    
+    av_freep(&s->mbskip_table);
+    av_freep(&s->bitstream_buffer);
+    av_freep(&s->tex_pb_buffer);
+    av_freep(&s->pb2_buffer);
+    av_freep(&s->edge_emu_buffer);
+    av_freep(&s->co_located_type_table);
+    av_freep(&s->field_mv_table);
+    av_freep(&s->field_select_table);
+    av_freep(&s->avctx->stats_out);
+    av_freep(&s->ac_stats);
+    
     for(i=0;i<3;i++) {
-        if (s->last_picture_base[i])
-           free(s->last_picture_base[i]);
-       if (s->next_picture_base[i])
-           free(s->next_picture_base[i]);
-        if (s->has_b_frames)
-            free(s->aux_picture_base[i]);
+        int j;
+        if(!(s->flags&CODEC_FLAG_DR1)){
+            av_freep(&s->last_picture_base[i]);
+            av_freep(&s->next_picture_base[i]);
+            av_freep(&s->aux_picture_base[i]);
+        }
+        s->last_picture_base[i]=
+        s->next_picture_base[i]=
+        s->aux_picture_base [i] = NULL;
+        s->last_picture[i]=
+        s->next_picture[i]=
+        s->aux_picture [i] = NULL;
+
+        for(j=0; j<REORDER_BUFFER_SIZE; j++){
+            av_freep(&s->picture_buffer[j][i]);
+        }
     }
     s->context_initialized = 0;
 }
@@ -278,11 +465,17 @@ int MPV_encode_init(AVCodecContext *avctx)
     MpegEncContext *s = avctx->priv_data;
     int i;
 
+    avctx->pix_fmt = PIX_FMT_YUV420P;
+
     s->bit_rate = avctx->bit_rate;
     s->bit_rate_tolerance = avctx->bit_rate_tolerance;
     s->frame_rate = avctx->frame_rate;
     s->width = avctx->width;
     s->height = avctx->height;
+    if(avctx->gop_size > 600){
+        fprintf(stderr, "Warning keyframe interval too large! reducing it ...\n");
+        avctx->gop_size=600;
+    }
     s->gop_size = avctx->gop_size;
     s->rtp_mode = avctx->rtp_mode;
     s->rtp_payload_size = avctx->rtp_payload_size;
@@ -294,33 +487,67 @@ int MPV_encode_init(AVCodecContext *avctx)
     s->qcompress= avctx->qcompress;
     s->qblur= avctx->qblur;
     s->avctx = avctx;
-    
+    s->aspect_ratio_info= avctx->aspect_ratio_info;
+    if (avctx->aspect_ratio_info == FF_ASPECT_EXTENDED)
+    {
+       s->aspected_width = avctx->aspected_width;
+       s->aspected_height = avctx->aspected_height;
+    }
+    s->flags= avctx->flags;
+    s->max_b_frames= avctx->max_b_frames;
+    s->b_frame_strategy= avctx->b_frame_strategy;
+    s->codec_id= avctx->codec->id;
+    s->luma_elim_threshold  = avctx->luma_elim_threshold;
+    s->chroma_elim_threshold= avctx->chroma_elim_threshold;
+    s->strict_std_compliance= avctx->strict_std_compliance;
+    s->data_partitioning= avctx->flags & CODEC_FLAG_PART;
+    s->mpeg_quant= avctx->mpeg_quant;
+
     if (s->gop_size <= 1) {
         s->intra_only = 1;
         s->gop_size = 12;
     } else {
         s->intra_only = 0;
     }
-    s->full_search = motion_estimation_method;
 
+    /* ME algorithm */
+    if (avctx->me_method == 0)
+        /* For compatibility */
+        s->me_method = motion_estimation_method;
+    else
+        s->me_method = avctx->me_method;
+
+    /* Fixed QSCALE */
     s->fixed_qscale = (avctx->flags & CODEC_FLAG_QSCALE);
     
+    s->adaptive_quant= (   s->avctx->lumi_masking
+                        || s->avctx->dark_masking
+                        || s->avctx->temporal_cplx_masking 
+                        || s->avctx->spatial_cplx_masking
+                        || s->avctx->p_masking)
+                       && !s->fixed_qscale;
+    
+    s->progressive_sequence= !(avctx->flags & CODEC_FLAG_INTERLACED_DCT);
+
     switch(avctx->codec->id) {
     case CODEC_ID_MPEG1VIDEO:
         s->out_format = FMT_MPEG1;
+        avctx->delay=0; //FIXME not sure, should check the spec
         break;
     case CODEC_ID_MJPEG:
         s->out_format = FMT_MJPEG;
         s->intra_only = 1; /* force intra only for jpeg */
         s->mjpeg_write_tables = 1; /* write all tables */
+       s->mjpeg_data_only_frames = 0; /* write all the needed headers */
         s->mjpeg_vsample[0] = 2; /* set up default sampling factors */
         s->mjpeg_vsample[1] = 1; /* the only currently supported values */
         s->mjpeg_vsample[2] = 1; 
-        s->mjpeg_hsample[0] = 2; 
+        s->mjpeg_hsample[0] = 2;
         s->mjpeg_hsample[1] = 1; 
         s->mjpeg_hsample[2] = 1; 
         if (mjpeg_init(s) < 0)
             return -1;
+        avctx->delay=0;
         break;
     case CODEC_ID_H263:
         if (h263_get_picture_format(s->width, s->height) == 7) {
@@ -328,6 +555,7 @@ int MPV_encode_init(AVCodecContext *avctx)
             return -1;
         }
         s->out_format = FMT_H263;
+        avctx->delay=0;
         break;
     case CODEC_ID_H263P:
         s->out_format = FMT_H263;
@@ -335,36 +563,76 @@ int MPV_encode_init(AVCodecContext *avctx)
         s->rtp_payload_size = 1200; 
         s->h263_plus = 1;
         s->unrestricted_mv = 1;
+        s->h263_aic = 1;
         
         /* These are just to be sure */
         s->umvplus = 0;
         s->umvplus_dec = 0;
+        avctx->delay=0;
         break;
     case CODEC_ID_RV10:
         s->out_format = FMT_H263;
         s->h263_rv10 = 1;
+        avctx->delay=0;
         break;
     case CODEC_ID_MPEG4:
         s->out_format = FMT_H263;
         s->h263_pred = 1;
         s->unrestricted_mv = 1;
+        s->has_b_frames= s->max_b_frames ? 1 : 0;
+        s->low_delay=0;
+        avctx->delay= s->low_delay ? 0 : (s->max_b_frames + 1); 
+        break;
+    case CODEC_ID_MSMPEG4V1:
+        s->out_format = FMT_H263;
+        s->h263_msmpeg4 = 1;
+        s->h263_pred = 1;
+        s->unrestricted_mv = 1;
+        s->msmpeg4_version= 1;
+        avctx->delay=0;
+        break;
+    case CODEC_ID_MSMPEG4V2:
+        s->out_format = FMT_H263;
+        s->h263_msmpeg4 = 1;
+        s->h263_pred = 1;
+        s->unrestricted_mv = 1;
+        s->msmpeg4_version= 2;
+        avctx->delay=0;
+        break;
+    case CODEC_ID_MSMPEG4V3:
+        s->out_format = FMT_H263;
+        s->h263_msmpeg4 = 1;
+        s->h263_pred = 1;
+        s->unrestricted_mv = 1;
+        s->msmpeg4_version= 3;
+        avctx->delay=0;
+        break;
+    case CODEC_ID_WMV1:
+        s->out_format = FMT_H263;
+        s->h263_msmpeg4 = 1;
+        s->h263_pred = 1;
+        s->unrestricted_mv = 1;
+        s->msmpeg4_version= 4;
+        avctx->delay=0;
         break;
-    case CODEC_ID_MSMPEG4:
+    case CODEC_ID_WMV2:
         s->out_format = FMT_H263;
         s->h263_msmpeg4 = 1;
         s->h263_pred = 1;
         s->unrestricted_mv = 1;
+        s->msmpeg4_version= 5;
+        avctx->delay=0;
         break;
     default:
         return -1;
     }
-
+    
     { /* set up some save defaults, some codecs might override them later */
         static int done=0;
         if(!done){
             int i;
             done=1;
-            memset(default_mv_penalty, 0, sizeof(UINT16)*MAX_FCODE*(2*MAX_MV+1));
+            memset(default_mv_penalty, 0, sizeof(UINT16)*(MAX_FCODE+1)*(2*MAX_MV+1));
             memset(default_fcode_tab , 0, sizeof(UINT8)*(2*MAX_MV+1));
 
             for(i=-16; i<16; i++){
@@ -374,9 +642,11 @@ int MPV_encode_init(AVCodecContext *avctx)
     }
     s->mv_penalty= default_mv_penalty;
     s->fcode_tab= default_fcode_tab;
-
-    if (s->out_format == FMT_H263)
-        h263_encode_init(s);
+    s->y_dc_scale_table=
+    s->c_dc_scale_table= ff_mpeg1_dc_scale_table;
+    /* dont use mv_penalty table for crap MV as it would be confused */
+    if (s->me_method < ME_EPZS) s->mv_penalty = default_mv_penalty;
 
     s->encoding = 1;
 
@@ -384,20 +654,46 @@ int MPV_encode_init(AVCodecContext *avctx)
     if (MPV_common_init(s) < 0)
         return -1;
     
+    if (s->out_format == FMT_H263)
+        h263_encode_init(s);
+    else if (s->out_format == FMT_MPEG1)
+        ff_mpeg1_encode_init(s);
+    if(s->msmpeg4_version)
+        ff_msmpeg4_encode_init(s);
+
     /* init default q matrix */
     for(i=0;i<64;i++) {
-        s->intra_matrix[i] = default_intra_matrix[i];
-        s->non_intra_matrix[i] = default_non_intra_matrix[i];
+        int j= s->idct_permutation[i];
+        if(s->codec_id==CODEC_ID_MPEG4 && s->mpeg_quant){
+            s->intra_matrix[j] = ff_mpeg4_default_intra_matrix[i];
+            s->inter_matrix[j] = ff_mpeg4_default_non_intra_matrix[i];
+        }else if(s->out_format == FMT_H263){
+            s->intra_matrix[j] =
+            s->inter_matrix[j] = ff_mpeg1_default_non_intra_matrix[i];
+        }else{ /* mpeg1 */
+            s->intra_matrix[j] = ff_mpeg1_default_intra_matrix[i];
+            s->inter_matrix[j] = ff_mpeg1_default_non_intra_matrix[i];
+        }
+    }
+
+    /* precompute matrix */
+    /* for mjpeg, we do include qscale in the matrix */
+    if (s->out_format != FMT_MJPEG) {
+        convert_matrix(s, s->q_intra_matrix, s->q_intra_matrix16, s->q_intra_matrix16_bias, 
+                       s->intra_matrix, s->intra_quant_bias, 1, 31);
+        convert_matrix(s, s->q_inter_matrix, s->q_inter_matrix16, s->q_inter_matrix16_bias, 
+                       s->inter_matrix, s->inter_quant_bias, 1, 31);
     }
 
-    /* rate control init */
-    rate_control_init(s);
+    if(ff_rate_control_init(s) < 0)
+        return -1;
 
     s->picture_number = 0;
     s->picture_in_gop_number = 0;
     s->fake_picture_number = 0;
     /* motion detector init */
     s->f_code = 1;
+    s->b_code = 1;
 
     return 0;
 }
@@ -409,6 +705,9 @@ int MPV_encode_end(AVCodecContext *avctx)
 #ifdef STATS
     print_stats();
 #endif
+
+    ff_rate_control_uninit(s);
+
     MPV_common_end(s);
     if (s->out_format == FMT_MJPEG)
         mjpeg_close(s);
@@ -417,6 +716,7 @@ int MPV_encode_end(AVCodecContext *avctx)
 }
 
 /* draw the edges of width 'w' of an image of size width, height */
+//FIXME check that this is ok for mpeg4 interlaced
 static void draw_edges_c(UINT8 *buf, int wrap, int width, int height, int w)
 {
     UINT8 *ptr, *last_line;
@@ -445,45 +745,171 @@ static void draw_edges_c(UINT8 *buf, int wrap, int width, int height, int w)
 }
 
 /* generic function for encode/decode called before a frame is coded/decoded */
-void MPV_frame_start(MpegEncContext *s)
+void MPV_frame_start(MpegEncContext *s, AVCodecContext *avctx)
 {
     int i;
     UINT8 *tmp;
 
     s->mb_skiped = 0;
+    s->decoding_error=0;
+    avctx->mbskip_table= s->mbskip_table;
+
+    if(avctx->flags&CODEC_FLAG_DR1){
+        avctx->get_buffer_callback(avctx, s->width, s->height, s->pict_type);
+
+        s->linesize  = avctx->dr_stride;
+        s->uvlinesize= avctx->dr_uvstride;
+        s->ip_buffer_count= avctx->dr_ip_buffer_count;
+    }
+    avctx->dr_ip_buffer_count= s->ip_buffer_count;
+    
     if (s->pict_type == B_TYPE) {
         for(i=0;i<3;i++) {
+            if(avctx->flags&CODEC_FLAG_DR1)
+                s->aux_picture[i]= avctx->dr_buffer[i];
+            
+            //FIXME the following should never be needed, the decoder should drop b frames if no reference is available
+            if(s->next_picture[i]==NULL)
+                s->next_picture[i]= s->aux_picture[i];
+            if(s->last_picture[i]==NULL)
+                s->last_picture[i]= s->next_picture[i];
+
             s->current_picture[i] = s->aux_picture[i];
         }
     } else {
-        s->last_non_b_pict_type= s->pict_type;
         for(i=0;i<3;i++) {
             /* swap next and last */
-            tmp = s->last_picture[i];
+            if(avctx->flags&CODEC_FLAG_DR1)
+                tmp= avctx->dr_buffer[i];
+            else
+                tmp = s->last_picture[i];
+
             s->last_picture[i] = s->next_picture[i];
             s->next_picture[i] = tmp;
             s->current_picture[i] = tmp;
+
+            if(s->last_picture[i]==NULL)
+                s->last_picture[i]= s->next_picture[i];
+
+            s->last_dr_opaque= s->next_dr_opaque;
+            s->next_dr_opaque= avctx->dr_opaque_frame;
+
+            if(s->has_b_frames && s->last_dr_opaque && s->codec_id!=CODEC_ID_SVQ1)
+                avctx->dr_opaque_frame= s->last_dr_opaque;
+            else
+                avctx->dr_opaque_frame= s->next_dr_opaque;
         }
     }
+    /* set dequantizer, we cant do it during init as it might change for mpeg4
+       and we cant do it in the header decode as init isnt called for mpeg4 there yet */
+    if(s->out_format == FMT_H263){
+        if(s->mpeg_quant)
+            s->dct_unquantize = s->dct_unquantize_mpeg2;
+        else
+            s->dct_unquantize = s->dct_unquantize_h263;
+    }else 
+        s->dct_unquantize = s->dct_unquantize_mpeg1;
 }
 
 /* generic function for encode/decode called after a frame has been coded/decoded */
 void MPV_frame_end(MpegEncContext *s)
 {
+    s->avctx->key_frame   = (s->pict_type == I_TYPE);
+    s->avctx->pict_type   = s->pict_type;
+
     /* draw edge for correct motion prediction if outside */
-    if (s->pict_type != B_TYPE && !s->intra_only) {
-      if(s->avctx==NULL || s->avctx->codec->id!=CODEC_ID_MPEG4 || s->divx_version==500){
-        draw_edges(s->current_picture[0], s->linesize, s->mb_width*16, s->mb_height*16, EDGE_WIDTH);
-        draw_edges(s->current_picture[1], s->linesize/2, s->mb_width*8, s->mb_height*8, EDGE_WIDTH/2);
-        draw_edges(s->current_picture[2], s->linesize/2, s->mb_width*8, s->mb_height*8, EDGE_WIDTH/2);
-      }else{
-        /* mpeg4? / opendivx / xvid */
-        draw_edges(s->current_picture[0], s->linesize, s->width, s->height, EDGE_WIDTH);
-        draw_edges(s->current_picture[1], s->linesize/2, s->width/2, s->height/2, EDGE_WIDTH/2);
-        draw_edges(s->current_picture[2], s->linesize/2, s->width/2, s->height/2, EDGE_WIDTH/2);
-      }
+    if (s->pict_type != B_TYPE && !s->intra_only && !(s->flags&CODEC_FLAG_EMU_EDGE)) {
+        draw_edges(s->current_picture[0], s->linesize  , s->h_edge_pos   , s->v_edge_pos   , EDGE_WIDTH  );
+        draw_edges(s->current_picture[1], s->uvlinesize, s->h_edge_pos>>1, s->v_edge_pos>>1, EDGE_WIDTH/2);
+        draw_edges(s->current_picture[2], s->uvlinesize, s->h_edge_pos>>1, s->v_edge_pos>>1, EDGE_WIDTH/2);
     }
     emms_c();
+    
+    s->last_pict_type    = s->pict_type;
+    if(s->pict_type!=B_TYPE){
+        s->last_non_b_pict_type= s->pict_type;
+        s->num_available_buffers++;
+        if(s->num_available_buffers>2) s->num_available_buffers= 2;
+    }
+}
+
+/* reorder input for encoding */
+void reorder_input(MpegEncContext *s, AVPicture *pict)
+{
+    int i, j, index;
+            
+    if(s->max_b_frames > FF_MAX_B_FRAMES) s->max_b_frames= FF_MAX_B_FRAMES;
+
+//        delay= s->max_b_frames+1; (or 0 if no b frames cuz decoder diff)
+
+    for(j=0; j<REORDER_BUFFER_SIZE-1; j++){
+        s->coded_order[j]= s->coded_order[j+1];
+    }
+    s->coded_order[j].picture[0]= s->coded_order[j].picture[1]= s->coded_order[j].picture[2]= NULL; //catch uninitalized buffers
+    s->coded_order[j].pict_type=0;
+
+    switch(s->input_pict_type){
+    default: 
+    case I_TYPE:
+    case S_TYPE:
+    case P_TYPE:
+        index= s->max_b_frames - s->b_frames_since_non_b;
+        s->b_frames_since_non_b=0;
+        break;            
+    case B_TYPE:
+        index= s->max_b_frames + 1;
+        s->b_frames_since_non_b++;
+        break;          
+    }
+//printf("index:%d type:%d strides: %d %d\n", index, s->input_pict_type, pict->linesize[0], s->linesize);
+    if(   (index==0 || (s->flags&CODEC_FLAG_INPUT_PRESERVED))
+       && pict->linesize[0] == s->linesize
+       && pict->linesize[1] == s->uvlinesize
+       && pict->linesize[2] == s->uvlinesize){
+//printf("ptr\n");
+        for(i=0; i<3; i++){
+            s->coded_order[index].picture[i]= pict->data[i];
+        }
+    }else{
+//printf("copy\n");
+        for(i=0; i<3; i++){
+            uint8_t *src = pict->data[i];
+            uint8_t *dest;
+            int src_wrap = pict->linesize[i];
+            int dest_wrap = s->linesize;
+            int w = s->width;
+            int h = s->height;
+
+            if(index==0) dest= s->last_picture[i]+16; //is current_picture indeed but the switch hapens after reordering
+            else         dest= s->picture_buffer[s->picture_buffer_index][i];
+
+            if (i >= 1) {
+                dest_wrap >>= 1;
+                w >>= 1;
+                h >>= 1;
+            }
+
+            s->coded_order[index].picture[i]= dest;
+            for(j=0;j<h;j++) {
+                memcpy(dest, src, w);
+                dest += dest_wrap;
+                src += src_wrap;
+            }
+        }
+        if(index!=0){
+            s->picture_buffer_index++;
+            if(s->picture_buffer_index >= REORDER_BUFFER_SIZE) s->picture_buffer_index=0;
+        }
+    }
+    s->coded_order[index].pict_type = s->input_pict_type;
+    s->coded_order[index].qscale    = s->input_qscale;
+    s->coded_order[index].force_type= s->force_input_type;
+    s->coded_order[index].picture_in_gop_number= s->input_picture_in_gop_number;
+    s->coded_order[index].picture_number= s->input_picture_number;
+
+    for(i=0; i<3; i++){
+        s->new_picture[i]= s->coded_order[0].picture[i];
+    }
 }
 
 int MPV_encode_picture(AVCodecContext *avctx,
@@ -491,87 +917,109 @@ int MPV_encode_picture(AVCodecContext *avctx,
 {
     MpegEncContext *s = avctx->priv_data;
     AVPicture *pict = data;
-    int i, j;
 
