summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2013-10-03Merge "Using vp9_zero instead of vpx_memset."Dmitry Kovalev
2013-10-03Merge "Making decode_modes_b function more straightforward."Dmitry Kovalev
2013-10-03Using vp9_zero instead of vpx_memset.Dmitry Kovalev
Change-Id: I9a0d0e9c3459954aa7b9c68f92cc5d56385ebd18
2013-10-03Merge "Removing vpx_codec_impl_{top, bottom}.h files."Dmitry Kovalev
2013-10-03Merge "Speed setting review."Paul Wilkins
2013-10-03Merge "make use last partition consider motion"Paul Wilkins
2013-10-03Merge "mips dsp-ase r2 vp9 decoder convolve module optimizations"Johann
2013-10-03Merge "BITSTREAM - RESTORING BILINEAR INTERPOLATION FILTER SUPPORT"Dmitry Kovalev
2013-10-03Speed setting review.Paul Wilkins
Substantial reworking of the speed vs quality trade offs for speed 1 and 2. In this patch I am attempting to freeze the "quality" meaning of speeds 1 and 2 relative to speed 0 so that in future we can better evaluate progress. I am targeting : Speed 1 quality ~-5% vs speed 0. Speed 2 quality ~-10% vs speed 0 It is inevitable that quality will still fluctuate a little as we adjust settings and add new features, but we will attempt to keep as close as possible to these values. Above speed 2 things will remain a bit more fluid for now. In this patch speed 1 is approximately 4-5x as fast as speed 0. This is similar to before but the quality hit is a lot less. Likewise speed 2 is approximately 2x as fast as speed 1 but is similar in quality to the previous speed 1 configuration. Also slight change to behavior of FLAG_EARLY_TERMINATE to insure all reference frames get at least one rd test. Important for very low variance regions. WIP :- Added a new speed level with old speed 4 becoming speed 5. Speed 3 and 4 tradeoffs still WIP Change-Id: Ic7a38dd7b5b63ab1501f9352411972f480ac6264
2013-10-03make use last partition consider motionJim Bankoski
This commit causes use last partition to consider whether a 64x64 has motion that might make a new partitioning worth while. Change-Id: I3a57bedef4f3cd961fadbfa96651c206fa36da4a
2013-10-03Merge "Improved auto_partition_range."Paul Wilkins
2013-10-02BITSTREAM - RESTORING BILINEAR INTERPOLATION FILTER SUPPORTDmitry Kovalev
Adding appropriate test vector vp90-2-06-bilinear.webm. Change-Id: Ia3bbf57318e0cc61a1b724fe751e3f9c7e11b337
2013-10-02mips dsp-ase r2 vp9 decoder convolve module optimizationsParag Salasakar
Change-Id: I401536778e3c68ba2b3ae3955c689d005e1f1d59
2013-10-02Merge "Adding const to function arguments."Dmitry Kovalev
2013-10-02Merge "Removing unused vp9_coeff_stats_model typedef."Dmitry Kovalev
2013-10-02Merge "Moving get_token_alloc function from common to the encoder."Dmitry Kovalev
2013-10-02Merge "Removing memset calls inside idct/iht functions."Dmitry Kovalev
2013-10-02Removing memset calls inside idct/iht functions.Dmitry Kovalev
Making appropriate memset inside decode_block now. Change-Id: I8e944194668c830de08271c8fb6e413251c201d8
2013-10-02Deprecate unused mode count variablesJingning Han
Remove mode_check_freq and mode_test_hit_counts from VP9_COMP. Change-Id: Iabfd9f841444cd9bf19ac761a9795f140082ce0b
2013-10-02BITSTREAM - CLARIFICATION OF MV SIZE RANGEJingning Han
The codec should effectively run with motion vector of range (-2048, 2047) in full pixels, for sequences of 1080p and below. Add assertions to clarify this behavior. Change-Id: Ia0cac28249f587d8f8882205228fa480263ab313
2013-10-02Merge "Adding read_intra_mode_{y, uv} functions for clarity."Dmitry Kovalev
2013-10-01Adding SSE2 optimized vp9_short_idct32x32_1_add function.Dmitry Kovalev
Change-Id: I4b1c6bb9ff615f5872b96ed07dbf0f5e18e63643
2013-10-01Adding read_intra_mode_{y, uv} functions for clarity.Dmitry Kovalev
Change-Id: I92fd32476c472e54f52b8d7602a98262b25e6eaf
2013-10-01Merge "vp9_thread nolintify lint issue I can't fix easily"Jim Bankoski
2013-10-01Merge "vp9_block.h cpplint issues resolved"Jim Bankoski
2013-10-01Merge "cpplint issue in vp9_rdopt.h"Jim Bankoski
2013-10-01Merge "cpplint issues in vp9_onyx_int.h"Jim Bankoski
2013-10-01Making decode_modes_b function more straightforward.Dmitry Kovalev
