* XXX should be optimized and moved to DSPContext
* TODO handle out of edge ME
*/
-static inline int block_cmp(uint8_t *src, int stride, uint8_t *src2, int stride2, int bw, int bh)
+static inline int block_cmp(uint8_t *src, int stride, uint8_t *src2, int stride2, int bw, int bh, int *xored)
{
int sum = 0;
int i, j;
uint8_t histogram[256]={0};
+ *xored = 0;
for(j = 0; j < bh; j++){
- for(i = 0; i < bw; i++)
- histogram[src[i] ^ src2[i]]++;
+ for(i = 0; i < bw; i++){
+ int t = src[i] ^ src2[i];
+ histogram[t]++;
+ *xored |= t;
+ }
src += stride;
src2 += stride2;
}
* TODO make better ME decisions
*/
static int zmbv_me(ZmbvEncContext *c, uint8_t *src, int sstride, uint8_t *prev, int pstride,
- int x, int y, int *mx, int *my)
+ int x, int y, int *mx, int *my, int *xored)
{
int dx, dy, tx, ty, tv, bv, bw, bh;
*mx = *my = 0;
bw = FFMIN(ZMBV_BLOCK, c->avctx->width - x);
bh = FFMIN(ZMBV_BLOCK, c->avctx->height - y);
- bv = block_cmp(src, sstride, prev, pstride, bw, bh);
+ bv = block_cmp(src, sstride, prev, pstride, bw, bh, xored);
if(!bv) return 0;
for(ty = FFMAX(y - c->range, 0); ty < FFMIN(y + c->range, c->avctx->height - bh); ty++){
for(tx = FFMAX(x - c->range, 0); tx < FFMIN(x + c->range, c->avctx->width - bw); tx++){
if(tx == x && ty == y) continue; // we already tested this block
dx = tx - x;
dy = ty - y;
- tv = block_cmp(src, sstride, prev + dx + dy*pstride, pstride, bw, bh);
+ tv = block_cmp(src, sstride, prev + dx + dy*pstride, pstride, bw, bh, xored);
if(tv < bv){
bv = tv;
*mx = dx;
work_size += avctx->width;
}
}else{
- int x, y, bh2, bw2;
+ int x, y, bh2, bw2, xored;
uint8_t *tsrc, *tprev;
uint8_t *mv;
int mx, my, bv;
tsrc = src + x;
tprev = prev + x;
- bv = zmbv_me(c, tsrc, p->linesize[0], tprev, c->pstride, x, y, &mx, &my);
- mv[0] = (mx << 1) | !!bv;
+ bv = zmbv_me(c, tsrc, p->linesize[0], tprev, c->pstride, x, y, &mx, &my, &xored);
+ mv[0] = (mx << 1) | !!xored;
mv[1] = my << 1;
tprev += mx + my * c->pstride;
- if(bv){
+ if(xored){
for(j = 0; j < bh2; j++){
for(i = 0; i < bw2; i++)
c->work_buf[work_size++] = tsrc[i] ^ tprev[i];