-    if (s->fixed_qscale) 
-        s->qscale = avctx->quality;
+    s->input_qscale = avctx->quality;
 
     init_put_bits(&s->pb, buf, buf_size, NULL, NULL);
 
-    if (!s->intra_only) {
-        /* first picture of GOP is intra */
-        if (s->picture_in_gop_number >= s->gop_size){
-            s->picture_in_gop_number=0;
-            s->pict_type = I_TYPE;
-        }else
-            s->pict_type = P_TYPE;
-    } else {
-        s->pict_type = I_TYPE;
+    if(avctx->flags&CODEC_FLAG_TYPE){
+        s->input_pict_type=
+        s->force_input_type= avctx->key_frame ? I_TYPE : P_TYPE;
+    }else if(s->flags&CODEC_FLAG_PASS2){
+        s->input_pict_type=
+        s->force_input_type= s->rc_context.entry[s->input_picture_number].new_pict_type;
+    }else{
+        s->force_input_type=0;
+        if (!s->intra_only) {
+            /* first picture of GOP is intra */
+            if (s->input_picture_in_gop_number % s->gop_size==0){
+                s->input_pict_type = I_TYPE;
+            }else if(s->max_b_frames==0){
+                s->input_pict_type = P_TYPE;
+            }else{
+                if(s->b_frames_since_non_b < s->max_b_frames) //FIXME more IQ
+                    s->input_pict_type = B_TYPE;
+                else
+                    s->input_pict_type = P_TYPE;
+            }
+        } else {
+            s->input_pict_type = I_TYPE;
+        }
     }
+
+    if(s->input_pict_type==I_TYPE)
+        s->input_picture_in_gop_number=0;
     
-    MPV_frame_start(s);
+    reorder_input(s, pict);
     
-    for(i=0;i<3;i++) {
-        UINT8 *src = pict->data[i];
-        UINT8 *dest = s->current_picture[i];
-        int src_wrap = pict->linesize[i];
-        int dest_wrap = s->linesize;
-        int w = s->width;
-        int h = s->height;
+    /* output? */
+    if(s->coded_order[0].picture[0]){
 
-        if (i >= 1) {
-            dest_wrap >>= 1;
-            w >>= 1;
-            h >>= 1;
-        }
+        s->pict_type= s->coded_order[0].pict_type;
+        if (s->fixed_qscale) /* the ratecontrol needs the last qscale so we dont touch it for CBR */
+            s->qscale= s->coded_order[0].qscale;
+        s->force_type= s->coded_order[0].force_type;
+        s->picture_in_gop_number= s->coded_order[0].picture_in_gop_number;
+        s->picture_number= s->coded_order[0].picture_number;
 
-        if(dest_wrap==src_wrap){
-            s->new_picture[i] = pict->data[i];
-        } else {
-            for(j=0;j<h;j++) {
-                memcpy(dest, src, w);
-                dest += dest_wrap;
-                src += src_wrap;
-            }
-            s->new_picture[i] = s->current_picture[i];
-           }
-    }
+        MPV_frame_start(s, avctx);
 
-    encode_picture(s, s->picture_number);
-    avctx->key_frame = (s->pict_type == I_TYPE);
+        encode_picture(s, s->picture_number);
+        
+        avctx->real_pict_num  = s->picture_number;
+        avctx->header_bits = s->header_bits;
+        avctx->mv_bits     = s->mv_bits;
+        avctx->misc_bits   = s->misc_bits;
+        avctx->i_tex_bits  = s->i_tex_bits;
+        avctx->p_tex_bits  = s->p_tex_bits;
+        avctx->i_count     = s->i_count;
+        avctx->p_count     = s->mb_num - s->i_count - s->skip_count; //FIXME f/b_count in avctx
+        avctx->skip_count  = s->skip_count;
+
+        MPV_frame_end(s);
+
+        if (s->out_format == FMT_MJPEG)
+            mjpeg_picture_trailer(s);
+
+        if(!s->fixed_qscale)
+            avctx->quality = s->qscale;
+        
+        if(s->flags&CODEC_FLAG_PASS1)
+            ff_write_pass1_stats(s);
     
-    MPV_frame_end(s);
-    s->picture_number++;
-    s->picture_in_gop_number++;
+    }
 
-    if (s->out_format == FMT_MJPEG)
-        mjpeg_picture_trailer(s);
+    s->input_picture_number++;
+    s->input_picture_in_gop_number++;
 
     flush_put_bits(&s->pb);
-    s->last_frame_bits= s->frame_bits;
     s->frame_bits  = (pbBufPtr(&s->pb) - s->pb.buf) * 8;
+    
     s->total_bits += s->frame_bits;
+    avctx->frame_bits  = s->frame_bits;
+//printf("fcode: %d, type: %d, head: %d, mv: %d, misc: %d, frame: %d, itex: %d, ptex: %d\n", 
+//s->f_code, avctx->key_frame, s->header_bits, s->mv_bits, s->misc_bits, s->frame_bits, s->i_tex_bits, s->p_tex_bits);
+#if 0 //dump some stats to stats.txt for testing/debuging
+if(s->max_b_frames==0)
+{
+    static FILE *f=NULL;
+    if(!f) f= fopen("stats.txt", "wb");
+    get_psnr(pict->data, s->current_picture,
+             pict->linesize, s->linesize, avctx);
+    fprintf(f, "%7d, %7d, %2.4f\n", pbBufPtr(&s->pb) - s->pb.buf, s->qscale, avctx->psnr_y);
+}
+#endif
 
-    avctx->quality = s->qscale;
     if (avctx->get_psnr) {
         /* At this point pict->data should have the original frame   */
         /* an s->current_picture should have the coded/decoded frame */
         get_psnr(pict->data, s->current_picture,
                  pict->linesize, s->linesize, avctx);
+//        printf("%f\n", avctx->psnr_y);
     }
     return pbBufPtr(&s->pb) - s->pb.buf;
 }
 
-static inline int clip(int a, int amin, int amax)
-{
-    if (a < amin)
-        return amin;
-    else if (a > amax)
-        return amax;
-    else
-        return a;
-}
-
 static inline void gmc1_motion(MpegEncContext *s,
                                UINT8 *dest_y, UINT8 *dest_cb, UINT8 *dest_cr,
                                int dest_offset,
@@ -579,8 +1027,9 @@ static inline void gmc1_motion(MpegEncContext *s,
                                int h)
 {
     UINT8 *ptr;
-    int dxy, offset, mx, my, src_x, src_y, height, linesize;
+    int offset, src_x, src_y, linesize, uvlinesize;
     int motion_x, motion_y;
+    int emu=0;
 
     if(s->real_sprite_warping_points>1) printf("more than 1 warp point isnt supported\n");
     motion_x= s->sprite_offset[0][0];
@@ -597,9 +1046,18 @@ static inline void gmc1_motion(MpegEncContext *s,
         motion_y =0;
     
     linesize = s->linesize;
+    uvlinesize = s->uvlinesize;
     ptr = ref_picture[0] + (src_y * linesize) + src_x + src_offset;
 
     dest_y+=dest_offset;
+    if(s->flags&CODEC_FLAG_EMU_EDGE){
+        if(src_x<0 || src_y<0 || src_x + (motion_x&15) + 16 > s->h_edge_pos
+                              || src_y + (motion_y&15) + h  > s->v_edge_pos){
+            emulated_edge_mc(s, ptr, linesize, 17, h+1, src_x, src_y, s->h_edge_pos, s->v_edge_pos);
+            ptr= s->edge_emu_buffer;
+            emu=1;
+        }
+    }
     gmc1(dest_y  , ptr  , linesize, h, motion_x&15, motion_y&15, s->no_rounding);
     gmc1(dest_y+8, ptr+8, linesize, h, motion_x&15, motion_y&15, s->no_rounding);
 
@@ -616,47 +1074,132 @@ static inline void gmc1_motion(MpegEncContext *s,
     if (src_y == s->height>>1)
         motion_y =0;
 
-    offset = (src_y * linesize>>1) + src_x + (src_offset>>1);
+    offset = (src_y * uvlinesize) + src_x + (src_offset>>1);
     ptr = ref_picture[1] + offset;
-    gmc1(dest_cb + (dest_offset>>1), ptr, linesize>>1, h>>1, motion_x&15, motion_y&15, s->no_rounding);
+    if(emu){
+        emulated_edge_mc(s, ptr, uvlinesize, 9, (h>>1)+1, src_x, src_y, s->h_edge_pos>>1, s->v_edge_pos>>1);
+        ptr= s->edge_emu_buffer;
+    }
+    gmc1(dest_cb + (dest_offset>>1), ptr, uvlinesize, h>>1, motion_x&15, motion_y&15, s->no_rounding);
+    
     ptr = ref_picture[2] + offset;
-    gmc1(dest_cr + (dest_offset>>1), ptr, linesize>>1, h>>1, motion_x&15, motion_y&15, s->no_rounding);
+    if(emu){
+        emulated_edge_mc(s, ptr, uvlinesize, 9, (h>>1)+1, src_x, src_y, s->h_edge_pos>>1, s->v_edge_pos>>1);
+        ptr= s->edge_emu_buffer;
+    }
+    gmc1(dest_cr + (dest_offset>>1), ptr, uvlinesize, h>>1, motion_x&15, motion_y&15, s->no_rounding);
     
     return;
 }
 
+static void emulated_edge_mc(MpegEncContext *s, UINT8 *src, int linesize, int block_w, int block_h, 
+                                    int src_x, int src_y, int w, int h){
+    int x, y;
+    int start_y, start_x, end_y, end_x;
+    UINT8 *buf= s->edge_emu_buffer;
+    
+    if(src_y>= h){
+        src+= (h-1-src_y)*linesize;
+        src_y=h-1;
+    }else if(src_y<=-block_h){
+        src+= (1-block_h-src_y)*linesize;
+        src_y=1-block_h;
+    }
+    if(src_x>= w){
+        src+= (w-1-src_x);
+        src_x=w-1;
+    }else if(src_x<=-block_w){
+        src+= (1-block_w-src_x);
+        src_x=1-block_w;
+    }
+
+    start_y= MAX(0, -src_y);
+    start_x= MAX(0, -src_x);
+    end_y= MIN(block_h, h-src_y);
+    end_x= MIN(block_w, w-src_x);
+
+    // copy existing part
+    for(y=start_y; y<end_y; y++){
+        for(x=start_x; x<end_x; x++){
+            buf[x + y*linesize]= src[x + y*linesize];
+        }
+    }
+
+    //top
+    for(y=0; y<start_y; y++){
+        for(x=start_x; x<end_x; x++){
+            buf[x + y*linesize]= buf[x + start_y*linesize];
+        }
+    }
+
+    //bottom
+    for(y=end_y; y<block_h; y++){
+        for(x=start_x; x<end_x; x++){
+            buf[x + y*linesize]= buf[x + (end_y-1)*linesize];
+        }
+    }
+                                    
+    for(y=0; y<block_h; y++){
+       //left
+        for(x=0; x<start_x; x++){
+            buf[x + y*linesize]= buf[start_x + y*linesize];
+        }
+       
+       //right
+        for(x=end_x; x<block_w; x++){
+            buf[x + y*linesize]= buf[end_x - 1 + y*linesize];
+        }
+    }
+}
+
+
 /* apply one mpeg motion vector to the three components */
 static inline void mpeg_motion(MpegEncContext *s,
                                UINT8 *dest_y, UINT8 *dest_cb, UINT8 *dest_cr,
                                int dest_offset,
                                UINT8 **ref_picture, int src_offset,
-                               int field_based, op_pixels_func *pix_op,
+                               int field_based, op_pixels_func (*pix_op)[4],
                                int motion_x, int motion_y, int h)
 {
     UINT8 *ptr;
-    int dxy, offset, mx, my, src_x, src_y, height, linesize;
+    int dxy, offset, mx, my, src_x, src_y, height, v_edge_pos, linesize, uvlinesize;
+    int emu=0;
+#if 0    
 if(s->quarter_sample)
 {
     motion_x>>=1;
     motion_y>>=1;
 }
+#endif
     dxy = ((motion_y & 1) << 1) | (motion_x & 1);
     src_x = s->mb_x * 16 + (motion_x >> 1);
     src_y = s->mb_y * (16 >> field_based) + (motion_y >> 1);
                 
     /* WARNING: do no forget half pels */
     height = s->height >> field_based;
+    v_edge_pos = s->v_edge_pos >> field_based;
     src_x = clip(src_x, -16, s->width);
     if (src_x == s->width)
         dxy &= ~1;
     src_y = clip(src_y, -16, height);
     if (src_y == height)
         dxy &= ~2;
-    linesize = s->linesize << field_based;
+    linesize   = s->linesize << field_based;
+    uvlinesize = s->uvlinesize << field_based;
     ptr = ref_picture[0] + (src_y * linesize) + (src_x) + src_offset;
     dest_y += dest_offset;
-    pix_op[dxy](dest_y, ptr, linesize, h);
-    pix_op[dxy](dest_y + 8, ptr + 8, linesize, h);
+
+    if(s->flags&CODEC_FLAG_EMU_EDGE){
+        if(src_x<0 || src_y<0 || src_x + (motion_x&1) + 16 > s->h_edge_pos
+                              || src_y + (motion_y&1) + h  > v_edge_pos){
+            emulated_edge_mc(s, ptr, linesize, 17, h+1, src_x, src_y, s->h_edge_pos, v_edge_pos);
+            ptr= s->edge_emu_buffer;
+            emu=1;
+        }
+    }
+    pix_op[0][dxy](dest_y, ptr, linesize, h);
+
+    if(s->flags&CODEC_FLAG_GRAY) return;
 
     if (s->out_format == FMT_H263) {
         dxy = 0;
@@ -682,30 +1225,40 @@ if(s->quarter_sample)
     src_y = clip(src_y, -8, height >> 1);
     if (src_y == (height >> 1))
         dxy &= ~2;
-
-    offset = (src_y * (linesize >> 1)) + src_x + (src_offset >> 1);
+    offset = (src_y * uvlinesize) + src_x + (src_offset >> 1);
     ptr = ref_picture[1] + offset;
-    pix_op[dxy](dest_cb + (dest_offset >> 1), ptr, linesize >> 1, h >> 1);
+    if(emu){
+        emulated_edge_mc(s, ptr, uvlinesize, 9, (h>>1)+1, src_x, src_y, s->h_edge_pos>>1, v_edge_pos>>1);
+        ptr= s->edge_emu_buffer;
+    }
+    pix_op[1][dxy](dest_cb + (dest_offset >> 1), ptr, uvlinesize, h >> 1);
+
     ptr = ref_picture[2] + offset;
-    pix_op[dxy](dest_cr + (dest_offset >> 1), ptr, linesize >> 1, h >> 1);
+    if(emu){
+        emulated_edge_mc(s, ptr, uvlinesize, 9, (h>>1)+1, src_x, src_y, s->h_edge_pos>>1, v_edge_pos>>1);
+        ptr= s->edge_emu_buffer;
+    }
+    pix_op[1][dxy](dest_cr + (dest_offset >> 1), ptr, uvlinesize, h >> 1);
 }
 
 static inline void qpel_motion(MpegEncContext *s,
                                UINT8 *dest_y, UINT8 *dest_cb, UINT8 *dest_cr,
                                int dest_offset,
                                UINT8 **ref_picture, int src_offset,
-                               int field_based, op_pixels_func *pix_op,
-                               qpel_mc_func *qpix_op,
+                               int field_based, op_pixels_func (*pix_op)[4],
+                               qpel_mc_func (*qpix_op)[16],
                                int motion_x, int motion_y, int h)
 {
     UINT8 *ptr;
-    int dxy, offset, mx, my, src_x, src_y, height, linesize;
+    int dxy, offset, mx, my, src_x, src_y, height, v_edge_pos, linesize, uvlinesize;
+    int emu=0;
 
     dxy = ((motion_y & 3) << 2) | (motion_x & 3);
     src_x = s->mb_x * 16 + (motion_x >> 2);
     src_y = s->mb_y * (16 >> field_based) + (motion_y >> 2);
 
     height = s->height >> field_based;
+    v_edge_pos = s->v_edge_pos >> field_based;
     src_x = clip(src_x, -16, s->width);
     if (src_x == s->width)
         dxy &= ~3;
@@ -713,25 +1266,46 @@ static inline void qpel_motion(MpegEncContext *s,
     if (src_y == height)
         dxy &= ~12;
     linesize = s->linesize << field_based;
+    uvlinesize = s->uvlinesize << field_based;
     ptr = ref_picture[0] + (src_y * linesize) + src_x + src_offset;
     dest_y += dest_offset;
 //printf("%d %d %d\n", src_x, src_y, dxy);
-    qpix_op[dxy](dest_y                 , ptr                 , linesize, linesize, motion_x&3, motion_y&3);
-    qpix_op[dxy](dest_y              + 8, ptr              + 8, linesize, linesize, motion_x&3, motion_y&3);
-    qpix_op[dxy](dest_y + linesize*8    , ptr + linesize*8    , linesize, linesize, motion_x&3, motion_y&3);
-    qpix_op[dxy](dest_y + linesize*8 + 8, ptr + linesize*8 + 8, linesize, linesize, motion_x&3, motion_y&3);
-    
-    mx= (motion_x>>1) | (motion_x&1);
-    my= (motion_y>>1) | (motion_y&1);
-
-    dxy = 0;
-    if ((mx & 3) != 0)
-        dxy |= 1;
-    if ((my & 3) != 0)
-        dxy |= 2;
-    mx = mx >> 2;
-    my = my >> 2;
     
