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

Bug fix in motion vector limit calculation

Browse Source 
Motion vector limits are calculated using right shifts, which
could give wrong results for negative numbers. James Berry's
test on one clip showed encoder produced some artifacts. This
change fixed that.

Change-Id: I035fc02280b10455b7f6eb388f7c2e33b796b018
This commit is contained in:
Yunqing Wang
2011-06-30 11:20:13 -04:00
parent 034cea5e72
commit b748045470
4 changed files with 35 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vp8/encoder/mcomp.h
View File

@@ -22,7 +22,7 @@ extern void accum_mv_refs(MB_PREDICTION_MODE, const int near_mv_ref_cts[4]);
#define MAX_MVSEARCH_STEPS 8 // The maximum number of steps in a step search given the largest allowed initial step
#define MAX_FULL_PEL_VAL ((1 << (MAX_MVSEARCH_STEPS+3)) - 8) // Max full pel mv specified in 1/8 pel units
#define MAX_FULL_PEL_VAL ((1 << (MAX_MVSEARCH_STEPS)) - 1) // Max full pel mv specified in 1 pel units
#define MAX_FIRST_STEP (1 << (MAX_MVSEARCH_STEPS-1)) // Maximum size of the first step in full pel units
extern void print_mode_context(void);