1. 14 Aug, 2013 2 commits
  2. 13 Aug, 2013 5 commits
  3. 12 Aug, 2013 11 commits
  4. 11 Aug, 2013 1 commit
    • Mans Rullgard's avatar
      vp9: neon: optimise convolve8_horiz functions · b84dc949
      Mans Rullgard authored
      Each iteration of the horizontal loop reuses 7 of the 11 source
      values.  Loading only the 4 new values saves some time.
      
      Also add preload for source data.
      
      Overall 4% faster on Chromebook.
      
      Change-Id: I8f69e749f2b7f79e9734620dcee51dbfcd716b44
      b84dc949
  5. 10 Aug, 2013 3 commits
  6. 09 Aug, 2013 11 commits
  7. 08 Aug, 2013 7 commits
    • Dmitry Kovalev's avatar
      Using MV struct instead of int[2] array. · 47fad4c2
      Dmitry Kovalev authored
      Change-Id: Iab951c555037e36b154f319f351c5e67f9abb931
      47fad4c2
    • Dmitry Kovalev's avatar
      Removing unneeded intermediate entropy_nodes_adapt var. · ac008f00
      Dmitry Kovalev authored
      Change-Id: I541a178d997b4541e0e2d4d5b854e2ed6b113c3a
      ac008f00
    • Deb Mukherjee's avatar
      Adds a new subpel motion function · 1ba91a84
      Deb Mukherjee authored
      Adds a new subpel motion estimation function that uses a 2-level
      tree-structured decision tree to eliminate redundant computations.
      It searches fewer points than iterative search (which can search
      the same point multiple times) but has the same quality roughly.
      
      This is made the default setting at speeds 0 and 1, while at
      speed 2 and above only a 1-level search is used.
      
      Also includes various cleanups for consistency and redundancy removal.
      
      Results:
      derf: +0.012% psnr
      stdhd: +0.09% psnr
      Speedup of about 2-3%
      
      Change-Id: Iedde4866f5475586dea0f0ba4cb7428fba24eee9
      1ba91a84
    • Adrian Grange's avatar
      Moved fast motion search level decision to function · 83ee80c0
      Adrian Grange authored
      Moving this block of code into a function makes the
      code easier to read and change.
      
      Change-Id: If4ede570cce1eab1982b188c4d3e4fd3d4db236e
      83ee80c0
    • Adrian Grange's avatar
      Simplify & fix potential bug in rd_pick_partition · aae6a4c8
      Adrian Grange authored
      Different partitionings were not being evaluated against
      best_rd and there were unnecessary calls to RDCOST. This
      could have resulted in a non-optimal partioning being
      selected.
      
      I simplified the variables used to track the rate,
      distortion and RD values throughout the function.
      
      Change-Id: Ifa7085ee80d824e86791432a5bc6d8fea5a3e313
      aae6a4c8
    • Dmitry Kovalev's avatar
      Removing plane_block_{width, height}_log2by4 functions. · 61c33d0a
      Dmitry Kovalev authored
      Change-Id: I040b82b8e32aee272d10cbb021c7ba1c76343d7a
      61c33d0a
    • Dmitry Kovalev's avatar
      Cleanup inside vp9_reconinter.c. · a766d891
      Dmitry Kovalev authored
      Using block width and block height instead of their logarithms. Using
      SUBPEL_BITS and SUBPEL_SHIFTS constants instead of magic numbers.
      
      Change-Id: I4e10e93c907c8a5e1cb27dfe74d1fcdcc4995448
      a766d891