+    if(s->flags&CODEC_FLAG_EMU_EDGE){
+        if(src_x<0 || src_y<0 || src_x + (motion_x&3) + 16 > s->h_edge_pos
+                              || src_y + (motion_y&3) + h  > v_edge_pos){
+            emulated_edge_mc(s, ptr, linesize, 17, h+1, src_x, src_y, s->h_edge_pos, v_edge_pos);
+            ptr= s->edge_emu_buffer;
+            emu=1;
+        }
+    }
+    if(!field_based)
+        qpix_op[0][dxy](dest_y, ptr, linesize);
+    else{
+        //damn interlaced mode
+        //FIXME boundary mirroring is not exactly correct here
+        qpix_op[1][dxy](dest_y  , ptr  , linesize);
+        qpix_op[1][dxy](dest_y+8, ptr+8, linesize);
+    }
+
+    if(s->flags&CODEC_FLAG_GRAY) return;
+
+    if(field_based){
+        mx= motion_x/2;
+        my= motion_y>>1;
+    }else if(s->divx_version){
+        mx= (motion_x>>1)|(motion_x&1);
+        my= (motion_y>>1)|(motion_y&1);
+    }else{
+        mx= motion_x/2;
+        my= motion_y/2;
+    }
+    mx= (mx>>1)|(mx&1);
+    my= (my>>1)|(my&1);
+    dxy= (mx&1) | ((my&1)<<1);
+    mx>>=1;
+    my>>=1;
+
     src_x = s->mb_x * 8 + mx;
     src_y = s->mb_y * (8 >> field_based) + my;
     src_x = clip(src_x, -8, s->width >> 1);
@@ -741,22 +1315,32 @@ static inline void qpel_motion(MpegEncContext *s,
     if (src_y == (height >> 1))
         dxy &= ~2;
 
-    offset = (src_y * (linesize >> 1)) + src_x + (src_offset >> 1);
+    offset = (src_y * uvlinesize) + src_x + (src_offset >> 1);
     ptr = ref_picture[1] + offset;
-    pix_op[dxy](dest_cb + (dest_offset >> 1), ptr, linesize >> 1, h >> 1);
+    if(emu){
+        emulated_edge_mc(s, ptr,  uvlinesize, 9, (h>>1)+1, src_x, src_y, s->h_edge_pos>>1, v_edge_pos>>1);
+        ptr= s->edge_emu_buffer;
+    }
+    pix_op[1][dxy](dest_cb + (dest_offset >> 1), ptr,  uvlinesize, h >> 1);
+    
     ptr = ref_picture[2] + offset;
-    pix_op[dxy](dest_cr + (dest_offset >> 1), ptr, linesize >> 1, h >> 1);
+    if(emu){
+        emulated_edge_mc(s, ptr,  uvlinesize, 9, (h>>1)+1, src_x, src_y, s->h_edge_pos>>1, v_edge_pos>>1);
+        ptr= s->edge_emu_buffer;
+    }
+    pix_op[1][dxy](dest_cr + (dest_offset >> 1), ptr,  uvlinesize, h >> 1);
 }
 
 
 static inline void MPV_motion(MpegEncContext *s, 
                               UINT8 *dest_y, UINT8 *dest_cb, UINT8 *dest_cr,
                               int dir, UINT8 **ref_picture, 
-                              op_pixels_func *pix_op, qpel_mc_func *qpix_op)
+                              op_pixels_func (*pix_op)[4], qpel_mc_func (*qpix_op)[16])
 {
     int dxy, offset, mx, my, src_x, src_y, motion_x, motion_y;
     int mb_x, mb_y, i;
     UINT8 *ptr, *dest;
+    int emu=0;
 
     mb_x = s->mb_x;
     mb_y = s->mb_y;
@@ -764,19 +1348,10 @@ static inline void MPV_motion(MpegEncContext *s,
     switch(s->mv_type) {
     case MV_TYPE_16X16:
         if(s->mcsel){
-#if 0
-            mpeg_motion(s, dest_y, dest_cb, dest_cr, 0,
-                        ref_picture, 0,
-                        0, pix_op,
-                        s->sprite_offset[0][0]>>3,
-                        s->sprite_offset[0][1]>>3,
-                        16);
-#else
             gmc1_motion(s, dest_y, dest_cb, dest_cr, 0,
                         ref_picture, 0,
                         16);
-#endif
-        }else if(s->quarter_sample && dir==0){ //FIXME
+        }else if(s->quarter_sample){
             qpel_motion(s, dest_y, dest_cb, dest_cr, 0,
                         ref_picture, 0,
                         0, pix_op, qpix_op,
@@ -789,33 +1364,76 @@ static inline void MPV_motion(MpegEncContext *s,
         }           
         break;
     case MV_TYPE_8X8:
-        for(i=0;i<4;i++) {
-            motion_x = s->mv[dir][i][0];
-            motion_y = s->mv[dir][i][1];
+        mx = 0;
+        my = 0;
+        if(s->quarter_sample){
+            for(i=0;i<4;i++) {
+                motion_x = s->mv[dir][i][0];
+                motion_y = s->mv[dir][i][1];
+
+                dxy = ((motion_y & 3) << 2) | (motion_x & 3);
+                src_x = mb_x * 16 + (motion_x >> 2) + (i & 1) * 8;
+                src_y = mb_y * 16 + (motion_y >> 2) + (i >>1) * 8;
+                    
+                /* WARNING: do no forget half pels */
+                src_x = clip(src_x, -16, s->width);
+                if (src_x == s->width)
+                    dxy &= ~3;
+                src_y = clip(src_y, -16, s->height);
+                if (src_y == s->height)
+                    dxy &= ~12;
+                    
+                ptr = ref_picture[0] + (src_y * s->linesize) + (src_x);
+                if(s->flags&CODEC_FLAG_EMU_EDGE){
+                    if(src_x<0 || src_y<0 || src_x + (motion_x&3) + 8 > s->h_edge_pos
+                                          || src_y + (motion_y&3) + 8 > s->v_edge_pos){
+                        emulated_edge_mc(s, ptr, s->linesize, 9, 9, src_x, src_y, s->h_edge_pos, s->v_edge_pos);
+                        ptr= s->edge_emu_buffer;
+                    }
+                }
+                dest = dest_y + ((i & 1) * 8) + (i >> 1) * 8 * s->linesize;
+                qpix_op[1][dxy](dest, ptr, s->linesize);
 
-            dxy = ((motion_y & 1) << 1) | (motion_x & 1);
-            src_x = mb_x * 16 + (motion_x >> 1) + (i & 1) * 8;
-            src_y = mb_y * 16 + (motion_y >> 1) + ((i >> 1) & 1) * 8;
+                mx += s->mv[dir][i][0]/2;
+                my += s->mv[dir][i][1]/2;
+            }
+        }else{
+            for(i=0;i<4;i++) {
+                motion_x = s->mv[dir][i][0];
+                motion_y = s->mv[dir][i][1];
+
+                dxy = ((motion_y & 1) << 1) | (motion_x & 1);
+                src_x = mb_x * 16 + (motion_x >> 1) + (i & 1) * 8;
+                src_y = mb_y * 16 + (motion_y >> 1) + (i >>1) * 8;
                     
-            /* WARNING: do no forget half pels */
-            src_x = clip(src_x, -16, s->width);
-            if (src_x == s->width)
-                dxy &= ~1;
-            src_y = clip(src_y, -16, s->height);
-            if (src_y == s->height)
-                dxy &= ~2;
+                /* WARNING: do no forget half pels */
+                src_x = clip(src_x, -16, s->width);
+                if (src_x == s->width)
+                    dxy &= ~1;
+                src_y = clip(src_y, -16, s->height);
+                if (src_y == s->height)
+                    dxy &= ~2;
                     
-            ptr = ref_picture[0] + (src_y * s->linesize) + (src_x);
-            dest = dest_y + ((i & 1) * 8) + (i >> 1) * 8 * s->linesize;
-            pix_op[dxy](dest, ptr, s->linesize, 8);
+                ptr = ref_picture[0] + (src_y * s->linesize) + (src_x);
+                if(s->flags&CODEC_FLAG_EMU_EDGE){
+                    if(src_x<0 || src_y<0 || src_x + (motion_x&1) + 8 > s->h_edge_pos
+                                          || src_y + (motion_y&1) + 8 > s->v_edge_pos){
+                        emulated_edge_mc(s, ptr, s->linesize, 9, 9, src_x, src_y, s->h_edge_pos, s->v_edge_pos);
+                        ptr= s->edge_emu_buffer;
+                    }
+                }
+                dest = dest_y + ((i & 1) * 8) + (i >> 1) * 8 * s->linesize;
+                pix_op[1][dxy](dest, ptr, s->linesize, 8);
+
+                mx += s->mv[dir][i][0];
+                my += s->mv[dir][i][1];
+            }
         }
+
+        if(s->flags&CODEC_FLAG_GRAY) break;
         /* In case of 8X8, we construct a single chroma motion vector
            with a special rounding */
-        mx = 0;
-        my = 0;
         for(i=0;i<4;i++) {
-            mx += s->mv[dir][i][0];
-            my += s->mv[dir][i][1];
         }
         if (mx >= 0)
             mx = (h263_chroma_roundtab[mx & 0xf] + ((mx >> 3) & ~1));
@@ -842,24 +1460,50 @@ static inline void MPV_motion(MpegEncContext *s,
         if (src_y == s->height/2)
             dxy &= ~2;
         
-        offset = (src_y * (s->linesize >> 1)) + src_x;
+        offset = (src_y * (s->uvlinesize)) + src_x;
         ptr = ref_picture[1] + offset;
-        pix_op[dxy](dest_cb, ptr, s->linesize >> 1, 8);
+        if(s->flags&CODEC_FLAG_EMU_EDGE){
+                if(src_x<0 || src_y<0 || src_x + (dxy &1) + 8 > s->h_edge_pos>>1
+                                      || src_y + (dxy>>1) + 8 > s->v_edge_pos>>1){
+                    emulated_edge_mc(s, ptr, s->uvlinesize, 9, 9, src_x, src_y, s->h_edge_pos>>1, s->v_edge_pos>>1);
+                    ptr= s->edge_emu_buffer;
+                    emu=1;
+                }
+            }
+        pix_op[1][dxy](dest_cb, ptr, s->uvlinesize, 8);
+
         ptr = ref_picture[2] + offset;
-        pix_op[dxy](dest_cr, ptr, s->linesize >> 1, 8);
+        if(emu){
+            emulated_edge_mc(s, ptr, s->uvlinesize, 9, 9, src_x, src_y, s->h_edge_pos>>1, s->v_edge_pos>>1);
+            ptr= s->edge_emu_buffer;
+        }
+        pix_op[1][dxy](dest_cr, ptr, s->uvlinesize, 8);
         break;
     case MV_TYPE_FIELD:
         if (s->picture_structure == PICT_FRAME) {
-            /* top field */
-            mpeg_motion(s, dest_y, dest_cb, dest_cr, 0,
-                        ref_picture, s->field_select[dir][0] ? s->linesize : 0,
-                        1, pix_op,
-                        s->mv[dir][0][0], s->mv[dir][0][1], 8);
-            /* bottom field */
-            mpeg_motion(s, dest_y, dest_cb, dest_cr, s->linesize,
-                        ref_picture, s->field_select[dir][1] ? s->linesize : 0,
-                        1, pix_op,
-                        s->mv[dir][1][0], s->mv[dir][1][1], 8);
+            if(s->quarter_sample){
+                /* top field */
+                qpel_motion(s, dest_y, dest_cb, dest_cr, 0,
+                            ref_picture, s->field_select[dir][0] ? s->linesize : 0,
+                            1, pix_op, qpix_op,
+                            s->mv[dir][0][0], s->mv[dir][0][1], 8);
+                /* bottom field */
+                qpel_motion(s, dest_y, dest_cb, dest_cr, s->linesize,
+                            ref_picture, s->field_select[dir][1] ? s->linesize : 0,
+                            1, pix_op, qpix_op,
+                            s->mv[dir][1][0], s->mv[dir][1][1], 8);
+            }else{
+                /* top field */       
+                mpeg_motion(s, dest_y, dest_cb, dest_cr, 0,
+                            ref_picture, s->field_select[dir][0] ? s->linesize : 0,
+                            1, pix_op,
+                            s->mv[dir][0][0], s->mv[dir][0][1], 8);
+                /* bottom field */
+                mpeg_motion(s, dest_y, dest_cb, dest_cr, s->linesize,
+                            ref_picture, s->field_select[dir][1] ? s->linesize : 0,
+                            1, pix_op,
+                            s->mv[dir][1][0], s->mv[dir][1][1], 8);
+            }
         } else {
             
 
@@ -875,8 +1519,7 @@ static inline void put_dct(MpegEncContext *s,
 {
     if (!s->mpeg2)
         s->dct_unquantize(s, block, i, s->qscale);
-    ff_idct (block);
-    put_pixels_clamped(block, dest, line_size);
+    s->idct_put (dest, line_size, block);
 }
 
 /* add block[] to dest[] */
@@ -884,14 +1527,53 @@ static inline void add_dct(MpegEncContext *s,
                            DCTELEM *block, int i, UINT8 *dest, int line_size)
 {
     if (s->block_last_index[i] >= 0) {
-        if (!s->mpeg2)
-            if(s->encoding || (!s->h263_msmpeg4))
-                s->dct_unquantize(s, block, i, s->qscale);
-        ff_idct (block);
-        add_pixels_clamped(block, dest, line_size);
+        s->idct_add (dest, line_size, block);
     }
 }
 
+static inline void add_dequant_dct(MpegEncContext *s, 
+                           DCTELEM *block, int i, UINT8 *dest, int line_size)
+{
+    if (s->block_last_index[i] >= 0) {
+        s->dct_unquantize(s, block, i, s->qscale);
+
+        s->idct_add (dest, line_size, block);
+    }
+}
+
+/**
+ * cleans dc, ac, coded_block for the current non intra MB
+ */
+void ff_clean_intra_table_entries(MpegEncContext *s)
+{
+    int wrap = s->block_wrap[0];
+    int xy = s->block_index[0];
+    
+    s->dc_val[0][xy           ] = 
+    s->dc_val[0][xy + 1       ] = 
+    s->dc_val[0][xy     + wrap] =
+    s->dc_val[0][xy + 1 + wrap] = 1024;
+    /* ac pred */
+    memset(s->ac_val[0][xy       ], 0, 32 * sizeof(INT16));
+    memset(s->ac_val[0][xy + wrap], 0, 32 * sizeof(INT16));
+    if (s->msmpeg4_version>=3) {
+        s->coded_block[xy           ] =
+        s->coded_block[xy + 1       ] =
+        s->coded_block[xy     + wrap] =
+        s->coded_block[xy + 1 + wrap] = 0;
+    }
+    /* chroma */
+    wrap = s->block_wrap[4];
+    xy = s->mb_x + 1 + (s->mb_y + 1) * wrap;
+    s->dc_val[1][xy] =
+    s->dc_val[2][xy] = 1024;
+    /* ac pred */
+    memset(s->ac_val[1][xy], 0, 16 * sizeof(INT16));
+    memset(s->ac_val[2][xy], 0, 16 * sizeof(INT16));
+    
+    s->mbintra_table[s->mb_x + s->mb_y*s->mb_width]= 0;
+}
+
 /* generic function called after a macroblock has been parsed by the
    decoder or after it has been encoded by the encoder.
 
