1. 23 Aug, 2016 1 commit
    • Urvang Joshi's avatar
      Palette code: remove the use of same if condition twice. · f1906e96
      Urvang Joshi authored
      rd_pick_palette_intra_sby() method is called only when,
      cpi->common.allow_screen_content_tools is on. So, no need to check that
      again. We just use an assert() instead to still be safe.
      
      Change-Id: I19785c2aac016798c8d331bbe91971b3806b73a8
      f1906e96
  2. 22 Aug, 2016 1 commit
  3. 19 Aug, 2016 3 commits
  4. 15 Aug, 2016 1 commit
  5. 12 Aug, 2016 1 commit
  6. 10 Aug, 2016 3 commits
  7. 09 Aug, 2016 1 commit
    • Urvang Joshi's avatar
      Palette code cleanup: · d0000208
      Urvang Joshi authored
      - Avoid some memcpy()s
      - Remove indices array
      - Make pre_indices array local
      - Avoid rounding twice
      - Other small simplifications
      
      Change-Id: Iac3236daaad04f21f54054cdd9504de13b942a07
      d0000208
  8. 05 Aug, 2016 2 commits
    • Yue Chen's avatar
      Speed filter intra mode search in EXT_INTRA experiment · f6a5c274
      Yue Chen authored
      (1) Key frame: skip filter intra modes whose directional pred
          version is relatively bad (rd >= 1.125 * best_rd)
      (2) Inter frame: do not check filter intra modes if best_intra_rd
          >= 1.25 * best_rd
      
      Encoding time overhead is reduced by:
      4.9% (9.2%->4.3%, soccer_cif)
      Coding gains drop by 0.021% on lowres and by 0.076% on midres
      
      Change-Id: I29b6f7d3d3dc4b362c6d63bc447e6a429ba5dc66
      f6a5c274
    • Urvang Joshi's avatar
      Palette code: simpler and faster duplicate removal · 016a5daa
      Urvang Joshi authored
      Change-Id: I0c1baa5ca73c1f067d69239d3e31d1050b4706d2
      016a5daa
  9. 04 Aug, 2016 2 commits
    • Urvang Joshi's avatar
      Make palette code faster: replace nested for loops by a single memcpy(). · a0a23b7f
      Urvang Joshi authored
      Change-Id: Ia14df45a35c98d680822454fbb8d1763884c1852
      a0a23b7f
    • Zoe Liu's avatar
      Code refactoring on Macros related to ref frame numbers · 1af28f02
      Zoe Liu authored
      We have renamed following Macros to avoid name confusion:
      
      REFS_PER_FRAME --> INTER_REFS_PER_FRAME
      (= ALTREF_FRAME - LAST_FRAME + 1)
      MAX_REF_FRAMES --> TOTAL_REFS_PER_FRAME
      (= ALTREF_FRAME - INTRA_FRAME + 1)
      
      INTER_REFS_PER_FRAME specifies the maximum number of reference frames
      that each Inter frame may use.
      TOTAL_REFS_PER_FRAME is equal to INTER_REFS_PER_FRAME + 1, which
      counts the INTRA_FRAME.
      
      Further, at the encoder side, since REF_FRAMES specifies the maximum
      number of the reference frames that the encoder may store, REF_FRAMES
      is usually larger than INTER_REFS_PER_FRAME. For example, in the
      ext-refs experiment, REF_FRAMES == 8, which allows the encoder to
      store maximum 8 reference frames in the buffer, but
      INTER_REFS_PER_FRAME equals to 6, which allows each Inter frame may
      use up to 6 frames out of the 8 buffered frames as its references.
      Hence, in order to explore the possibility to store more reference
      frames in future patches, we modified a couple of array sizes to
      accomodate the case that the number of buffered reference frames is
      not always equal to the number of the references that are being used
      by each Inter frame.
      
      Change-Id: I19e42ef608946cc76ebfd3e965a05f4b9b93a0b3
      1af28f02
  10. 02 Aug, 2016 2 commits
  11. 29 Jul, 2016 1 commit
  12. 23 Jul, 2016 1 commit
  13. 21 Jul, 2016 1 commit
    • Debargha Mukherjee's avatar
      Rectangular transforms 4x8 & 8x4 · e5848dea
      Debargha Mukherjee authored
      Added a new expt rect-tx to be used in conjunction with ext-tx.
      [rect-tx is a temporary config flag and will eventually be
      merged into ext-tx once it works correctly with all other
      experiments].
      
      Added 4x8 and 8x4 tranforms for use initially with rectangular
      sub8x8 y blocks as part of this experiment.
      
      There is about a -0.2% BDRATE improvement on lowres, others pending.
      
      When var-tx is on rectangular transforms are currently not used.
      That will be enabled in a subsequent patch.
      
      Change-Id: Iaf3f88ede2740ffe6a0ffb1ef5fc01a16cd0283a
      e5848dea
  14. 18 Jul, 2016 2 commits
    • hui su's avatar
      Extra round of subpel MV search around second best full-pixel MV · 9a470241
      hui su authored
      Keep track of the best and second best full pixel motion vector
      candidates, and do subpel search around both of them.
      
