generic-library/vpx
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Fix non-4:2:0 chroma MV calculation for SPLITMV

Browse Source 
The previous code was somewhat vestigial for 16x16 MI units, but was
incorrect when called with chroma blocks larger than 4x4 because the
block index caused a reference to a non-existent BMI. This patch uses
the same MV for all chroma subblocks in SPLITMV mode, which is
suboptimal for non-4:2:0 subsamplings, but as SPLITMV may be removed
in the near future, will use this as a stop gap.

Change-Id: I3211cee5ccf1cfb426e5eef5353b0ce5bb92b4cd
This commit is contained in:
John Koleszar 2013-05-08 16:14:14 -07:00 committed by Gerrit Code Review
parent 6dab4d205e
commit a37ee9d2e8
1 changed files with 16 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

32
vp9/common/vp9_reconinter.c
View File

@ -257,22 +257,19 @@ static INLINE int round_mv_comp_q4(int value) {
return (value < 0 ? value - 2 : value + 2) / 4;
}
#define IDX1 2
#define IDX2 3
static int mi_mv_pred_row_q4(MACROBLOCKD *mb, int off, int idx) {
const int temp = mb->mode_info_context->bmi[off + 0].as_mv[idx].as_mv.row +
mb->mode_info_context->bmi[off + 1].as_mv[idx].as_mv.row +
mb->mode_info_context->bmi[off + IDX1].as_mv[idx].as_mv.row +
mb->mode_info_context->bmi[off + IDX2].as_mv[idx].as_mv.row;
static int mi_mv_pred_row_q4(MACROBLOCKD *mb, int idx) {
const int temp = mb->mode_info_context->bmi[0].as_mv[idx].as_mv.row +
mb->mode_info_context->bmi[1].as_mv[idx].as_mv.row +
mb->mode_info_context->bmi[2].as_mv[idx].as_mv.row +
mb->mode_info_context->bmi[3].as_mv[idx].as_mv.row;
return round_mv_comp_q4(temp);
}
static int mi_mv_pred_col_q4(MACROBLOCKD *mb, int off, int idx) {
const int temp = mb->mode_info_context->bmi[off + 0].as_mv[idx].as_mv.col +
mb->mode_info_context->bmi[off + 1].as_mv[idx].as_mv.col +
mb->mode_info_context->bmi[off + IDX1].as_mv[idx].as_mv.col +
mb->mode_info_context->bmi[off + IDX2].as_mv[idx].as_mv.col;
static int mi_mv_pred_col_q4(MACROBLOCKD *mb, int idx) {
const int temp = mb->mode_info_context->bmi[0].as_mv[idx].as_mv.col +
mb->mode_info_context->bmi[1].as_mv[idx].as_mv.col +
mb->mode_info_context->bmi[2].as_mv[idx].as_mv.col +
mb->mode_info_context->bmi[3].as_mv[idx].as_mv.col;
return round_mv_comp_q4(temp);
}
@ -351,9 +348,12 @@ static void build_inter_predictors(int plane, int block,
if (plane == 0) {
mv = &xd->mode_info_context->bmi[block].as_mv[which_mv].as_mv;
} else {
const int y_block = (block & 2) * 4 + (block & 1) * 2;
split_chroma_mv.row = mi_mv_pred_row_q4(xd, y_block, which_mv);
split_chroma_mv.col = mi_mv_pred_col_q4(xd, y_block, which_mv);
// TODO(jkoleszar): All chroma MVs in SPLITMV mode are taken as the
// same MV (the average of the 4 luma MVs) but we could do something
// smarter for non-4:2:0. Just punt for now, pending the changes to get
// rid of SPLITMV mode entirely.
split_chroma_mv.row = mi_mv_pred_row_q4(xd, which_mv);
split_chroma_mv.col = mi_mv_pred_col_q4(xd, which_mv);
mv = &split_chroma_mv;
}
} else {