@@ -905,77 +1587,67 @@ static inline void add_dct(MpegEncContext *s,
 void MPV_decode_mb(MpegEncContext *s, DCTELEM block[6][64])
 {
     int mb_x, mb_y;
-    int dct_linesize, dct_offset;
-    op_pixels_func *op_pix;
-    qpel_mc_func *op_qpix;
+    const int mb_xy = s->mb_y * s->mb_width + s->mb_x;
 
     mb_x = s->mb_x;
     mb_y = s->mb_y;
 
 #ifdef FF_POSTPROCESS
+    /* Obsolete. Exists for compatibility with mplayer only. */
     quant_store[mb_y][mb_x]=s->qscale;
     //printf("[%02d][%02d] %d\n",mb_x,mb_y,s->qscale);
+#else
+    /* even more obsolete, exists for mplayer xp only */
+    if(s->avctx->quant_store) s->avctx->quant_store[mb_y*s->avctx->qstride+mb_x] = s->qscale;
 #endif
+    s->qscale_table[mb_xy]= s->qscale;
 
     /* update DC predictors for P macroblocks */
     if (!s->mb_intra) {
         if (s->h263_pred || s->h263_aic) {
-          if(s->mbintra_table[mb_x + mb_y*s->mb_width])
-          {
-            int wrap, xy, v;
-            s->mbintra_table[mb_x + mb_y*s->mb_width]=0;
-            wrap = 2 * s->mb_width + 2;
-            xy = 2 * mb_x + 1 +  (2 * mb_y + 1) * wrap;
-            v = 1024;
-            
-            s->dc_val[0][xy] = v;
-            s->dc_val[0][xy + 1] = v;
-            s->dc_val[0][xy + wrap] = v;
-            s->dc_val[0][xy + 1 + wrap] = v;
-            /* ac pred */
-            memset(s->ac_val[0][xy], 0, 16 * sizeof(INT16));
-            memset(s->ac_val[0][xy + 1], 0, 16 * sizeof(INT16));
-            memset(s->ac_val[0][xy + wrap], 0, 16 * sizeof(INT16));
-            memset(s->ac_val[0][xy + 1 + wrap], 0, 16 * sizeof(INT16));
-            if (s->h263_msmpeg4) {
-                s->coded_block[xy] = 0;
-                s->coded_block[xy + 1] = 0;
-                s->coded_block[xy + wrap] = 0;
-                s->coded_block[xy + 1 + wrap] = 0;
-            }
-            /* chroma */
-            wrap = s->mb_width + 2;
-            xy = mb_x + 1 + (mb_y + 1) * wrap;
-            s->dc_val[1][xy] = v;
-            s->dc_val[2][xy] = v;
-            /* ac pred */
-            memset(s->ac_val[1][xy], 0, 16 * sizeof(INT16));
-            memset(s->ac_val[2][xy], 0, 16 * sizeof(INT16));
-          }
+            if(s->mbintra_table[mb_xy])
+                ff_clean_intra_table_entries(s);
         } else {
-            s->last_dc[0] = 128 << s->intra_dc_precision;
-            s->last_dc[1] = 128 << s->intra_dc_precision;
+            s->last_dc[0] =
+            s->last_dc[1] =
             s->last_dc[2] = 128 << s->intra_dc_precision;
         }
     }
     else if (s->h263_pred || s->h263_aic)
-        s->mbintra_table[mb_x + mb_y*s->mb_width]=1;
+        s->mbintra_table[mb_xy]=1;
 
     /* update motion predictor, not for B-frames as they need the motion_val from the last P/S-Frame */
-    if (s->out_format == FMT_H263) {
-      if(s->pict_type!=B_TYPE){
-        int xy, wrap, motion_x, motion_y;
+    if (s->out_format == FMT_H263 && s->pict_type!=B_TYPE) { //FIXME move into h263.c if possible, format specific stuff shouldnt be here
         
-        wrap = 2 * s->mb_width + 2;
-        xy = 2 * mb_x + 1 + (2 * mb_y + 1) * wrap;
-        if (s->mb_intra) {
-            motion_x = 0;
-            motion_y = 0;
-            goto motion_init;
-        } else if (s->mv_type == MV_TYPE_16X16) {
-            motion_x = s->mv[0][0][0];
-            motion_y = s->mv[0][0][1];
-        motion_init:
+        const int wrap = s->block_wrap[0];
+        const int xy = s->block_index[0];
+        const int mb_index= s->mb_x + s->mb_y*s->mb_width;
+        if(s->mv_type == MV_TYPE_8X8){
+            s->co_located_type_table[mb_index]= CO_LOCATED_TYPE_4MV;
+        } else {
+            int motion_x, motion_y;
+            if (s->mb_intra) {
+                motion_x = 0;
+                motion_y = 0;
+                if(s->co_located_type_table)
+                    s->co_located_type_table[mb_index]= 0;
+            } else if (s->mv_type == MV_TYPE_16X16) {
+                motion_x = s->mv[0][0][0];
+                motion_y = s->mv[0][0][1];
+                if(s->co_located_type_table)
+                    s->co_located_type_table[mb_index]= 0;
+            } else /*if (s->mv_type == MV_TYPE_FIELD)*/ {
+                int i;
+                motion_x = s->mv[0][0][0] + s->mv[0][1][0];
+                motion_y = s->mv[0][0][1] + s->mv[0][1][1];
+                motion_x = (motion_x>>1) | (motion_x&1);
+                for(i=0; i<2; i++){
+                    s->field_mv_table[mb_index][i][0]= s->mv[0][i][0];
+                    s->field_mv_table[mb_index][i][1]= s->mv[0][i][1];
+                    s->field_select_table[mb_index][i]= s->field_select[0][i];
+                }
+                s->co_located_type_table[mb_index]= CO_LOCATED_TYPE_FIELDMV;
+            }
             /* no update if 8X8 because it has been done during parsing */
             s->motion_val[xy][0] = motion_x;
             s->motion_val[xy][1] = motion_y;
@@ -986,30 +1658,36 @@ void MPV_decode_mb(MpegEncContext *s, DCTELEM block[6][64])
             s->motion_val[xy + 1 + wrap][0] = motion_x;
             s->motion_val[xy + 1 + wrap][1] = motion_y;
         }
-      }
     }
     
-    if (!s->intra_only) {
+    if (!(s->encoding && (s->intra_only || s->pict_type==B_TYPE))) {
         UINT8 *dest_y, *dest_cb, *dest_cr;
-        UINT8 *mbskip_ptr;
-
-        /* avoid copy if macroblock skipped in last frame too */
-        if (!s->encoding && s->pict_type != B_TYPE) {
-            mbskip_ptr = &s->mbskip_table[s->mb_y * s->mb_width + s->mb_x];
+        int dct_linesize, dct_offset;
+        op_pixels_func (*op_pix)[4];
+        qpel_mc_func (*op_qpix)[16];
+
+        /* avoid copy if macroblock skipped in last frame too 
+           dont touch it for B-frames as they need the skip info from the next p-frame */
+        if (s->pict_type != B_TYPE) {
+            UINT8 *mbskip_ptr = &s->mbskip_table[mb_xy];
             if (s->mb_skiped) {
                 s->mb_skiped = 0;
-                /* if previous was skipped too, then nothing to do ! */
-                if (*mbskip_ptr != 0) 
+
+                (*mbskip_ptr) ++; /* indicate that this time we skiped it */
+                if(*mbskip_ptr >99) *mbskip_ptr= 99;
+
+                /* if previous was skipped too, then nothing to do ! 
+                   skip only during decoding as we might trash the buffers during encoding a bit */
+                if (*mbskip_ptr >= s->ip_buffer_count  && !s->encoding) 
                     goto the_end;
-                *mbskip_ptr = 1; /* indicate that this time we skiped it */
             } else {
                 *mbskip_ptr = 0; /* not skipped */
             }
         }
 
-        dest_y = s->current_picture[0] + (mb_y * 16 * s->linesize) + mb_x * 16;
-        dest_cb = s->current_picture[1] + (mb_y * 8 * (s->linesize >> 1)) + mb_x * 8;
-        dest_cr = s->current_picture[2] + (mb_y * 8 * (s->linesize >> 1)) + mb_x * 8;
+        dest_y = s->current_picture [0] + (mb_y * 16* s->linesize  ) + mb_x * 16;
+        dest_cb = s->current_picture[1] + (mb_y * 8 * s->uvlinesize) + mb_x * 8;
+        dest_cr = s->current_picture[2] + (mb_y * 8 * s->uvlinesize) + mb_x * 8;
 
         if (s->interlaced_dct) {
             dct_linesize = s->linesize * 2;
@@ -1021,33 +1699,51 @@ void MPV_decode_mb(MpegEncContext *s, DCTELEM block[6][64])
 
         if (!s->mb_intra) {
             /* motion handling */
-            if (!s->no_rounding){
-                op_pix = put_pixels_tab;
-                op_qpix= qpel_mc_rnd_tab;
-            }else{
-                op_pix = put_no_rnd_pixels_tab;
-                op_qpix= qpel_mc_no_rnd_tab;
-            }
+            /* decoding or more than one mb_type (MC was allready done otherwise) */
+            if((!s->encoding) || (s->mb_type[mb_xy]&(s->mb_type[mb_xy]-1))){
+                if ((!s->no_rounding) || s->pict_type==B_TYPE){                
+                    op_pix = put_pixels_tab;
+                    op_qpix= put_qpel_pixels_tab;
+                }else{
+                    op_pix = put_no_rnd_pixels_tab;
+                    op_qpix= put_no_rnd_qpel_pixels_tab;
+                }
 
-            if (s->mv_dir & MV_DIR_FORWARD) {
-                MPV_motion(s, dest_y, dest_cb, dest_cr, 0, s->last_picture, op_pix, op_qpix);
-                if (!s->no_rounding) 
+                if (s->mv_dir & MV_DIR_FORWARD) {
+                    MPV_motion(s, dest_y, dest_cb, dest_cr, 0, s->last_picture, op_pix, op_qpix);
                     op_pix = avg_pixels_tab;
-                else
-                    op_pix = avg_no_rnd_pixels_tab;
-            }
-            if (s->mv_dir & MV_DIR_BACKWARD) {
-                MPV_motion(s, dest_y, dest_cb, dest_cr, 1, s->next_picture, op_pix, op_qpix);
+                    op_qpix= avg_qpel_pixels_tab;
+                }
+                if (s->mv_dir & MV_DIR_BACKWARD) {
+                    MPV_motion(s, dest_y, dest_cb, dest_cr, 1, s->next_picture, op_pix, op_qpix);
+                }
             }
 
-            /* add dct residue */
-            add_dct(s, block[0], 0, dest_y, dct_linesize);
-            add_dct(s, block[1], 1, dest_y + 8, dct_linesize);
-            add_dct(s, block[2], 2, dest_y + dct_offset, dct_linesize);
-            add_dct(s, block[3], 3, dest_y + dct_offset + 8, dct_linesize);
+            /* skip dequant / idct if we are really late ;) */
+            if(s->hurry_up>1) goto the_end;
 
-            add_dct(s, block[4], 4, dest_cb, s->linesize >> 1);
-            add_dct(s, block[5], 5, dest_cr, s->linesize >> 1);
+            /* add dct residue */
+            if(s->encoding || !(s->mpeg2 || s->h263_msmpeg4 || (s->codec_id==CODEC_ID_MPEG4 && !s->mpeg_quant))){
+                add_dequant_dct(s, block[0], 0, dest_y, dct_linesize);
+                add_dequant_dct(s, block[1], 1, dest_y + 8, dct_linesize);
+                add_dequant_dct(s, block[2], 2, dest_y + dct_offset, dct_linesize);
+                add_dequant_dct(s, block[3], 3, dest_y + dct_offset + 8, dct_linesize);
+
+                if(!(s->flags&CODEC_FLAG_GRAY)){
+                    add_dequant_dct(s, block[4], 4, dest_cb, s->uvlinesize);
+                    add_dequant_dct(s, block[5], 5, dest_cr, s->uvlinesize);
+                }
+            } else {
+                add_dct(s, block[0], 0, dest_y, dct_linesize);
+                add_dct(s, block[1], 1, dest_y + 8, dct_linesize);
+                add_dct(s, block[2], 2, dest_y + dct_offset, dct_linesize);
+                add_dct(s, block[3], 3, dest_y + dct_offset + 8, dct_linesize);
+
+                if(!(s->flags&CODEC_FLAG_GRAY)){
+                    add_dct(s, block[4], 4, dest_cb, s->uvlinesize);
+                    add_dct(s, block[5], 5, dest_cr, s->uvlinesize);
+                }
+            }
         } else {
             /* dct only in intra block */
             put_dct(s, block[0], 0, dest_y, dct_linesize);
@@ -1055,127 +1751,721 @@ void MPV_decode_mb(MpegEncContext *s, DCTELEM block[6][64])
             put_dct(s, block[2], 2, dest_y + dct_offset, dct_linesize);
             put_dct(s, block[3], 3, dest_y + dct_offset + 8, dct_linesize);
 
-            put_dct(s, block[4], 4, dest_cb, s->linesize >> 1);
-            put_dct(s, block[5], 5, dest_cr, s->linesize >> 1);
+            if(!(s->flags&CODEC_FLAG_GRAY)){
+                put_dct(s, block[4], 4, dest_cb, s->uvlinesize);
+                put_dct(s, block[5], 5, dest_cr, s->uvlinesize);
+            }
         }
     }
  the_end:
-    emms_c();
+    emms_c(); //FIXME remove
 }
 
-static void encode_picture(MpegEncContext *s, int picture_number)
+static inline void dct_single_coeff_elimination(MpegEncContext *s, int n, int threshold)
 {
-    int mb_x, mb_y, wrap, last_gob, pdif = 0;
-    UINT8 *ptr;
-    int i, motion_x, motion_y;
+    static const char tab[64]=
+        {3,2,2,1,1,1,1,1,
+         1,1,1,1,1,1,1,1,
+         1,1,1,1,1,1,1,1,
+         0,0,0,0,0,0,0,0,
+         0,0,0,0,0,0,0,0,
+         0,0,0,0,0,0,0,0,
+         0,0,0,0,0,0,0,0,
+         0,0,0,0,0,0,0,0};
+    int score=0;
+    int run=0;
+    int i;
+    DCTELEM *block= s->block[n];
+    const int last_index= s->block_last_index[n];
+    int skip_dc;
+
+    if(threshold<0){
+        skip_dc=0;
+        threshold= -threshold;
+    }else
+        skip_dc=1;
+
+    /* are all which we could set to zero are allready zero? */
+    if(last_index<=skip_dc - 1) return;
+
+    for(i=0; i<=last_index; i++){
+        const int j = s->intra_scantable.permutated[i];
+        const int level = ABS(block[j]);
+        if(level==1){
+            if(skip_dc && i==0) continue;
+            score+= tab[run];
+            run=0;
+        }else if(level>1){
+            return;
+        }else{
+            run++;
+        }
+    }
+    if(score >= threshold) return;
+    for(i=skip_dc; i<=last_index; i++){
+        const int j = s->intra_scantable.permutated[i];
+        block[j]=0;
+    }
+    if(block[0]) s->block_last_index[n]= 0;
+    else         s->block_last_index[n]= -1;
+}
 
-    s->picture_number = picture_number;
+static inline void clip_coeffs(MpegEncContext *s, DCTELEM *block, int last_index)
+{
+    int i;
+    const int maxlevel= s->max_qcoeff;
+    const int minlevel= s->min_qcoeff;
+    
+    if(s->mb_intra){
+        i=1; //skip clipping of intra dc
+    }else
+        i=0;
+    
+    for(;i<=last_index; i++){
+        const int j= s->intra_scantable.permutated[i];
+        int level = block[j];
+       
+        if     (level>maxlevel) level=maxlevel;
+        else if(level<minlevel) level=minlevel;
+        block[j]= level;
+    }
+}
 
-    s->last_mc_mb_var = s->mc_mb_var;
-    /* Reset the average MB variance */
-    s->avg_mb_var = 0;
-    s->mc_mb_var = 0;
+static inline void requantize_coeffs(MpegEncContext *s, DCTELEM block[64], int oldq, int newq, int n)
+{
+    int i;
 
-    /* Estimate motion for every MB */
-    for(mb_y=0; mb_y < s->mb_height; mb_y++) {
-        for(mb_x=0; mb_x < s->mb_width; mb_x++) {
-            int xy= mb_y * s->mb_width + mb_x;
-            s->mb_x = mb_x;
-            s->mb_y = mb_y;
+    if(s->mb_intra){
+        i=1; //skip clipping of intra dc
+         //FIXME requantize, note (mpeg1/h263/h263p-aic dont need it,...)
+    }else
+        i=0;
+    
+    for(;i<=s->block_last_index[n]; i++){
+        const int j = s->intra_scantable.permutated[i];
+        int level = block[j];
+        
+        block[j]= ROUNDED_DIV(level*oldq, newq);
+    }
 
-            /* compute motion vector and macro block type (intra or non intra) */
-            motion_x = 0;
-            motion_y = 0;
-            if (s->pict_type == P_TYPE) {
-                s->mb_intra = estimate_motion(s, mb_x, mb_y,
-                                              &motion_x,
-                                              &motion_y);
-            } else {
-                s->mb_intra = 1;
-            }
-            /* Store MB type and MV */
-            s->mb_type[xy] = s->mb_intra;
-            s->mv_table[0][xy] = motion_x;
-            s->mv_table[1][xy] = motion_y;
+    for(i=s->block_last_index[n]; i>=0; i--){
+        const int j = s->intra_scantable.permutated[i];
+        if(block[j]) break;
+    }
+    s->block_last_index[n]= i;
+}
+
+static inline void auto_requantize_coeffs(MpegEncContext *s, DCTELEM block[6][64])
+{
+    int i,n, newq;
+    const int maxlevel= s->max_qcoeff;
+    const int minlevel= s->min_qcoeff;
+    int largest=0, smallest=0;
+
+    assert(s->adaptive_quant);
+    
+    for(n=0; n<6; n++){
+        if(s->mb_intra){
+            i=1; //skip clipping of intra dc
+             //FIXME requantize, note (mpeg1/h263/h263p-aic dont need it,...)
+        }else
+            i=0;
+
+        for(;i<=s->block_last_index[n]; i++){
+            const int j = s->intra_scantable.permutated[i];
+            int level = block[n][j];
+            if(largest  < level) largest = level;
+            if(smallest > level) smallest= level;
         }
     }
-    emms_c();
+    
+    for(newq=s->qscale+1; newq<32; newq++){
+        if(   ROUNDED_DIV(smallest*s->qscale, newq) >= minlevel
+           && ROUNDED_DIV(largest *s->qscale, newq) <= maxlevel) 
+            break;
+    }
+        
+    if(s->out_format==FMT_H263){
+        /* h263 like formats cannot change qscale by more than 2 easiely */
+        if(s->avctx->qmin + 2 < newq)
+            newq= s->avctx->qmin + 2;
+    }
 
-    if(s->avg_mb_var < s->mc_mb_var && s->pict_type != B_TYPE){ //FIXME subtract MV bits
-        int i;
-        s->pict_type= I_TYPE;
-        s->picture_in_gop_number=0;
-        for(i=0; i<s->mb_num; i++){
-            s->mb_type[i] = 1;
-            s->mv_table[0][i] = 0;
-            s->mv_table[1][i] = 0;
+    for(n=0; n<6; n++){
+        requantize_coeffs(s, block[n], s->qscale, newq, n);
+        clip_coeffs(s, block[n], s->block_last_index[n]);
+    }
+     
+    s->dquant+= newq - s->qscale;
+    s->qscale= newq;
+}
+#if 0
+static int pix_vcmp16x8(UINT8 *s, int stride){ //FIXME move to dsputil & optimize
+    int score=0;
+    int x,y;
+    
+    for(y=0; y<7; y++){
+        for(x=0; x<16; x+=4){
+            score+= ABS(s[x  ] - s[x  +stride]) + ABS(s[x+1] - s[x+1+stride]) 
+                   +ABS(s[x+2] - s[x+2+stride]) + ABS(s[x+3] - s[x+3+stride]);
         }
+        s+= stride;
     }
+    
+    return score;
+}
 