      Compression improvement:
      lowres 0.22%   midres 0.23%   hdres 0.18%
      
      No noticeable encoding speed changes observed on lowres test clips.
      
      Change-Id: I5f4df2a03d1db061cfdfdba6138b27e9ea91f089
      9a470241
    • Sarah Parker's avatar
      Add buf0, width, height fields to buf_2d · 166c3250
      Sarah Parker authored
      These are needed for the warping function in the global motion
      experiment.
      
      Change-Id: Iaab176d0c0b90f6b938e2bac48b24c07e87e3cd9
      166c3250
  15. 14 Jul, 2016 2 commits
  16. 13 Jul, 2016 2 commits
    • hui su's avatar
      Refactor codes about motion search · 581636d7
      hui su authored
      1. Add "best_mv" in MACROBLOCK to store the best motion vector
      during motion search, so that we don't need to pass its pointer
      to various motion search functions.
      
      2. Declare some functions as static when possible.
      
      3. Fix some indents.
      
      Change-Id: I0778146c0866cbc55e245988c59222577ea8260e
      581636d7
    • Geza Lore's avatar
      Optimize and cleanup obmc predictor and rd search. · 4c4f04ac
      Geza Lore authored
      Use vpx_blend_a64_hmask and vpx_blend_a64_vmask to speed up
      computing the obmc predictor. Clean up calc_target_weighted_pred.
      
      Encoder speedup: 1.3%
      Decoder speedup: 6.5%
      
      Change-Id: I0c774fe53d22399e92a10d1daf3af0010d88d2c5
      4c4f04ac
  17. 11 Jul, 2016 2 commits
  18. 08 Jul, 2016 2 commits
    • Jingning Han's avatar
      Integrate ext-interp into dual filter framework · e3a2aeb0
      Jingning Han authored
      The combination of the two experiments improves the compression
      performance gains:
      
      lowres 2.5%
      midres 2.1%
      
      Change-Id: Id26c0a9474ce08893aa1d946365c7ff850fab57a
      e3a2aeb0
    • Jingning Han's avatar
      Properly reset rate and distortion value for zero pred residual case · 1bf039cc
      Jingning Han authored
      When the prediction residuals are all zero, reset the coeff rate
      cost and the distortion value to be zero. This change doesn't affect
      lowres set significantly, but improves several clips in the midres
      set, like sintel_480p and mobisode2_480p, by a few percents. The
      average performance for midres set is improved by 0.2%.
      
      Change-Id: Idd5ebf2652e556a1b1c569fe3c48dacef3f11c32
      1bf039cc
  19. 04 Jul, 2016 1 commit
    • Geza Lore's avatar
      Refactoring in preparation for OBMC optimizations. · 007aa7dd
      Geza Lore authored
      - Use int32_t instead of int in vpx_obmc{variance,sad} functions
      - Remove weigthed_src and obmc mask strides and assume contiguous
        buffers. These inputs can always be packed as contiguous arrays.
      
      Change-Id: I74c09b3fb3337f13d39e13a9cb61e140536f345d
      007aa7dd
  20. 01 Jul, 2016 1 commit
  21. 30 Jun, 2016 1 commit
    • Geza Lore's avatar
      Reject ext-inter compound modes based on modelled RD. · 532304e4
      Geza Lore authored
      Reject ext-inter compound modes before doing full rate distortion
      evaluation, if the corresponding single reference modes had a lower
      modelled RD.
      
      ext-inter speedup up to TBD.
      
      Coding performance: TBD
      
      Change-Id: I358bfb879c5ebe5e7afbf6f540cc784f8de14857
      532304e4
  22. 29 Jun, 2016 1 commit
    • Jingning Han's avatar
      Fix shift value in dist_block with hbd · 49222c37
      Jingning Han authored
      This offset value related to the bit depth has been taken care of
      inside the function vp10_highbd_block_error.
      
      Change-Id: I58dd8a53380ba4529d59837e56a951bc81a2962e
      49222c37
  23. 28 Jun, 2016 2 commits
    • Geza Lore's avatar
      Remove skip_txfm optimization. · 92922be8
      Geza Lore authored
      Commit 0d6980d7 removed some use
      of the skip_txfm optimization, and the rest are not productive.
      
      The current use of this optimization is only used with --good
      and --cpu-used >= 3, however the overhead of this is higher than the
      speedup it yields.
      
      Removing this, and subsequently simplifying model_rd_for_sb yields
      a net encoder speedup:
      --cpu-used=0    ~1.5% faster
      --cpu-used=3    ~2.0% faster
      
      The code simplification is also significant.
      
      Change-Id: I1dd668c32de15a2e912c59c42379d0f9e1032ff8
      92922be8
    • Debargha Mukherjee's avatar
      Quantization fix for new-quant/var-tx · f3dfa0c3
      Debargha Mukherjee authored
      Also use the fp quantizer consistently
      
      lowres: -0.07 BDRATE improvement
      
      Change-Id: I9174f6ad54a74d38541004b99cb3689d0c09be55
      f3dfa0c3
  24. 27 Jun, 2016 1 commit
  25. 24 Jun, 2016 2 commits
  26. 23 Jun, 2016 1 commit