| 2008-07-04 |
Fiona Glaser | Update file headers throughout x264 |
blob | commitdiff | raw |
| 2008-07-03 |
Fiona Glaser | denoise_dct asm |
blob | commitdiff | raw | diff to current |
| 2008-07-02 |
Fiona Glaser | Optimizations and cosmetics in macroblock.c |
blob | commitdiff | raw | diff to current |
| 2008-06-24 |
Fiona Glaser | Convert NNZ to raster order and other optimizations |
blob | commitdiff | raw | diff to current |
| 2008-06-03 |
Fiona Glaser | omit redundant mc after non-rdo dct size decision,... |
blob | commitdiff | raw | diff to current |
| 2008-05-17 |
Fiona Glaser | explicit write combining, because gcc fails at optimizi... |
blob | commitdiff | raw | diff to current |
| 2008-05-17 |
Fiona Glaser | force unroll macroblock_load_pic_pointers |
blob | commitdiff | raw | diff to current |
| 2008-05-17 |
Fiona Glaser | r836 borked lossless cabac nnz |
blob | commitdiff | raw | diff to current |
| 2008-05-14 |
Fiona Glaser | faster probe_skip |
blob | commitdiff | raw | diff to current |
| 2008-04-27 |
Fiona Glaser | remove some redundant nnz counts |
blob | commitdiff | raw | diff to current |
| 2008-04-27 |
Fiona Glaser | omit P/B-skip mc from macroblock_encode if the pixels... |
blob | commitdiff | raw | diff to current |
| 2008-03-30 |
Loren Merritt | r784 borked lossless dc zigzag |
blob | commitdiff | raw | diff to current |
| 2008-03-22 |
Loren Merritt | cosmetics in DECLARE_ALIGNED |
blob | commitdiff | raw | diff to current |
| 2008-03-22 |
Loren Merritt | don't distinguish between luma4x4 and luma4x4ac |
blob | commitdiff | raw | diff to current |
| 2008-03-20 |
Loren Merritt | reduce zigzag arrays from int to int16_t |
blob | commitdiff | raw | diff to current |
| 2008-03-19 |
Fiona Glaser | skip intra pred+dct+quant in cases where it's redundant... |
blob | commitdiff | raw | diff to current |
| 2008-03-03 |
Loren Merritt | increase the alignment of the i8x8 edge cache, needed... |
blob | commitdiff | raw | diff to current |
| 2007-12-02 |
Loren Merritt | reduce the data type used in some tables. 16KB smaller... |
blob | commitdiff | raw | diff to current |
| 2007-11-19 |
Loren Merritt | cosmetics: reorder mc_luma/mc_chroma/get_ref arguments... |
blob | commitdiff | raw | diff to current |
| 2007-07-17 |
Loren Merritt | move os/compiler specific defines to their own header |
blob | commitdiff | raw | diff to current |
| 2007-04-04 |
Loren Merritt | 2x faster quant. 2% overall. |
blob | commitdiff | raw | diff to current |
| 2007-04-04 |
Loren Merritt | when encoding with a cqm, probe_skip now also uses... |
blob | commitdiff | raw | diff to current |
| 2007-01-11 |
Loren Merritt | make array_non_zero() compatible with -fstrict-aliasing |
blob | commitdiff | raw | diff to current |
| 2006-11-18 |
Eric Petit | Merges Guillaume Poirier's AltiVec changes: |
blob | commitdiff | raw | diff to current |
| 2006-10-29 |
Loren Merritt | compute chroma_qp only once per mb |
blob | commitdiff | raw | diff to current |
| 2006-10-29 |
Loren Merritt | rd refinement of intra chroma direction (enabled in... |
blob | commitdiff | raw | diff to current |
| 2006-10-09 |
Loren Merritt | compute pskip_mv only once per macroblock, and store it |
blob | commitdiff | raw | diff to current |
| 2006-10-01 |
Loren Merritt | allow custom deadzones for non-trellis quantization. |
blob | commitdiff | raw | diff to current |
| 2006-10-01 |
Loren Merritt | move zigzag scan functions to dsp function pointers. |
blob | commitdiff | raw | diff to current |
| 2006-10-01 |
Loren Merritt | support interlace. uses MBAFF syntax, but is not adapti... |
blob | commitdiff | raw | diff to current |
| 2006-08-16 |
Loren Merritt | cosmetics |
blob | commitdiff | raw | diff to current |
| 2006-05-10 |
Loren Merritt | faster intra search: filter i8x8 edges only once, and... |
blob | commitdiff | raw | diff to current |
| 2006-04-20 |
Loren Merritt | 10l in r463 (q0 i16x16 dc was permuted) |
blob | commitdiff | raw | diff to current |
| 2006-04-19 |