-    /* find best f_code */
-    if(s->pict_type==P_TYPE){
-        int mv_num[8];
-        int i;
-        int loose=0;
-        UINT8 * fcode_tab= s->fcode_tab;
+static int pix_diff_vcmp16x8(UINT8 *s1, UINT8*s2, int stride){ //FIXME move to dsputil & optimize
+    int score=0;
+    int x,y;
+    
+    for(y=0; y<7; y++){
+        for(x=0; x<16; x++){
+            score+= ABS(s1[x  ] - s2[x ] - s1[x  +stride] + s2[x +stride]);
+        }
+        s1+= stride;
+        s2+= stride;
+    }
+    
+    return score;
+}
+#else
+#define SQ(a) ((a)*(a))
 
-        for(i=0; i<8; i++) mv_num[i]=0;
+static int pix_vcmp16x8(UINT8 *s, int stride){ //FIXME move to dsputil & optimize
+    int score=0;
+    int x,y;
+    
+    for(y=0; y<7; y++){
+        for(x=0; x<16; x+=4){
+            score+= SQ(s[x  ] - s[x  +stride]) + SQ(s[x+1] - s[x+1+stride]) 
+                   +SQ(s[x+2] - s[x+2+stride]) + SQ(s[x+3] - s[x+3+stride]);
+        }
+        s+= stride;
+    }
+    
+    return score;
+}
+
+static int pix_diff_vcmp16x8(UINT8 *s1, UINT8*s2, int stride){ //FIXME move to dsputil & optimize
+    int score=0;
+    int x,y;
+    
+    for(y=0; y<7; y++){
+        for(x=0; x<16; x++){
+            score+= SQ(s1[x  ] - s2[x ] - s1[x  +stride] + s2[x +stride]);
+        }
+        s1+= stride;
+        s2+= stride;
+    }
+    
+    return score;
+}
+
+#endif
+static void encode_mb(MpegEncContext *s, int motion_x, int motion_y)
+{
+    const int mb_x= s->mb_x;
+    const int mb_y= s->mb_y;
+    int i;
+    int skip_dct[6];
+    int dct_offset   = s->linesize*8; //default for progressive frames
+    
+    for(i=0; i<6; i++) skip_dct[i]=0;
+    
+    if(s->adaptive_quant){
+        s->dquant= s->qscale_table[mb_x + mb_y*s->mb_width] - s->qscale;
 
-        for(i=0; i<s->mb_num; i++){
-            if(s->mb_type[i] == 0){
-                mv_num[ fcode_tab[s->mv_table[0][i] + MAX_MV] ]++;
-                mv_num[ fcode_tab[s->mv_table[1][i] + MAX_MV] ]++;
-//printf("%d %d %d\n", s->mv_table[0][i], fcode_tab[s->mv_table[0][i] + MAX_MV], i);
+        if(s->out_format==FMT_H263){
+            if     (s->dquant> 2) s->dquant= 2;
+            else if(s->dquant<-2) s->dquant=-2;
+        }
+            
+        if(s->codec_id==CODEC_ID_MPEG4){        
+            if(!s->mb_intra){
+                assert(s->dquant==0 || s->mv_type!=MV_TYPE_8X8);
+
+                if(s->mv_dir&MV_DIRECT)
+                    s->dquant=0;
             }
-//else printf("I");
         }
+        s->qscale+= s->dquant;
+        s->y_dc_scale= s->y_dc_scale_table[ s->qscale ];
+        s->c_dc_scale= s->c_dc_scale_table[ s->qscale ];
+    }
 
-        for(i=MAX_FCODE; i>1; i--){
-            loose+= mv_num[i];
-            if(loose > 4) break;
+    if (s->mb_intra) {
+        UINT8 *ptr;
+        int wrap_y;
+        int emu=0;
+
+        wrap_y = s->linesize;
+        ptr = s->new_picture[0] + (mb_y * 16 * wrap_y) + mb_x * 16;
+
+        if(mb_x*16+16 > s->width || mb_y*16+16 > s->height){
+            emulated_edge_mc(s, ptr, wrap_y, 16, 16, mb_x*16, mb_y*16, s->width, s->height);
+            ptr= s->edge_emu_buffer;
+            emu=1;
+        }
+        
+        if(s->flags&CODEC_FLAG_INTERLACED_DCT){
+            int progressive_score, interlaced_score;
+            
+            progressive_score= pix_vcmp16x8(ptr, wrap_y  ) + pix_vcmp16x8(ptr + wrap_y*8, wrap_y );
+            interlaced_score = pix_vcmp16x8(ptr, wrap_y*2) + pix_vcmp16x8(ptr + wrap_y  , wrap_y*2);
+            
+            if(progressive_score > interlaced_score + 100){
+                s->interlaced_dct=1;
+            
+                dct_offset= wrap_y;
+                wrap_y<<=1;
+            }else
+                s->interlaced_dct=0;
+        }
+        
+        get_pixels(s->block[0], ptr                 , wrap_y);
+        get_pixels(s->block[1], ptr              + 8, wrap_y);
+        get_pixels(s->block[2], ptr + dct_offset    , wrap_y);
+        get_pixels(s->block[3], ptr + dct_offset + 8, wrap_y);
+
+        if(s->flags&CODEC_FLAG_GRAY){
+            skip_dct[4]= 1;
+            skip_dct[5]= 1;
+        }else{
+            int wrap_c = s->uvlinesize;
+            ptr = s->new_picture[1] + (mb_y * 8 * wrap_c) + mb_x * 8;
+            if(emu){
+                emulated_edge_mc(s, ptr, wrap_c, 8, 8, mb_x*8, mb_y*8, s->width>>1, s->height>>1);
+                ptr= s->edge_emu_buffer;
+            }
+            get_pixels(s->block[4], ptr, wrap_c);
+
+            ptr = s->new_picture[2] + (mb_y * 8 * wrap_c) + mb_x * 8;
+            if(emu){
+                emulated_edge_mc(s, ptr, wrap_c, 8, 8, mb_x*8, mb_y*8, s->width>>1, s->height>>1);
+                ptr= s->edge_emu_buffer;
+            }
+            get_pixels(s->block[5], ptr, wrap_c);
         }
-        s->f_code= i;
     }else{
-        s->f_code= 1;
+        op_pixels_func (*op_pix)[4];
+        qpel_mc_func (*op_qpix)[16];
+        UINT8 *dest_y, *dest_cb, *dest_cr;
+        UINT8 *ptr_y, *ptr_cb, *ptr_cr;
+        int wrap_y, wrap_c;
+        int emu=0;
+
+        dest_y  = s->current_picture[0] + (mb_y * 16 * s->linesize       ) + mb_x * 16;
+        dest_cb = s->current_picture[1] + (mb_y * 8  * (s->uvlinesize)) + mb_x * 8;
+        dest_cr = s->current_picture[2] + (mb_y * 8  * (s->uvlinesize)) + mb_x * 8;
+        wrap_y = s->linesize;
+        wrap_c = s->uvlinesize;
+        ptr_y  = s->new_picture[0] + (mb_y * 16 * wrap_y) + mb_x * 16;
+        ptr_cb = s->new_picture[1] + (mb_y * 8 * wrap_c) + mb_x * 8;
+        ptr_cr = s->new_picture[2] + (mb_y * 8 * wrap_c) + mb_x * 8;
+
+        if ((!s->no_rounding) || s->pict_type==B_TYPE){
+            op_pix = put_pixels_tab;
+            op_qpix= put_qpel_pixels_tab;
+        }else{
+            op_pix = put_no_rnd_pixels_tab;
+            op_qpix= put_no_rnd_qpel_pixels_tab;
+        }
+
+        if (s->mv_dir & MV_DIR_FORWARD) {
+            MPV_motion(s, dest_y, dest_cb, dest_cr, 0, s->last_picture, op_pix, op_qpix);
+            op_pix = avg_pixels_tab;
+            op_qpix= avg_qpel_pixels_tab;
+        }
+        if (s->mv_dir & MV_DIR_BACKWARD) {
+            MPV_motion(s, dest_y, dest_cb, dest_cr, 1, s->next_picture, op_pix, op_qpix);
+        }
+
+        if(mb_x*16+16 > s->width || mb_y*16+16 > s->height){
+            emulated_edge_mc(s, ptr_y, wrap_y, 16, 16, mb_x*16, mb_y*16, s->width, s->height);
+            ptr_y= s->edge_emu_buffer;
+            emu=1;
+        }
+        
+        if(s->flags&CODEC_FLAG_INTERLACED_DCT){
+            int progressive_score, interlaced_score;
+            
+            progressive_score= pix_diff_vcmp16x8(ptr_y           , dest_y           , wrap_y  ) 
+                             + pix_diff_vcmp16x8(ptr_y + wrap_y*8, dest_y + wrap_y*8, wrap_y  );
+            interlaced_score = pix_diff_vcmp16x8(ptr_y           , dest_y           , wrap_y*2)
+                             + pix_diff_vcmp16x8(ptr_y + wrap_y  , dest_y + wrap_y  , wrap_y*2);
+            
+            if(progressive_score > interlaced_score + 600){
+                s->interlaced_dct=1;
+            
+                dct_offset= wrap_y;
+                wrap_y<<=1;
+            }else
+                s->interlaced_dct=0;
+        }
+        
+        diff_pixels(s->block[0], ptr_y                 , dest_y                 , wrap_y);
+        diff_pixels(s->block[1], ptr_y              + 8, dest_y              + 8, wrap_y);
+        diff_pixels(s->block[2], ptr_y + dct_offset    , dest_y + dct_offset    , wrap_y);
+        diff_pixels(s->block[3], ptr_y + dct_offset + 8, dest_y + dct_offset + 8, wrap_y);
+        
+        if(s->flags&CODEC_FLAG_GRAY){
+            skip_dct[4]= 1;
+            skip_dct[5]= 1;
+        }else{
+            if(emu){
+                emulated_edge_mc(s, ptr_cb, wrap_c, 8, 8, mb_x*8, mb_y*8, s->width>>1, s->height>>1);
+                ptr_cb= s->edge_emu_buffer;
+            }
+            diff_pixels(s->block[4], ptr_cb, dest_cb, wrap_c);
+            if(emu){
+                emulated_edge_mc(s, ptr_cr, wrap_c, 8, 8, mb_x*8, mb_y*8, s->width>>1, s->height>>1);
+                ptr_cr= s->edge_emu_buffer;
+            }
+            diff_pixels(s->block[5], ptr_cr, dest_cr, wrap_c);
+        }
+
+        /* pre quantization */         
+        if(s->mc_mb_var[s->mb_width*mb_y+ mb_x]<2*s->qscale*s->qscale){
+            //FIXME optimize
+            if(pix_abs8x8(ptr_y               , dest_y               , wrap_y) < 20*s->qscale) skip_dct[0]= 1;
+            if(pix_abs8x8(ptr_y            + 8, dest_y            + 8, wrap_y) < 20*s->qscale) skip_dct[1]= 1;
+            if(pix_abs8x8(ptr_y +dct_offset   , dest_y +dct_offset   , wrap_y) < 20*s->qscale) skip_dct[2]= 1;
+            if(pix_abs8x8(ptr_y +dct_offset+ 8, dest_y +dct_offset+ 8, wrap_y) < 20*s->qscale) skip_dct[3]= 1;
+            if(pix_abs8x8(ptr_cb              , dest_cb              , wrap_y) < 20*s->qscale) skip_dct[4]= 1;
+            if(pix_abs8x8(ptr_cr              , dest_cr              , wrap_y) < 20*s->qscale) skip_dct[5]= 1;
+#if 0
+{
+ static int stat[7];
+ int num=0;
+ for(i=0; i<6; i++)
+  if(skip_dct[i]) num++;
+ stat[num]++;
+ if(s->mb_x==0 && s->mb_y==0){
+  for(i=0; i<7; i++){
+   printf("%6d %1d\n", stat[i], i);
+  }
+ }
+}
+#endif
+        }
+
     }
-//printf("f_code %d ///\n", s->f_code);
-    /* convert MBs with too long MVs to I-Blocks */
-    if(s->pict_type==P_TYPE){
-        int i;
-        const int f_code= s->f_code;
-        UINT8 * fcode_tab= s->fcode_tab;
-
-        for(i=0; i<s->mb_num; i++){
-            if(s->mb_type[i] == 0){
-                if(   fcode_tab[s->mv_table[0][i] + MAX_MV] > f_code
-                   || fcode_tab[s->mv_table[0][i] + MAX_MV] == 0
-                   || fcode_tab[s->mv_table[1][i] + MAX_MV] > f_code
-                   || fcode_tab[s->mv_table[1][i] + MAX_MV] == 0 ){
-                    s->mb_type[i] = 1;
-                    s->mv_table[0][i] = 0;
-                    s->mv_table[1][i] = 0;
-                }
+            
+#if 0
+            {
+                float adap_parm;
+                
+                adap_parm = ((s->avg_mb_var << 1) + s->mb_var[s->mb_width*mb_y+mb_x] + 1.0) /
+                            ((s->mb_var[s->mb_width*mb_y+mb_x] << 1) + s->avg_mb_var + 1.0);
+            
+                printf("\ntype=%c qscale=%2d adap=%0.2f dquant=%4.2f var=%4d avgvar=%4d", 
+                        (s->mb_type[s->mb_width*mb_y+mb_x] > 0) ? 'I' : 'P', 
+                        s->qscale, adap_parm, s->qscale*adap_parm,
+                        s->mb_var[s->mb_width*mb_y+mb_x], s->avg_mb_var);
             }
+#endif
+    /* DCT & quantize */
+    if(s->out_format==FMT_MJPEG){
+        for(i=0;i<6;i++) {
+            int overflow;
+            s->block_last_index[i] = s->dct_quantize(s, s->block[i], i, 8, &overflow);
+            if (overflow) clip_coeffs(s, s->block[i], s->block_last_index[i]);
         }
+    }else{
+        for(i=0;i<6;i++) {
+            if(!skip_dct[i]){
+                int overflow;
+                s->block_last_index[i] = s->dct_quantize(s, s->block[i], i, s->qscale, &overflow);
+            // FIXME we could decide to change to quantizer instead of clipping
+            // JS: I don't think that would be a good idea it could lower quality instead
+            //     of improve it. Just INTRADC clipping deserves changes in quantizer
+                if (overflow) clip_coeffs(s, s->block[i], s->block_last_index[i]);
+            }else
+                s->block_last_index[i]= -1;
+        }
+        if(s->luma_elim_threshold && !s->mb_intra)
+            for(i=0; i<4; i++)
+                dct_single_coeff_elimination(s, i, s->luma_elim_threshold);
+        if(s->chroma_elim_threshold && !s->mb_intra)
+            for(i=4; i<6; i++)
+                dct_single_coeff_elimination(s, i, s->chroma_elim_threshold);
     }
 
-//    printf("%d %d\n", s->avg_mb_var, s->mc_mb_var);
+    if((s->flags&CODEC_FLAG_GRAY) && s->mb_intra){
+        s->block_last_index[4]=
+        s->block_last_index[5]= 0;
+        s->block[4][0]=
+        s->block[5][0]= 128;
+    }
 
-    if (!s->fixed_qscale) 
-        s->qscale = rate_estimate_qscale(s);
+    /* huffman encode */
+    switch(s->out_format) {
+    case FMT_MPEG1:
+        mpeg1_encode_mb(s, s->block, motion_x, motion_y);
+        break;
+    case FMT_H263:
+        if (s->h263_msmpeg4)
+            msmpeg4_encode_mb(s, s->block, motion_x, motion_y);
+        else if(s->h263_pred)
+            mpeg4_encode_mb(s, s->block, motion_x, motion_y);
+        else
+            h263_encode_mb(s, s->block, motion_x, motion_y);
+        break;
+    case FMT_MJPEG:
+        mjpeg_encode_mb(s, s->block);
+        break;
+    }
+}
 