Moving out decode_tokens function calls and adding decode_blocks boolean variable. We only have to decode if eobtotal > 0, i.e. we have at least one non-zero coefficient. Also inlining and remove vp9_set_pred_flag_mbskip function. Change-Id: I7be38b12ee8206faf0beea2bbf4d52be42575b03
2013-10-01vp9_thread nolintify lint issue I can't fix easilyJim Bankoski
Change-Id: Ib19dabe697656e4d7e8403d91bedca7cd31d36bf
2013-10-01vp9_block.h cpplint issues resolvedJim Bankoski
Change-Id: Icc6a76a5be77f3e19918155bab3998e0aa32ccf5
2013-10-01cpplint issues in vp9_onyx_int.hJim Bankoski
Change-Id: I6c4058aebe834e1a12b7a3fb10484b9ebe60b349
2013-10-01cpplint issue in vp9_rdopt.hJim Bankoski
Change-Id: I84209d382ca5dfc537ee533cd792d8caa0e25cee
2013-10-01Merge "Fix linker warnings for bilinear filters"Matthew Heaney
2013-10-01Fix linker warnings for bilinear filtersMatthew Heaney
The declaration of the bilinear filters specified an alignment clause in the implementation file but not in the header. This turned out to be harmless, but it did cause linker warnings to be emitted when building on Windows. The (extern) declaration in the header was changed, to match the declaration in the implementation. Change-Id: I44be89b1572fe9a50fa47a42e4db9128c4897b04
2013-10-01Merge "Modify HORIZx16 macro in subpixel filter functions"Yunqing Wang
2013-10-01Modify HORIZx16 macro in subpixel filter functionsYunqing Wang
Interleaved the instructions, reduced register dependency, and prefetched the source data. This improved the decoder speed by 0.6% - 2%. Change-Id: I568067aa0c629b2e58219326899c82aedf7eccca
2013-10-01Moving get_token_alloc function from common to the encoder.Dmitry Kovalev
Also renaming mb_row -> mi_row, mb_col -> mi_col arguments and calculate mb_rows/mb_cols values from mi_rows/mi_cols. Change-Id: I6919a279f560648e23bc9a12f507d17c21ffd5d7
2013-10-01fix build with MSVCYaowu Xu
near is a key word, changed to use nearmv instead. Change-Id: Ib54438c431b2b2521a62fc7b61a9c127dd7bc01e
2013-09-30Removing unused vp9_coeff_stats_model typedef.Dmitry Kovalev
Change-Id: I6973e7121b6393379b5759f288632e8eab763d3e
2013-09-30Adding const to function arguments.Dmitry Kovalev
Function list: tx_counts_to_branch_counts_32x32 tx_counts_to_branch_counts_8x8 tx_counts_to_branch_counts_8x8 update_ct update_ct2 update_mode_probs Change-Id: I120d8945a34378cf285d6bd415e23de1d522cf2f
2013-09-30Merge "Using array of motion vectors instead of separate variables."Dmitry Kovalev
2013-09-30Merge "Removing vp9_add_constant_residual_{8x8, 16x16, 32x32} functions."Dmitry Kovalev
2013-09-30Fix rectangular partition check in speed 1Jingning Han
Make encoder skip rectangular partition check in speed 1 and above, when early termination was triggered in partition split. Thanks Guillaume (gmartres@) for catching this issue. This change makes bus_cif at 2000kbps speed 1 runtime goes down from 25612ms to 23438ms (about 9% speed-up), at the expense of -0.235% performance down. Change-Id: I98613fad081a261d30d5fa206f934ca70601c180
2013-09-30Using array of motion vectors instead of separate variables.Dmitry Kovalev
Change-Id: I7380a089105f658257bbb3e30a525da168e76952
2013-09-30Removing vpx_codec_impl_{top, bottom}.h files.Dmitry Kovalev
It doesn't seem reasonable to have these files as our API part. Just inlining them in the source. Change-Id: Iff970bb25e72e49e7ac21990824dbf4ef8bfd2e2
2013-09-30Merge "Using size_t for memory buffer size."Dmitry Kovalev
2013-09-30Removing vp9_add_constant_residual_{8x8, 16x16, 32x32} functions.Dmitry Kovalev
We don't need these functions anymore. The only one which was actually used is vp9_add_constant_residual_32x32. Addition of vp9_short_idct32x32_1_add eliminates this single usage. SSE2 optimized version of vp9_short_idct32x32_1_add will be added in the next patch set, right now it is only C implementation. Now we have all idct functions implemented in a consistent manner. Change-Id: I63df79a13cf62aa2c9360a7a26933c100f9ebda3
2013-09-30Merge "systemdependent lint issue resolved"Jim Bankoski
2013-09-30Merge changes I2b2af1dd,Id2cc5c82Jim Bankoski
* changes: fixed cpp lint issue in vp9_postproc_x86 nolintify intrinsic idct file
2013-09-30Merge "cpplint issues in vp9_loopfilter.h"Jim Bankoski
generated by cgit v1.2.3 (git 2.25.1) at 2024-10-11 23:25:17 +0000