Loren Merritt | Before, we eliminated dct blocks containing only a... |
blob | commitdiff | raw | diff to current |
| 2006-03-24 |
Loren Merritt | use FDEC_STRIDE instead of a parameter in mmx dct |
blob | commitdiff | raw | diff to current |
| 2006-03-20 |
Loren Merritt | RD subpel motion estimation (--subme 7) |
blob | commitdiff | raw | diff to current |
| 2006-03-09 |
Loren Merritt | keep transposed dct coefs. ~1% overall speedup. |
blob | commitdiff | raw | diff to current |
| 2006-03-06 |
Loren Merritt | no --nr in intra blocks. intra prediction doesn't work... |
blob | commitdiff | raw | diff to current |
| 2006-02-27 |
Loren Merritt | copy current macroblock to a smaller buffer, to improve... |
blob | commitdiff | raw | diff to current |
| 2006-02-27 |
Loren Merritt | copy current macroblock to a smaller buffer, to improve... |
blob | commitdiff | raw | diff to current |
| 2006-01-17 |
Loren Merritt | cosmetics: simplify #includes |
blob | commitdiff | raw | diff to current |
| 2006-01-10 |
Loren Merritt | dct-domain noise reduction (ported from lavc) |
blob | commitdiff | raw | diff to current |
| 2005-11-22 |
Loren Merritt | skip dequant+idct of decimated blocks. |
blob | commitdiff | raw | diff to current |
| 2005-11-07 |
Loren Merritt | use array_non_zero() when we don't need a full array_no... |
blob | commitdiff | raw | diff to current |
| 2005-11-06 |
Loren Merritt | mmx dequant. up to 3% speedup w/ RD. |
blob | commitdiff | raw | diff to current |
| 2005-11-03 |
Loren Merritt | Trellis RD quantization. |
blob | commitdiff | raw | diff to current |
| 2005-10-30 |
Loren Merritt | faster decimation |
blob | commitdiff | raw | diff to current |
| 2005-10-16 |
Loren Merritt | cosmetics: remove #if0'ed code |
blob | commitdiff | raw | diff to current |
| 2005-10-09 |
Loren Merritt | cosmetics: merge some duplicate tables |
blob | commitdiff | raw | diff to current |
| 2005-10-08 |
Loren Merritt | cosmetics: move some stuff from macroblock_encode to... |
blob | commitdiff | raw | diff to current |
| 2005-09-24 |
Loren Merritt | allow i4x4 and i8x8 down-left prediction with emulated... |
blob | commitdiff | raw | diff to current |
| 2005-09-14 |
Loren Merritt | MMX quantization functions, and optimization of the... |
blob | commitdiff | raw | diff to current |
| 2005-06-20 |
Loren Merritt | custom quant matrices |
blob | commitdiff | raw | diff to current |
| 2005-06-17 |
Loren Merritt | lossless mode enabled at qp=0 |
blob | commitdiff | raw | diff to current |
| 2005-06-14 |
Loren Merritt | rate-distortion optimized MB types in I- and P-frames... |
blob | commitdiff | raw | diff to current |
| 2005-06-05 |
Loren Merritt | SATD-based decision for 8x8 transform in inter-MBs. |
blob | commitdiff | raw | diff to current |
| 2005-06-03 |
Loren Merritt | 8x8 transform and 8x8 intra prediction. |
blob | commitdiff | raw | diff to current |
| 2005-02-13 |
Loren Merritt | Cache half-pixel interpolated reference frames, to... |
blob | commitdiff | raw | diff to current |
| 2005-01-28 |
Loren Merritt | Clipping mvs to within picture + emulated border when... |
blob | commitdiff | raw | diff to current |
| 2005-01-27 |
Loren Merritt | Fix clipping of mvs in probe_pskip. (Previously it... |
blob | commitdiff | raw | diff to current |
| 2004-12-31 |
Loren Merritt | remove relative include paths, to avoid conflicts with... |
blob | commitdiff | raw | diff to current |
| 2004-12-17 |
Loren Merritt | implement macroblock types B_SKIP, B_DIRECT, B_8x8 |
blob | commitdiff | raw | diff to current |
| 2004-12-14 |
Loren Merritt | rename 'core/' to 'common/', which avoids conflicts... |
blob | commitdiff | raw | diff to current |
| 2004-07-28 |
Eric Petit | BeOS fixes (no stdint.h, no libm) |
blob | commitdiff | raw | diff to current |
| 2004-07-23 |
Laurent Aimar | * encoder.c, analyse.c, macroblock: fixed when using... |
blob | commitdiff | raw | diff to current |
| 2004-06-03 |
Laurent Aimar | * all: re-import of the CVS. |
blob | commitdiff | raw | diff to current |
|
projects / x264 / history