-    /* precompute matrix */
+void ff_copy_bits(PutBitContext *pb, UINT8 *src, int length)
+{
+    int bytes= length>>4;
+    int bits= length&15;
+    int i;
+
+    if(length==0) return;
+
+    for(i=0; i<bytes; i++) put_bits(pb, 16, be2me_16(((uint16_t*)src)[i]));
+    put_bits(pb, bits, be2me_16(((uint16_t*)src)[i])>>(16-bits));
+}
+
+static inline void copy_context_before_encode(MpegEncContext *d, MpegEncContext *s, int type){
+    int i;
+
+    memcpy(d->last_mv, s->last_mv, 2*2*2*sizeof(int)); //FIXME is memcpy faster then a loop?
+
+    /* mpeg1 */
+    d->mb_incr= s->mb_incr;
+    for(i=0; i<3; i++)
+        d->last_dc[i]= s->last_dc[i];
+    
+    /* statistics */
+    d->mv_bits= s->mv_bits;
+    d->i_tex_bits= s->i_tex_bits;
+    d->p_tex_bits= s->p_tex_bits;
+    d->i_count= s->i_count;
+    d->f_count= s->f_count;
+    d->b_count= s->b_count;
+    d->skip_count= s->skip_count;
+    d->misc_bits= s->misc_bits;
+    d->last_bits= 0;
+
+    d->mb_skiped= s->mb_skiped;
+}
+
+static inline void copy_context_after_encode(MpegEncContext *d, MpegEncContext *s, int type){
+    int i;
+
+    memcpy(d->mv, s->mv, 2*4*2*sizeof(int)); 
+    memcpy(d->last_mv, s->last_mv, 2*2*2*sizeof(int)); //FIXME is memcpy faster then a loop?
+    
+    /* mpeg1 */
+    d->mb_incr= s->mb_incr;
+    for(i=0; i<3; i++)
+        d->last_dc[i]= s->last_dc[i];
+    
+    /* statistics */
+    d->mv_bits= s->mv_bits;
+    d->i_tex_bits= s->i_tex_bits;
+    d->p_tex_bits= s->p_tex_bits;
+    d->i_count= s->i_count;
+    d->f_count= s->f_count;
+    d->b_count= s->b_count;
+    d->skip_count= s->skip_count;
+    d->misc_bits= s->misc_bits;
+
+    d->mb_intra= s->mb_intra;
+    d->mb_skiped= s->mb_skiped;
+    d->mv_type= s->mv_type;
+    d->mv_dir= s->mv_dir;
+    d->pb= s->pb;
+    if(s->data_partitioning){
+        d->pb2= s->pb2;
+        d->tex_pb= s->tex_pb;
+    }
+    d->block= s->block;
+    for(i=0; i<6; i++)
+        d->block_last_index[i]= s->block_last_index[i];
+}
+
+static inline void encode_mb_hq(MpegEncContext *s, MpegEncContext *backup, MpegEncContext *best, int type, 
+                           PutBitContext pb[2], PutBitContext pb2[2], PutBitContext tex_pb[2],
+                           int *dmin, int *next_block, int motion_x, int motion_y)
+{
+    int bits_count;
+    
+    copy_context_before_encode(s, backup, type);
+
+    s->block= s->blocks[*next_block];
+    s->pb= pb[*next_block];
+    if(s->data_partitioning){
+        s->pb2   = pb2   [*next_block];
+        s->tex_pb= tex_pb[*next_block];
+    }
+
+    encode_mb(s, motion_x, motion_y);
+
+    bits_count= get_bit_count(&s->pb);
+    if(s->data_partitioning){
+        bits_count+= get_bit_count(&s->pb2);
+        bits_count+= get_bit_count(&s->tex_pb);
+    }
+
+    if(bits_count<*dmin){
+        *dmin= bits_count;
+        *next_block^=1;
+
+        copy_context_after_encode(best, s, type);
+    }
+}
+
+static void encode_picture(MpegEncContext *s, int picture_number)
+{
+    int mb_x, mb_y, last_gob, pdif = 0;
+    int i;
+    int bits;
+    MpegEncContext best_s, backup_s;
+    UINT8 bit_buf[2][3000];
+    UINT8 bit_buf2[2][3000];
+    UINT8 bit_buf_tex[2][3000];
+    PutBitContext pb[2], pb2[2], tex_pb[2];
+
+    for(i=0; i<2; i++){
+        init_put_bits(&pb    [i], bit_buf    [i], 3000, NULL, NULL);
+        init_put_bits(&pb2   [i], bit_buf2   [i], 3000, NULL, NULL);
+        init_put_bits(&tex_pb[i], bit_buf_tex[i], 3000, NULL, NULL);
+    }
+
+    s->picture_number = picture_number;
+
+    s->block_wrap[0]=
+    s->block_wrap[1]=
+    s->block_wrap[2]=
+    s->block_wrap[3]= s->mb_width*2 + 2;
+    s->block_wrap[4]=
+    s->block_wrap[5]= s->mb_width + 2;
+    
+    /* Reset the average MB variance */
+    s->mb_var_sum = 0;
+    s->mc_mb_var_sum = 0;
+
+    /* we need to initialize some time vars before we can encode b-frames */
+    if (s->h263_pred && !s->h263_msmpeg4)
+        ff_set_mpeg4_time(s, s->picture_number); 
+
+    s->scene_change_score=0;
+    
+    s->qscale= (int)(s->frame_qscale + 0.5); //FIXME qscale / ... stuff for ME ratedistoration
+
+    /* Estimate motion for every MB */
+    if(s->pict_type != I_TYPE){
+        for(mb_y=0; mb_y < s->mb_height; mb_y++) {
+            s->block_index[0]= s->block_wrap[0]*(mb_y*2 + 1) - 1;
+            s->block_index[1]= s->block_wrap[0]*(mb_y*2 + 1);
+            s->block_index[2]= s->block_wrap[0]*(mb_y*2 + 2) - 1;
+            s->block_index[3]= s->block_wrap[0]*(mb_y*2 + 2);
+            for(mb_x=0; mb_x < s->mb_width; mb_x++) {
+                s->mb_x = mb_x;
+                s->mb_y = mb_y;
+                s->block_index[0]+=2;
+                s->block_index[1]+=2;
+                s->block_index[2]+=2;
+                s->block_index[3]+=2;
+
+                /* compute motion vector & mb_type and store in context */
+                if(s->pict_type==B_TYPE)
+                    ff_estimate_b_frame_motion(s, mb_x, mb_y);
+                else
+                    ff_estimate_p_frame_motion(s, mb_x, mb_y);
+//                s->mb_type[mb_y*s->mb_width + mb_x]=MB_TYPE_INTER;
+            }
+        }
+        emms_c();
+    }else /* if(s->pict_type == I_TYPE) */{
+        /* I-Frame */
+        //FIXME do we need to zero them?
+        memset(s->motion_val[0], 0, sizeof(INT16)*(s->mb_width*2 + 2)*(s->mb_height*2 + 2)*2);
+        memset(s->p_mv_table   , 0, sizeof(INT16)*(s->mb_width+2)*(s->mb_height+2)*2);
+        memset(s->mb_type      , MB_TYPE_INTRA, sizeof(UINT8)*s->mb_width*s->mb_height);
+        
+        if(!s->fixed_qscale){
+            /* finding spatial complexity for I-frame rate control */
+            for(mb_y=0; mb_y < s->mb_height; mb_y++) {
+                for(mb_x=0; mb_x < s->mb_width; mb_x++) {
+                    int xx = mb_x * 16;
+                    int yy = mb_y * 16;
+                    uint8_t *pix = s->new_picture[0] + (yy * s->linesize) + xx;
+                    int varc;
+                    int sum = pix_sum(pix, s->linesize);
+    
+                    sum= (sum+8)>>4;
+                    varc = (pix_norm1(pix, s->linesize) - sum*sum + 500 + 128)>>8;
+
+                    s->mb_var [s->mb_width * mb_y + mb_x] = varc;
+                    s->mb_mean[s->mb_width * mb_y + mb_x] = (sum+7)>>4;
+                    s->mb_var_sum    += varc;
+                }
+            }
+        }
+    }
+    if(s->scene_change_score > 0 && s->pict_type == P_TYPE){
+        s->pict_type= I_TYPE;
+        memset(s->mb_type   , MB_TYPE_INTRA, sizeof(UINT8)*s->mb_width*s->mb_height);
+        if(s->max_b_frames==0){
+            s->input_pict_type= I_TYPE;
+            s->input_picture_in_gop_number=0;
+        }
+//printf("Scene change detected, encoding as I Frame %d %d\n", s->mb_var_sum, s->mc_mb_var_sum);
+    }
+    
+    if(s->pict_type==P_TYPE || s->pict_type==S_TYPE) 
+        s->f_code= ff_get_best_fcode(s, s->p_mv_table, MB_TYPE_INTER);
+        ff_fix_long_p_mvs(s);
+    if(s->pict_type==B_TYPE){
+        s->f_code= ff_get_best_fcode(s, s->b_forw_mv_table, MB_TYPE_FORWARD);
+        s->b_code= ff_get_best_fcode(s, s->b_back_mv_table, MB_TYPE_BACKWARD);
+
+        ff_fix_long_b_mvs(s, s->b_forw_mv_table, s->f_code, MB_TYPE_FORWARD);
+        ff_fix_long_b_mvs(s, s->b_back_mv_table, s->b_code, MB_TYPE_BACKWARD);
+        ff_fix_long_b_mvs(s, s->b_bidir_forw_mv_table, s->f_code, MB_TYPE_BIDIR);
+        ff_fix_long_b_mvs(s, s->b_bidir_back_mv_table, s->b_code, MB_TYPE_BIDIR);
+    }
+    
+    if (s->fixed_qscale) 
+        s->frame_qscale = s->avctx->quality;
+    else
+        s->frame_qscale = ff_rate_estimate_qscale(s);
+
+    if(s->adaptive_quant){
+        switch(s->codec_id){
+        case CODEC_ID_MPEG4:
+            ff_clean_mpeg4_qscales(s);
+            break;
+        case CODEC_ID_H263:
+        case CODEC_ID_H263P:
+            ff_clean_h263_qscales(s);
+            break;
+        }
+
+        s->qscale= s->qscale_table[0];
+    }else
+        s->qscale= (int)(s->frame_qscale + 0.5);
+        
     if (s->out_format == FMT_MJPEG) {
         /* for mjpeg, we do include qscale in the matrix */
-        s->intra_matrix[0] = default_intra_matrix[0];
-        for(i=1;i<64;i++)
-            s->intra_matrix[i] = (default_intra_matrix[i] * s->qscale) >> 3;
-        convert_matrix(s->q_intra_matrix, s->q_intra_matrix16, s->intra_matrix, 8);
-    } else {
-        convert_matrix(s->q_intra_matrix, s->q_intra_matrix16, s->intra_matrix, s->qscale);
-        convert_matrix(s->q_non_intra_matrix, s->q_non_intra_matrix16, s->non_intra_matrix, s->qscale);
+        s->intra_matrix[0] = ff_mpeg1_default_intra_matrix[0];
+        for(i=1;i<64;i++){
+            int j= s->idct_permutation[i];
+
+            s->intra_matrix[j] = CLAMP_TO_8BIT((ff_mpeg1_default_intra_matrix[i] * s->qscale) >> 3);
+        }
+        convert_matrix(s, s->q_intra_matrix, s->q_intra_matrix16, 
+                       s->q_intra_matrix16_bias, s->intra_matrix, s->intra_quant_bias, 8, 8);
     }
 
+    s->last_bits= get_bit_count(&s->pb);
     switch(s->out_format) {
     case FMT_MJPEG:
         mjpeg_picture_header(s);
@@ -1194,7 +2484,18 @@ static void encode_picture(MpegEncContext *s, int picture_number)
         mpeg1_encode_picture_header(s, picture_number);
         break;
     }
-        
+    bits= get_bit_count(&s->pb);
+    s->header_bits= bits - s->last_bits;
+    s->last_bits= bits;
+    s->mv_bits=0;
+    s->misc_bits=0;
+    s->i_tex_bits=0;
+    s->p_tex_bits=0;
+    s->i_count=0;
+    s->f_count=0;
+    s->b_count=0;
+    s->skip_count=0;
+
     /* init last dc values */
     /* note: quant matrix value (8) is implied here */
     s->last_dc[0] = 128;
@@ -1203,8 +2504,6 @@ static void encode_picture(MpegEncContext *s, int picture_number)
     s->mb_incr = 1;
     s->last_mv[0][0][0] = 0;
     s->last_mv[0][0][1] = 0;
-    s->mv_type = MV_TYPE_16X16;
-    s->mv_dir = MV_DIR_FORWARD;
 
     /* Get the GOB height based on picture height */
     if (s->out_format == FMT_H263 && !s->h263_pred && !s->h263_msmpeg4) {
@@ -1214,30 +2513,37 @@ static void encode_picture(MpegEncContext *s, int picture_number)
             s->gob_index = 2;
         else
             s->gob_index = 4;
+    }else if(s->codec_id==CODEC_ID_MPEG4){
+        s->gob_index = 1;
     }
-        
-    s->avg_mb_var = s->avg_mb_var / s->mb_num;        
-    
-    s->block_wrap[0]=
-    s->block_wrap[1]=
-    s->block_wrap[2]=
-    s->block_wrap[3]= s->mb_width*2 + 2;
-    s->block_wrap[4]=
-    s->block_wrap[5]= s->mb_width + 2;
+
+    if(s->codec_id==CODEC_ID_MPEG4 && s->data_partitioning && s->pict_type!=B_TYPE)
+        ff_mpeg4_init_partitions(s);
+
+    s->resync_mb_x=0;
+    s->resync_mb_y=0;
     for(mb_y=0; mb_y < s->mb_height; mb_y++) {
-        /* Put GOB header based on RTP MTU */
+        /* Put GOB header based on RTP MTU for formats which support it per line (H263*)*/
         /* TODO: Put all this stuff in a separate generic function */
         if (s->rtp_mode) {
             if (!mb_y) {
                 s->ptr_lastgob = s->pb.buf;
                 s->ptr_last_mb_line = s->pb.buf;
             } else if (s->out_format == FMT_H263 && !s->h263_pred && !s->h263_msmpeg4 && !(mb_y % s->gob_index)) {
+                // MN: we could move the space check from h263 -> here, as its not h263 specific
                 last_gob = h263_encode_gob_header(s, mb_y);
                 if (last_gob) {
-                    s->first_gob_line = 1;
+                    s->first_slice_line = 1;
+                }else{
+                    /*MN: we reset it here instead at the end of each line cuz mpeg4 can have 
+                          slice lines starting & ending in the middle*/
+                    s->first_slice_line = 0;
                 }
             }
         }
+
+        s->y_dc_scale= s->y_dc_scale_table[ s->qscale ];
+        s->c_dc_scale= s->c_dc_scale_table[ s->qscale ];
         
         s->block_index[0]= s->block_wrap[0]*(mb_y*2 + 1) - 1;
         s->block_index[1]= s->block_wrap[0]*(mb_y*2 + 1);
@@ -1246,6 +2552,10 @@ static void encode_picture(MpegEncContext *s, int picture_number)
         s->block_index[4]= s->block_wrap[4]*(mb_y + 1)                    + s->block_wrap[0]*(s->mb_height*2 + 2);
         s->block_index[5]= s->block_wrap[4]*(mb_y + 1 + s->mb_height + 2) + s->block_wrap[0]*(s->mb_height*2 + 2);
         for(mb_x=0; mb_x < s->mb_width; mb_x++) {
+            const int mb_type= s->mb_type[mb_y * s->mb_width + mb_x];
+            const int xy= (mb_y+1) * (s->mb_width+2) + mb_x + 1;
+//            int d;
+            int dmin=10000000;
 
             s->mb_x = mb_x;
             s->mb_y = mb_y;
@@ -1255,126 +2565,216 @@ static void encode_picture(MpegEncContext *s, int picture_number)
             s->block_index[3]+=2;
             s->block_index[4]++;
             s->block_index[5]++;
-#if 0
-            /* compute motion vector and macro block type (intra or non intra) */
-            motion_x = 0;
-            motion_y = 0;
-            if (s->pict_type == P_TYPE) {
-                s->mb_intra = estimate_motion(s, mb_x, mb_y,
-                                              &motion_x,
-                                              &motion_y);
-            } else {
-                s->mb_intra = 1;
-            }
-#endif
-
-            s->mb_intra = s->mb_type[mb_y * s->mb_width + mb_x];
-            motion_x = s->mv_table[0][mb_y * s->mb_width + mb_x];
-            motion_y = s->mv_table[1][mb_y * s->mb_width + mb_x];
             
-            /* get the pixels */
-            wrap = s->linesize;
-            ptr = s->new_picture[0] + (mb_y * 16 * wrap) + mb_x * 16;
-            get_pixels(s->block[0], ptr, wrap);
-            get_pixels(s->block[1], ptr + 8, wrap);
-            get_pixels(s->block[2], ptr + 8 * wrap, wrap);
-            get_pixels(s->block[3], ptr + 8 * wrap + 8, wrap);
-            wrap = s->linesize >> 1;
-            ptr = s->new_picture[1] + (mb_y * 8 * wrap) + mb_x * 8;
-            get_pixels(s->block[4], ptr, wrap);
-
-            wrap = s->linesize >> 1;
-            ptr = s->new_picture[2] + (mb_y * 8 * wrap) + mb_x * 8;
-            get_pixels(s->block[5], ptr, wrap);
-
-            /* subtract previous frame if non intra */
-            if (!s->mb_intra) {
-                int dxy, offset, mx, my;
+            /* write gob / video packet header for formats which support it at any MB (MPEG4) */
+            if(s->rtp_mode && s->mb_y>0 && s->codec_id==CODEC_ID_MPEG4){
+                int pdif= pbBufPtr(&s->pb) - s->ptr_lastgob;
+
+                //the *2 is there so we stay below the requested size
+                if(pdif + s->mb_line_avgsize/s->mb_width >= s->rtp_payload_size){ 
+                    if(s->codec_id==CODEC_ID_MPEG4){
+                        if(s->data_partitioning && s->pict_type!=B_TYPE){
+                            ff_mpeg4_merge_partitions(s);
+                            ff_mpeg4_init_partitions(s);
+                        }
+                        ff_mpeg4_encode_video_packet_header(s);
+
+                        if(s->flags&CODEC_FLAG_PASS1){
+                            int bits= get_bit_count(&s->pb);
+                            s->misc_bits+= bits - s->last_bits;
+                            s->last_bits= bits;
+                        }
+                        ff_mpeg4_clean_buffers(s);
+                    }
+                    s->ptr_lastgob = pbBufPtr(&s->pb);
+                    s->first_slice_line=1;
+                    s->resync_mb_x=mb_x;
+                    s->resync_mb_y=mb_y;
+                }
 
-                dxy = ((motion_y & 1) << 1) | (motion_x & 1);
-                ptr = s->last_picture[0] + 
-                    ((mb_y * 16 + (motion_y >> 1)) * s->linesize) + 
-                    (mb_x * 16 + (motion_x >> 1));
-
-                sub_pixels_2(s->block[0], ptr, s->linesize, dxy);
-                sub_pixels_2(s->block[1], ptr + 8, s->linesize, dxy);
-                sub_pixels_2(s->block[2], ptr + s->linesize * 8, s->linesize, dxy);
-                sub_pixels_2(s->block[3], ptr + 8 + s->linesize * 8, s->linesize ,dxy);
-
-                if (s->out_format == FMT_H263) {
-                    /* special rounding for h263 */
-                    dxy = 0;
-                    if ((motion_x & 3) != 0)
-                        dxy |= 1;
-                    if ((motion_y & 3) != 0)
-                        dxy |= 2;
-                    mx = motion_x >> 2;
-                    my = motion_y >> 2;
-                } else {
-                    mx = motion_x / 2;
-                    my = motion_y / 2;
-                    dxy = ((my & 1) << 1) | (mx & 1);
-                    mx >>= 1;
-                    my >>= 1;
+                if(  (s->resync_mb_x   == s->mb_x)
+                   && s->resync_mb_y+1 == s->mb_y){
+                    s->first_slice_line=0; 
                 }
-                offset = ((mb_y * 8 + my) * (s->linesize >> 1)) + (mb_x * 8 + mx);
-                ptr = s->last_picture[1] + offset;
-                sub_pixels_2(s->block[4], ptr, s->linesize >> 1, dxy);
-                ptr = s->last_picture[2] + offset;
-                sub_pixels_2(s->block[5], ptr, s->linesize >> 1, dxy);
             }
-            emms_c();
-            
-#if 0
-            {
-                float adap_parm;
+
+            if(mb_type & (mb_type-1)){ // more than 1 MB type possible
+                int next_block=0;
+                int pb_bits_count, pb2_bits_count, tex_pb_bits_count;
+
+                copy_context_before_encode(&backup_s, s, -1);
+                backup_s.pb= s->pb;
+                best_s.data_partitioning= s->data_partitioning;
+                if(s->data_partitioning){
+                    backup_s.pb2= s->pb2;
+                    backup_s.tex_pb= s->tex_pb;
+                }
+
+                if(mb_type&MB_TYPE_INTER){
+                    s->mv_dir = MV_DIR_FORWARD;
+                    s->mv_type = MV_TYPE_16X16;
+                    s->mb_intra= 0;
+                    s->mv[0][0][0] = s->p_mv_table[xy][0];
+                    s->mv[0][0][1] = s->p_mv_table[xy][1];
+                    encode_mb_hq(s, &backup_s, &best_s, MB_TYPE_INTER, pb, pb2, tex_pb, 
+                                 &dmin, &next_block, s->mv[0][0][0], s->mv[0][0][1]);
+                }
+                if(mb_type&MB_TYPE_INTER4V){                 
+                    s->mv_dir = MV_DIR_FORWARD;
+                    s->mv_type = MV_TYPE_8X8;
+                    s->mb_intra= 0;
+                    for(i=0; i<4; i++){
+                        s->mv[0][i][0] = s->motion_val[s->block_index[i]][0];
+                        s->mv[0][i][1] = s->motion_val[s->block_index[i]][1];
+                    }
+                    encode_mb_hq(s, &backup_s, &best_s, MB_TYPE_INTER4V, pb, pb2, tex_pb, 
+                                 &dmin, &next_block, 0, 0);
+                }
+                if(mb_type&MB_TYPE_FORWARD){
+                    s->mv_dir = MV_DIR_FORWARD;
+                    s->mv_type = MV_TYPE_16X16;
+                    s->mb_intra= 0;
+                    s->mv[0][0][0] = s->b_forw_mv_table[xy][0];
+                    s->mv[0][0][1] = s->b_forw_mv_table[xy][1];
+                    encode_mb_hq(s, &backup_s, &best_s, MB_TYPE_FORWARD, pb, pb2, tex_pb, 
+                                 &dmin, &next_block, s->mv[0][0][0], s->mv[0][0][1]);
+                }
+                if(mb_type&MB_TYPE_BACKWARD){
+                    s->mv_dir = MV_DIR_BACKWARD;
+                    s->mv_type = MV_TYPE_16X16;
+                    s->mb_intra= 0;
+                    s->mv[1][0][0] = s->b_back_mv_table[xy][0];
+                    s->mv[1][0][1] = s->b_back_mv_table[xy][1];
+                    encode_mb_hq(s, &backup_s, &best_s, MB_TYPE_BACKWARD, pb, pb2, tex_pb, 
+                                 &dmin, &next_block, s->mv[1][0][0], s->mv[1][0][1]);
+                }
+                if(mb_type&MB_TYPE_BIDIR){
+                    s->mv_dir = MV_DIR_FORWARD | MV_DIR_BACKWARD;
+                    s->mv_type = MV_TYPE_16X16;
+                    s->mb_intra= 0;
+                    s->mv[0][0][0] = s->b_bidir_forw_mv_table[xy][0];
+                    s->mv[0][0][1] = s->b_bidir_forw_mv_table[xy][1];
+                    s->mv[1][0][0] = s->b_bidir_back_mv_table[xy][0];
+                    s->mv[1][0][1] = s->b_bidir_back_mv_table[xy][1];
+                    encode_mb_hq(s, &backup_s, &best_s, MB_TYPE_BIDIR, pb, pb2, tex_pb, 
+                                 &dmin, &next_block, 0, 0);
+                }
+                if(mb_type&MB_TYPE_DIRECT){
+                    s->mv_dir = MV_DIR_FORWARD | MV_DIR_BACKWARD | MV_DIRECT;
+                    s->mv_type = MV_TYPE_16X16; //FIXME
+                    s->mb_intra= 0;
+                    s->mv[0][0][0] = s->b_direct_forw_mv_table[xy][0];
+                    s->mv[0][0][1] = s->b_direct_forw_mv_table[xy][1];
+                    s->mv[1][0][0] = s->b_direct_back_mv_table[xy][0];
+                    s->mv[1][0][1] = s->b_direct_back_mv_table[xy][1];
+                    encode_mb_hq(s, &backup_s, &best_s, MB_TYPE_DIRECT, pb, pb2, tex_pb, 
+                                 &dmin, &next_block, s->b_direct_mv_table[xy][0], s->b_direct_mv_table[xy][1]);
+                }
+                if(mb_type&MB_TYPE_INTRA){
+                    s->mv_dir = MV_DIR_FORWARD;
+                    s->mv_type = MV_TYPE_16X16;
+                    s->mb_intra= 1;
+                    s->mv[0][0][0] = 0;
+                    s->mv[0][0][1] = 0;
+                    encode_mb_hq(s, &backup_s, &best_s, MB_TYPE_INTRA, pb, pb2, tex_pb, 
+                                 &dmin, &next_block, 0, 0);
+                    /* force cleaning of ac/dc pred stuff if needed ... */
+                    if(s->h263_pred || s->h263_aic)
+                        s->mbintra_table[mb_x + mb_y*s->mb_width]=1;
+                }
+                copy_context_after_encode(s, &best_s, -1);
                 
-                adap_parm = ((s->avg_mb_var << 1) + s->mb_var[s->mb_width*mb_y+mb_x] + 1.0) /
-                            ((s->mb_var[s->mb_width*mb_y+mb_x] << 1) + s->avg_mb_var + 1.0);
-            
-                printf("\ntype=%c qscale=%2d adap=%0.2f dquant=%4.2f var=%4d avgvar=%4d", 
-                        (s->mb_type[s->mb_width*mb_y+mb_x] > 0) ? 'I' : 'P', 
-                        s->qscale, adap_parm, s->qscale*adap_parm,
-                        s->mb_var[s->mb_width*mb_y+mb_x], s->avg_mb_var);
-            }
-#endif
-            /* DCT & quantize */
-            if (s->h263_msmpeg4) {
-                msmpeg4_dc_scale(s);
-            } else if (s->h263_pred) {
-                h263_dc_scale(s);
+                pb_bits_count= get_bit_count(&s->pb);
+                flush_put_bits(&s->pb);
+                ff_copy_bits(&backup_s.pb, bit_buf[next_block^1], pb_bits_count);
+                s->pb= backup_s.pb;
+                
+                if(s->data_partitioning){
+                    pb2_bits_count= get_bit_count(&s->pb2);
+                    flush_put_bits(&s->pb2);
+                    ff_copy_bits(&backup_s.pb2, bit_buf2[next_block^1], pb2_bits_count);
+                    s->pb2= backup_s.pb2;
+                    
+                    tex_pb_bits_count= get_bit_count(&s->tex_pb);
+                    flush_put_bits(&s->tex_pb);
+                    ff_copy_bits(&backup_s.tex_pb, bit_buf_tex[next_block^1], tex_pb_bits_count);
+                    s->tex_pb= backup_s.tex_pb;
+                }
+                s->last_bits= get_bit_count(&s->pb);
             } else {
-                /* default quantization values */
-                s->y_dc_scale = 8;
-                s->c_dc_scale = 8;
-            }
-            for(i=0;i<6;i++) {
-                s->block_last_index[i] = dct_quantize(s, s->block[i], i, s->qscale);
+                int motion_x, motion_y;
+                s->mv_type=MV_TYPE_16X16;
+                // only one MB-Type possible
+                switch(mb_type){
+                case MB_TYPE_INTRA:
+                    s->mv_dir = MV_DIR_FORWARD;
+                    s->mb_intra= 1;
+                    motion_x= s->mv[0][0][0] = 0;
+                    motion_y= s->mv[0][0][1] = 0;
+                    break;
+                case MB_TYPE_INTER:
+                    s->mv_dir = MV_DIR_FORWARD;
+                    s->mb_intra= 0;
+                    motion_x= s->mv[0][0][0] = s->p_mv_table[xy][0];
+                    motion_y= s->mv[0][0][1] = s->p_mv_table[xy][1];
+                    break;
+                case MB_TYPE_INTER4V:
+                    s->mv_dir = MV_DIR_FORWARD;
+                    s->mv_type = MV_TYPE_8X8;
+                    s->mb_intra= 0;
+                    for(i=0; i<4; i++){
+                        s->mv[0][i][0] = s->motion_val[s->block_index[i]][0];
+                        s->mv[0][i][1] = s->motion_val[s->block_index[i]][1];
+                    }
+                    motion_x= motion_y= 0;
+                    break;
+                case MB_TYPE_DIRECT:
+                    s->mv_dir = MV_DIR_FORWARD | MV_DIR_BACKWARD | MV_DIRECT;
+                    s->mb_intra= 0;
+                    motion_x=s->b_direct_mv_table[xy][0];
+                    motion_y=s->b_direct_mv_table[xy][1];
+                    s->mv[0][0][0] = s->b_direct_forw_mv_table[xy][0];
+                    s->mv[0][0][1] = s->b_direct_forw_mv_table[xy][1];
+                    s->mv[1][0][0] = s->b_direct_back_mv_table[xy][0];
+                    s->mv[1][0][1] = s->b_direct_back_mv_table[xy][1];
+                    break;
+                case MB_TYPE_BIDIR:
+                    s->mv_dir = MV_DIR_FORWARD | MV_DIR_BACKWARD;
+                    s->mb_intra= 0;
+                    motion_x=0;
+                    motion_y=0;
+                    s->mv[0][0][0] = s->b_bidir_forw_mv_table[xy][0];
+                    s->mv[0][0][1] = s->b_bidir_forw_mv_table[xy][1];
+                    s->mv[1][0][0] = s->b_bidir_back_mv_table[xy][0];
+                    s->mv[1][0][1] = s->b_bidir_back_mv_table[xy][1];
+                    break;
+                case MB_TYPE_BACKWARD:
+                    s->mv_dir = MV_DIR_BACKWARD;
+                    s->mb_intra= 0;
+                    motion_x= s->mv[1][0][0] = s->b_back_mv_table[xy][0];
+                    motion_y= s->mv[1][0][1] = s->b_back_mv_table[xy][1];
+                    break;
+                case MB_TYPE_FORWARD:
+                    s->mv_dir = MV_DIR_FORWARD;
+                    s->mb_intra= 0;
+                    motion_x= s->mv[0][0][0] = s->b_forw_mv_table[xy][0];
+                    motion_y= s->mv[0][0][1] = s->b_forw_mv_table[xy][1];
+//                    printf(" %d %d ", motion_x, motion_y);
+                    break;
+                default:
+                    motion_x=motion_y=0; //gcc warning fix
+                    printf("illegal MB type\n");
+                }
+                encode_mb(s, motion_x, motion_y);
             }
-
-            /* huffman encode */
-            switch(s->out_format) {
-            case FMT_MPEG1:
-                mpeg1_encode_mb(s, s->block, motion_x, motion_y);
-                break;
-            case FMT_H263:
-                if (s->h263_msmpeg4)
-                    msmpeg4_encode_mb(s, s->block, motion_x, motion_y);
-                else if(s->h263_pred)
-                    mpeg4_encode_mb(s, s->block, motion_x, motion_y);
-                else
-                    h263_encode_mb(s, s->block, motion_x, motion_y);
-                break;
-            case FMT_MJPEG:
-                mjpeg_encode_mb(s, s->block);
-                break;
+            /* clean the MV table in IPS frames for direct mode in B frames */
+            if(s->mb_intra /* && I,P,S_TYPE */){
+                s->p_mv_table[xy][0]=0;
+                s->p_mv_table[xy][1]=0;
             }
 
-            /* decompress blocks so that we keep the state of the decoder */
-            s->mv[0][0][0] = motion_x;
-            s->mv[0][0][1] = motion_y;
-
             MPV_decode_mb(s, s->block);
+//printf("MB %d %d bits\n", s->mb_x+s->mb_y*s->mb_width, get_bit_count(&s->pb));
         }
 
 
@@ -1388,13 +2788,20 @@ static void encode_picture(MpegEncContext *s, int picture_number)
             }
             //fprintf(stderr, "\nMB line: %d\tSize: %u\tAvg. Size: %u", s->mb_y, 
             //                    (s->pb.buf_ptr - s->ptr_last_mb_line), s->mb_line_avgsize);
-            s->first_gob_line = 0;
+            if(s->codec_id!=CODEC_ID_MPEG4) s->first_slice_line = 0; //FIXME clean
         }
     }
-    
-    if (s->h263_msmpeg4 && s->pict_type == I_TYPE)
+    emms_c();
+
+    if(s->codec_id==CODEC_ID_MPEG4 && s->data_partitioning && s->pict_type!=B_TYPE)
+        ff_mpeg4_merge_partitions(s);
+
+    if (s->msmpeg4_version && s->msmpeg4_version<4 && s->pict_type == I_TYPE)
         msmpeg4_encode_ext_header(s);
 
+    if(s->codec_id==CODEC_ID_MPEG4) 
+        ff_mpeg4_stuffing(&s->pb);
+
     //if (s->gob_number)
     //    fprintf(stderr,"\nNumber of GOB: %d", s->gob_number);
     
@@ -1408,117 +2815,75 @@ static void encode_picture(MpegEncContext *s, int picture_number)
         s->ptr_lastgob = pbBufPtr(&s->pb);
         //fprintf(stderr,"\nGOB: %2d size: %d (last)", s->gob_number, pdif);
     }
-
 }
 
 static int dct_quantize_c(MpegEncContext *s, 
                         DCTELEM *block, int n,
-                        int qscale)
+                        int qscale, int *overflow)
 {
     int i, j, level, last_non_zero, q;
     const int *qmat;
-    int minLevel, maxLevel;
-
-    if(s->avctx!=NULL && s->avctx->codec->id==CODEC_ID_MPEG4){
-       /* mpeg4 */
-        minLevel= -2048;
-       maxLevel= 2047;
-    }else if(s->out_format==FMT_MPEG1){
-       /* mpeg1 */
-        minLevel= -255;
-       maxLevel= 255;
-    }else if(s->out_format==FMT_MJPEG){
-       /* (m)jpeg */
-        minLevel= -1023;
-       maxLevel= 1023;
-    }else{
-       /* h263 / msmpeg4 */
-        minLevel= -128;
-       maxLevel= 127;
-    }
-
-    av_fdct (block);
+    int bias;
+    int max=0;
+    unsigned int threshold1, threshold2;
+    
+    s->fdct (block);
 
+#ifndef ARCH_ALPHA              /* Alpha uses unpermuted matrix */
     /* we need this permutation so that we correct the IDCT
        permutation. will be moved into DCT code */
-    block_permute(block);
+    block_permute(block, s->idct_permutation); //FIXME remove
+#endif
 
     if (s->mb_intra) {
-        if (n < 4)
-            q = s->y_dc_scale;
-        else
-            q = s->c_dc_scale;
-        q = q << 3;
-        
+        if (!s->h263_aic) {
+            if (n < 4)
+                q = s->y_dc_scale;
+            else
+                q = s->c_dc_scale;
+            q = q << 3;
+        } else
+            /* For AIC we skip quant/dequant of INTRADC */
+            q = 1 << 3;
+            
         /* note: block[0] is assumed to be positive */
         block[0] = (block[0] + (q >> 1)) / q;
         i = 1;
         last_non_zero = 0;
-        if (s->out_format == FMT_H263) {
-            qmat = s->q_non_intra_matrix;
-        } else {
-            qmat = s->q_intra_matrix;
-        }
+        qmat = s->q_intra_matrix[qscale];
+        bias= s->intra_quant_bias<<(QMAT_SHIFT - QUANT_BIAS_SHIFT);
     } else {
         i = 0;
         last_non_zero = -1;
-        qmat = s->q_non_intra_matrix;
+        qmat = s->q_inter_matrix[qscale];
+        bias= s->inter_quant_bias<<(QMAT_SHIFT - QUANT_BIAS_SHIFT);
     }
+    threshold1= (1<<QMAT_SHIFT) - bias - 1;
+    threshold2= (threshold1<<1);
 
     for(;i<64;i++) {
-        j = zigzag_direct[i];
+        j = s->intra_scantable.permutated[i];
         level = block[j];
         level = level * qmat[j];
-#ifdef PARANOID
-        {
-            static int count = 0;
-            int level1, level2, qmat1;
-            double val;
-            if (qmat == s->q_non_intra_matrix) {
-                qmat1 = default_non_intra_matrix[j] * s->qscale;
-            } else {
-                qmat1 = default_intra_matrix[j] * s->qscale;
-            }
-            if (av_fdct != jpeg_fdct_ifast)
-                val = ((double)block[j] * 8.0) / (double)qmat1;
-            else
-                val = ((double)block[j] * 8.0 * 2048.0) / 
-                    ((double)qmat1 * aanscales[j]);
-            level1 = (int)val;
-            level2 = level / (1 << (QMAT_SHIFT - 3));
-            if (level1 != level2) {
-                fprintf(stderr, "%d: quant error qlevel=%d wanted=%d level=%d qmat1=%d qmat=%d wantedf=%0.6f\n", 
-                        count, level2, level1, block[j], qmat1, qmat[j],
-                        val);
-                count++;
-            }
 
-        }
-#endif
-        /* XXX: slight error for the low range. Test should be equivalent to
-           (level <= -(1 << (QMAT_SHIFT - 3)) || level >= (1 <<
-           (QMAT_SHIFT - 3)))
-        */
-        if (((level << (31 - (QMAT_SHIFT - 3))) >> (31 - (QMAT_SHIFT - 3))) != 
-            level) {
-            level = level / (1 << (QMAT_SHIFT - 3));
-            /* XXX: currently, this code is not optimal. the range should be:
-               mpeg1: -255..255
-               mpeg2: -2048..2047
-               h263:  -128..127
-               mpeg4: -2048..2047
-            */
-            if (level > maxLevel)
-                level = maxLevel;
-            else if (level < minLevel)
-                level = minLevel;
-
-            block[j] = level;
+//        if(   bias+level >= (1<<(QMAT_SHIFT - 3))
+//           || bias-level >= (1<<(QMAT_SHIFT - 3))){
+        if(((unsigned)(level+threshold1))>threshold2){
+            if(level>0){
+                level= (bias + level)>>QMAT_SHIFT;
+                block[j]= level;
+            }else{
+                level= (bias - level)>>QMAT_SHIFT;
+                block[j]= -level;
+            }
+            max |=level;
             last_non_zero = i;
-        } else {
-            block[j] = 0;
+        }else{
+            block[j]=0;
         }
     }
+    *overflow= s->max_qcoeff < max; //overflow might have happend
+    
     return last_non_zero;
 }
 
@@ -1528,8 +2893,7 @@ static void dct_unquantize_mpeg1_c(MpegEncContext *s,
     int i, level, nCoeffs;
     const UINT16 *quant_matrix;
 
-    if(s->alternate_scan) nCoeffs= 64;
-    else nCoeffs= s->block_last_index[n]+1;
+    nCoeffs= s->block_last_index[n];
     
     if (s->mb_intra) {
         if (n < 4) 
@@ -1538,8 +2902,8 @@ static void dct_unquantize_mpeg1_c(MpegEncContext *s,
             block[0] = block[0] * s->c_dc_scale;
         /* XXX: only mpeg1 */
         quant_matrix = s->intra_matrix;
-        for(i=1;i<nCoeffs;i++) {
-            int j= zigzag_direct[i];
+        for(i=1;i<=nCoeffs;i++) {
+            int j= s->intra_scantable.permutated[i];
             level = block[j];
             if (level) {
                 if (level < 0) {
@@ -1560,9 +2924,9 @@ static void dct_unquantize_mpeg1_c(MpegEncContext *s,
         }
     } else {
         i = 0;
-        quant_matrix = s->non_intra_matrix;
-        for(;i<nCoeffs;i++) {
-            int j= zigzag_direct[i];
+        quant_matrix = s->inter_matrix;
+        for(;i<=nCoeffs;i++) {
+            int j= s->intra_scantable.permutated[i];
             level = block[j];
             if (level) {
                 if (level < 0) {
@@ -1586,33 +2950,96 @@ static void dct_unquantize_mpeg1_c(MpegEncContext *s,
     }
 }
 
+static void dct_unquantize_mpeg2_c(MpegEncContext *s, 
+                                   DCTELEM *block, int n, int qscale)
+{
+    int i, level, nCoeffs;
+    const UINT16 *quant_matrix;
+
+    if(s->alternate_scan) nCoeffs= 63;
+    else nCoeffs= s->block_last_index[n];
+    
+    if (s->mb_intra) {
+        if (n < 4) 
+            block[0] = block[0] * s->y_dc_scale;
+        else
+            block[0] = block[0] * s->c_dc_scale;
+        quant_matrix = s->intra_matrix;
+        for(i=1;i<=nCoeffs;i++) {
+            int j= s->intra_scantable.permutated[i];
+            level = block[j];
+            if (level) {
+                if (level < 0) {
+                    level = -level;
+                    level = (int)(level * qscale * quant_matrix[j]) >> 3;
+                    level = -level;
+                } else {
+                    level = (int)(level * qscale * quant_matrix[j]) >> 3;
+                }
+#ifdef PARANOID
+                if (level < -2048 || level > 2047)
+                    fprintf(stderr, "unquant error %d %d\n", i, level);
+#endif
+                block[j] = level;
+            }
+        }
+    } else {
+        int sum=-1;
+        i = 0;
+        quant_matrix = s->inter_matrix;
+        for(;i<=nCoeffs;i++) {
+            int j= s->intra_scantable.permutated[i];
+            level = block[j];
+            if (level) {
+                if (level < 0) {
+                    level = -level;
+                    level = (((level << 1) + 1) * qscale *
+                             ((int) (quant_matrix[j]))) >> 4;
+                    level = -level;
+                } else {
+                    level = (((level << 1) + 1) * qscale *
+                             ((int) (quant_matrix[j]))) >> 4;
+                }
+#ifdef PARANOID
+                if (level < -2048 || level > 2047)
+                    fprintf(stderr, "unquant error %d %d\n", i, level);
+#endif
+                block[j] = level;
+                sum+=level;
+            }
+        }
+        block[63]^=sum&1;
+    }
+}
+
+
 static void dct_unquantize_h263_c(MpegEncContext *s, 
                                   DCTELEM *block, int n, int qscale)
 {
     int i, level, qmul, qadd;
     int nCoeffs;
     
+    assert(s->block_last_index[n]>=0);
+    
+    qadd = (qscale - 1) | 1;
+    qmul = qscale << 1;
+    
     if (s->mb_intra) {
         if (!s->h263_aic) {
             if (n < 4) 
                 block[0] = block[0] * s->y_dc_scale;
             else
                 block[0] = block[0] * s->c_dc_scale;
-        }
+        }else
+            qadd = 0;
         i = 1;
-        nCoeffs= 64; //does not allways use zigzag table 
+        nCoeffs= 63; //does not allways use zigzag table 
     } else {
         i = 0;
-        nCoeffs= zigzag_end[ s->block_last_index[n] ];
+        nCoeffs= s->intra_scantable.raster_end[ s->block_last_index[n] ];
     }
 
-    qmul = s->qscale << 1;
-    if (s->h263_aic && s->mb_intra)
-        qadd = 0;
-    else
-        qadd = (s->qscale - 1) | 1;
-
-    for(;i<nCoeffs;i++) {
+    for(;i<=nCoeffs;i++) {
         level = block[i];
         if (level) {
             if (level < 0) {
@@ -1629,188 +3056,117 @@ static void dct_unquantize_h263_c(MpegEncContext *s,
     }
 }
 
-/* rate control */
-
-/* an I frame is I_FRAME_SIZE_RATIO bigger than a P frame */
-#define I_FRAME_SIZE_RATIO 3.0
-#define QSCALE_K           20
-
-static void rate_control_init(MpegEncContext *s)
+static void remove_ac(MpegEncContext *s, uint8_t *dest_y, uint8_t *dest_cb, uint8_t *dest_cr, int mb_x, int mb_y)
 {
-#if 1
-    emms_c();
-
-    //initial values, they dont really matter as they will be totally different within a few frames
-    s->i_pred.coeff= s->p_pred.coeff= 7.0;
-    s->i_pred.count= s->p_pred.count= 1.0;
-    
-    s->i_pred.decay= s->p_pred.decay= 0.4;
-    
-    // use more bits at the beginning, otherwise high motion at the begin will look like shit
-    s->qsum=100;
-    s->qcount=100;
-
-    s->short_term_qsum=0.001;
-    s->short_term_qcount=0.001;
-#else
-    s->wanted_bits = 0;
-
-    if (s->intra_only) {
-        s->I_frame_bits = ((INT64)s->bit_rate * FRAME_RATE_BASE) / s->frame_rate;
-        s->P_frame_bits = s->I_frame_bits;
-    } else {
-        s->P_frame_bits = (int) ((float)(s->gop_size * s->bit_rate) / 
-                                 (float)((float)s->frame_rate / FRAME_RATE_BASE * (I_FRAME_SIZE_RATIO + s->gop_size - 1)));
-        s->I_frame_bits = (int)(s->P_frame_bits * I_FRAME_SIZE_RATIO);
+    int dc, dcb, dcr, y, i;
+    for(i=0; i<4; i++){
+        dc= s->dc_val[0][mb_x*2+1 + (i&1) + (mb_y*2+1 + (i>>1))*(s->mb_width*2+2)];
+        for(y=0; y<8; y++){
+            int x;
+            for(x=0; x<8; x++){
+                dest_y[x + (i&1)*8 + (y + (i>>1)*8)*s->linesize]= dc/8;
+            }
+        }
+    }
+    dcb = s->dc_val[1][mb_x+1 + (mb_y+1)*(s->mb_width+2)];
+    dcr= s->dc_val[2][mb_x+1 + (mb_y+1)*(s->mb_width+2)];
+    for(y=0; y<8; y++){
+        int x;
+        for(x=0; x<8; x++){
+            dest_cb[x + y*(s->uvlinesize)]= dcb/8;
+            dest_cr[x + y*(s->uvlinesize)]= dcr/8;
+        }
     }
-
-#if defined(DEBUG)
-    printf("I_frame_size=%d P_frame_size=%d\n",
-           s->I_frame_bits, s->P_frame_bits);
-#endif
-#endif
-}
-
-static double predict(Predictor *p, double q, double var)
-{
-    return p->coeff*var / (q*p->count);
 }
 
-static void update_predictor(Predictor *p, double q, double var, double size)
+/**
+ * will conceal past errors, and allso drop b frames if needed
+ *
+ */
+void ff_conceal_past_errors(MpegEncContext *s, int unknown_pos)
 {
-    double new_coeff= size*q / (var + 1);
-    if(var<1000) return;
-/*{
-int pred= predict(p, q, var);
-int error= abs(pred-size);
-static double sum=0;
-static int count=0;
-if(count>5) sum+=error;
-count++;
-if(256*256*256*64%count==0){
-    printf("%d %f %f\n", count, sum/count, p->coeff);
-}
-}*/
-    p->count*= p->decay;
-    p->coeff*= p->decay;
-    p->count++;
-    p->coeff+= new_coeff;
-}
+    int mb_x= s->mb_x;
+    int mb_y= s->mb_y;
+    int mb_dist=0;
+    int i, intra_count=0, inter_count=0;
+    int intra_conceal= s->msmpeg4_version ? 50 : 50; //FIXME finetune
+    int inter_conceal= s->msmpeg4_version ? 50 : 50;
+
+    // for last block
+    if(mb_x>=s->mb_width)  mb_x= s->mb_width -1;
+    if(mb_y>=s->mb_height) mb_y= s->mb_height-1;
+
+    if(s->decoding_error==0 && unknown_pos){
+        if(s->data_partitioning && s->pict_type!=B_TYPE)
+                s->decoding_error= DECODING_AC_LOST;
+        else
+                s->decoding_error= DECODING_DESYNC;
+    }
 
-static int rate_estimate_qscale(MpegEncContext *s)
-{
-#if 1
-    int qmin= s->qmin;
-    int qmax= s->qmax;
-    int rate_q=5;
-    float q;
-    int qscale;
-    float br_compensation;
-    double diff;
-    double short_term_q;
-    double long_term_q;
-    int last_qscale= s->qscale;
-    double fps;
-    INT64 wanted_bits;
-    emms_c();
-    
-    fps= (double)s->frame_rate / FRAME_RATE_BASE;
-    wanted_bits= s->bit_rate*(double)s->picture_number/fps;
+    if(s->decoding_error==DECODING_DESYNC && s->pict_type!=B_TYPE) s->next_p_frame_damaged=1;
 
+    for(i=mb_x + mb_y*s->mb_width; i>=0; i--){
+        if(s->mbintra_table[i]) intra_count++;
+        else                    inter_count++;
+    }
     
-    if(s->picture_number>2){
-        /* update predictors */
-        if(s->last_pict_type == I_TYPE){
-        //FIXME
-        }else{ //P Frame
-//printf("%d %d %d %f\n", s->qscale, s->last_mc_mb_var, s->frame_bits, s->p_pred.coeff);
-            update_predictor(&s->p_pred, s->qscale, s->last_mc_mb_var, s->frame_bits);
-        }
+    if(s->decoding_error==DECODING_AC_LOST){
+        intra_conceal*=2;
+        inter_conceal*=2;
+    }else if(s->decoding_error==DECODING_ACDC_LOST){
+        intra_conceal*=2;
+        inter_conceal*=2;
     }
 
-    if(s->pict_type == I_TYPE){
-        //FIXME
-        rate_q= s->qsum/s->qcount;
-    }else{ //P Frame
-        int i;
-        int diff, best_diff=1000000000;
-        for(i=1; i<=31; i++){
-            diff= predict(&s->p_pred, i, s->mc_mb_var) - (double)s->bit_rate/fps;
-            if(diff<0) diff= -diff;
-            if(diff<best_diff){
-                best_diff= diff;
-                rate_q= i;
-            }
-        }
+    if(unknown_pos && (intra_count<inter_count)){
+        intra_conceal= inter_conceal= s->mb_num; 
+//        printf("%d %d\n",intra_count, inter_count);
     }
 
-    s->short_term_qsum*=s->qblur;
-    s->short_term_qcount*=s->qblur;
-
-    s->short_term_qsum+= rate_q;
-    s->short_term_qcount++;
-    short_term_q= s->short_term_qsum/s->short_term_qcount;
-    
-    long_term_q= s->qsum/s->qcount*s->total_bits/wanted_bits;
-
-//    q= (long_term_q - short_term_q)*s->qcompress + short_term_q;
-    q= 1/((1/long_term_q - 1/short_term_q)*s->qcompress + 1/short_term_q);
-
-    diff= s->total_bits - wanted_bits;
-    br_compensation= (s->bit_rate_tolerance - diff)/s->bit_rate_tolerance;
-    if(br_compensation<=0.0) br_compensation=0.001;
-    q/=br_compensation;
-
-    qscale= (int)(q + 0.5);
-    if     (qscale<qmin) qscale=qmin;
-    else if(qscale>qmax) qscale=qmax;
-    
-    if     (qscale<last_qscale-s->max_qdiff) qscale=last_qscale-s->max_qdiff;
-    else if(qscale>last_qscale+s->max_qdiff) qscale=last_qscale+s->max_qdiff;
+    fprintf(stderr, "concealing errors\n");
+
+    /* for all MBs from the current one back until the last resync marker */
+    for(; mb_y>=0 && mb_y>=s->resync_mb_y; mb_y--){
+        for(; mb_x>=0; mb_x--){
+            uint8_t *dest_y  = s->current_picture[0] + (mb_y * 16*  s->linesize      ) + mb_x * 16;
+            uint8_t *dest_cb = s->current_picture[1] + (mb_y * 8 * (s->uvlinesize)) + mb_x * 8;
+            uint8_t *dest_cr = s->current_picture[2] + (mb_y * 8 * (s->uvlinesize)) + mb_x * 8;
+            int mb_x_backup= s->mb_x; //FIXME pass xy to mpeg_motion
+            int mb_y_backup= s->mb_y;
+            s->mb_x=mb_x;
+            s->mb_y=mb_y;
+            if(s->mbintra_table[mb_y*s->mb_width + mb_x] && mb_dist<intra_conceal){
+                if(s->decoding_error==DECODING_AC_LOST){
+                    remove_ac(s, dest_y, dest_cb, dest_cr, mb_x, mb_y);
+//                    printf("remove ac to %d %d\n", mb_x, mb_y);
+                }else{
+                    mpeg_motion(s, dest_y, dest_cb, dest_cr, 0, 
+                                s->last_picture, 0, 0, put_pixels_tab,
+                                0/*mx*/, 0/*my*/, 16);
+                }
+            }
+            else if(!s->mbintra_table[mb_y*s->mb_width + mb_x] && mb_dist<inter_conceal){
+                int mx=0;
+                int my=0;
+
+                if(s->decoding_error!=DECODING_DESYNC){
+                    int xy= mb_x*2+1 + (mb_y*2+1)*(s->mb_width*2+2);
+                    mx= s->motion_val[ xy ][0];
+                    my= s->motion_val[ xy ][1];
+                }
 
-    s->qsum+= qscale;
-    s->qcount++;
+                mpeg_motion(s, dest_y, dest_cb, dest_cr, 0, 
+                            s->last_picture, 0, 0, put_pixels_tab,
+                            mx, my, 16);
+            }
+            s->mb_x= mb_x_backup;
+            s->mb_y= mb_y_backup;
 
-    s->last_pict_type= s->pict_type;
-//printf("q:%d diff:%d comp:%f rate_q:%d st_q:%d fvar:%d last_size:%d\n", qscale, (int)diff, br_compensation, 
-//       rate_q, (int)short_term_q, s->mc_mb_var, s->frame_bits);
-//printf("%d %d\n", s->bit_rate, (int)fps);
-    return qscale;
-#else
-    INT64 diff, total_bits = s->total_bits;
-    float q;
-    int qscale;
-    if (s->pict_type == I_TYPE) {
-        s->wanted_bits += s->I_frame_bits;
-    } else {
-        s->wanted_bits += s->P_frame_bits;
-    }
-    diff = s->wanted_bits - total_bits;
-    q = 31.0 - (float)diff / (QSCALE_K * s->mb_height * s->mb_width);
-    /* adjust for I frame */
-    if (s->pict_type == I_TYPE && !s->intra_only) {
-        q /= I_FRAME_SIZE_RATIO;
-    }
-
-    /* using a too small Q scale leeds to problems in mpeg1 and h263
-       because AC coefficients are clamped to 255 or 127 */
-    qmin = 3;
-    if (q < qmin)
-        q = qmin;
-    else if (q > 31)
-        q = 31;
-    qscale = (int)(q + 0.5);
-#if defined(DEBUG)
-    printf("\n%d: total=%0.0f wanted=%0.0f br=%0.1f diff=%d qest=%2.1f\n", 
-           s->picture_number, 
-           (double)total_bits, 
-           (double)s->wanted_bits,
-           (float)s->frame_rate / FRAME_RATE_BASE * 
-           total_bits / s->picture_number, 
-           (int)diff, q);
-#endif
-    return qscale;
-#endif
+            if(mb_x== s->resync_mb_x && mb_y== s->resync_mb_y) return;
+            if(!s->mbskip_table[mb_x + mb_y*s->mb_width]) mb_dist++;
+        }
+        mb_x=s->mb_width-1;
+    }
 }
 
 AVCodec mpeg1video_encoder = {
@@ -1873,10 +3229,50 @@ AVCodec mpeg4_encoder = {
     MPV_encode_end,
 };
 
-AVCodec msmpeg4_encoder = {
+AVCodec msmpeg4v1_encoder = {
+    "msmpeg4v1",
+    CODEC_TYPE_VIDEO,
+    CODEC_ID_MSMPEG4V1,
+    sizeof(MpegEncContext),
+    MPV_encode_init,
+    MPV_encode_picture,
+    MPV_encode_end,
+};
+
+AVCodec msmpeg4v2_encoder = {
+    "msmpeg4v2",
+    CODEC_TYPE_VIDEO,
+    CODEC_ID_MSMPEG4V2,
+    sizeof(MpegEncContext),
+    MPV_encode_init,
+    MPV_encode_picture,
+    MPV_encode_end,
+};
+
+AVCodec msmpeg4v3_encoder = {
     "msmpeg4",
     CODEC_TYPE_VIDEO,
-    CODEC_ID_MSMPEG4,
+    CODEC_ID_MSMPEG4V3,
+    sizeof(MpegEncContext),
+    MPV_encode_init,
+    MPV_encode_picture,
+    MPV_encode_end,
+};
+
+AVCodec wmv1_encoder = {
+    "wmv1",
+    CODEC_TYPE_VIDEO,
+    CODEC_ID_WMV1,
+    sizeof(MpegEncContext),
+    MPV_encode_init,
+    MPV_encode_picture,
+    MPV_encode_end,
+};
+
+AVCodec wmv2_encoder = {
+    "wmv2",
+    CODEC_TYPE_VIDEO,
+    CODEC_ID_WMV2,
     sizeof(MpegEncContext),
     MPV_encode_init,
     MPV_encode_picture,