1. 29 Jan, 2016 3 commits
  2. 28 Jan, 2016 4 commits
    • James Zern's avatar
      vp9/10: fix encoder crash on flush · 4f780e94
      James Zern authored
      the lookahead buffer allocation is deferred to receipt of the first
      frame to allow profile changes. if the encoder was flushed before
      supplying any frames the encoder would crash trying to dereference the
      NULL buffer. vp8 is unaffected.
      fixes mozilla bug:
      Change-Id: Icee4b64de760476eee0d33b568f0a1010335ff13
    • Marco's avatar
      vp9: encoder-breakout: Modify UV thresholds. · 1c2df9cd
      Marco authored
      Set UV thresholds to 0 for encoder breakout if
      block is detected as skin.
      Change-Id: Iaaa30b5cc8bae4cf611f8ea8eba693b730e32a3f
    • JackyChen's avatar
      VPX skin map improvement. · d1e3d046
      JackyChen authored
      Use multiple clusters instead of one and decrease
      the distance thresholds.
      Add a define to switch between models.
      Default is set to existing (1 cluster) model.
      Change-Id: I802cd9bb565437ae8983ef39453939f5d5073bb1
    • Marco's avatar
      vp9 non-rd mode: Modification for detected skin areas. · b39a599c
      Marco authored
      If a superblock contains alot of "skin" then force split
      of 64x64 partition, and make some adjustments in mode selection.
      This helps to reduce artifacts on moving face/skin areas at low bitrates.
      Little/no change in metrics: avgPSNR/SSIM down by ~0.12%.
      Small encoding time increase < 1%.
      Change-Id: Ic57f52148c3716f391419fab0530d916e4c1d186
  3. 27 Jan, 2016 2 commits
  4. 25 Jan, 2016 2 commits
    • Alex Converse's avatar
      vp9: Be more responsive to massive overshoot. · 4b2b90fe
      Alex Converse authored
      Limit oscilation detection in the case where overshoot is very very
      This keeps the 9-bit cost patch from breaking the DownUp reisze test.
      The patch pushed us to an 11% undershoot right before a scene cut
      causing a 1200% overshoot. (Whereas before we were undershooting by
      only 6% before overshooting by 1200%).
      Change-Id: Id90ccfab8aba872ccadc45b73b3bb097b895677f
    • Zoe Liu's avatar
      Cleaned a code in define_gf_group() for firstpass · 9e612763
      Zoe Liu authored
      Change-Id: I0b8eff9e74e3446024482d011ec4fb28b19136d3
  5. 22 Jan, 2016 3 commits
  6. 21 Jan, 2016 5 commits
  7. 20 Jan, 2016 2 commits
  8. 15 Jan, 2016 3 commits
  9. 14 Jan, 2016 2 commits
    • Yaowu Xu's avatar
      Properly set segmentation for error-resilient mode · 0dfe9618
      Yaowu Xu authored
      Change-Id: I0c85b637e51dc949a4e919f10a78870f23c2ca6a
    • Debargha Mukherjee's avatar
      Adding an aq mode for 360 videos · 02345be9
      Debargha Mukherjee authored
      Different quality levels are used for different regions in
      the frame depending on how far they are vertically from the
      center. Specifically, three segments are used based on the
      mi_row index with respect number to the number of mi_rows in
      the frame.
      Change-Id: Ifc8b777bc58ea8521dffc4640360c67d99f8d381
  10. 13 Jan, 2016 2 commits
  11. 07 Jan, 2016 2 commits
    • Yaowu Xu's avatar
      Enable encoder to avoid 8x4 or 4x8 partitions · 9cac17d1
      Yaowu Xu authored
      This commit enables encoder to avoid 8x4 and 4x8 partitions for
      scaled reference frames when libvpx is configured and built with
      Change-Id: I02ad65c386f5855f4325d72570c49164ed52f413
    • Yaowu Xu's avatar
      Fix a typo · 650a2d76
      Yaowu Xu authored
      Change-Id: I12de2dd5e5f375551804166188d76a9ad8067b41
  12. 06 Jan, 2016 1 commit
  13. 05 Jan, 2016 1 commit
    • Marco's avatar
      vp9-skin detection: Refactoring. · a8b7c6aa
      Marco authored
      Add function to compute skin map for a given block, as its
      used in several places (cyclic refresh, noise estimation, and denoising).
      Change-Id: Ied622908df43b6927f7fafc6c019d1867f2a24eb
  14. 04 Jan, 2016 1 commit
    • Marco's avatar
      vp9-svc: Set initial values for ext_buffer/flag indices. · e5dfca02
      Marco authored
      Set initial values for these parameters in the vp9_init_layer_context().
      This also fixes an issue in the svc-bypass mode when frame flags are
      passed via the vpx_codec_encode().
      Change-Id: I0968f04672f8d3d2fe2cea6b8a23f79f80d7a8b1
  15. 22 Dec, 2015 1 commit
    • Marco's avatar
      aq-mode=3: Don't reset segment if block is determined to be skin. · b121a3e7
      Marco authored
      For coding block sizes <=16X16, if the block is determined to be skin,
      then always allow for that block to be candidate for refresh. So if that
      block happens to be on the boost segment(s), segment won't get reset to 0
      and delta-q will be applied.
      PSNR/SSIM metrics neutral (little/no change) on RTC clips.
      Speed increase small/negligible (< 1%).
      Some visual improvement on faces in a few RTC clips.
      Change-Id: I6bf0fce6f39d820b491ce05d7c017ad168fce7d6
  16. 18 Dec, 2015 1 commit
    • Marco's avatar
      Non-rd speed >=5: Include H/V intra for bsize=16x16. · c8a2c31e
      Marco authored
      H/V intra mode was only enabled for bsize < 16x16,
      enable it also for bsize=16x16.
      Metrics are neutral with this change:
      Overall very small gain (0.1%), small visual gain on some RTC clips.
      Change-Id: Ib2d7a44382433bfc11cf324aa3cc5c382ea9e088
  17. 16 Dec, 2015 2 commits
  18. 15 Dec, 2015 3 commits
    • Marco Paniconi's avatar
      Revert "Add "unknown" status for noise estimation." · f3e7539c
      Marco Paniconi authored
      This reverts commit e15fedb9.
      Change-Id: Ibf2bce008c727a9754f88814b7630095fa7b8253
    • paulwilkins's avatar
      Fixed interval, fixed Q 1 pass test patch. · fc50d95b
      paulwilkins authored
      For testing implemented a fixed pattern and delta, 1 pass,
      fixed Q, low delay mode.
      This has not in any way been tuned or optimized.
      Change-Id: Idf5ee179b277fa15d07a97f14f2ce5bbaae80a04
    • paulwilkins's avatar
      1 pass VBR mode bug fix. · cea5e1c1
      paulwilkins authored
      The one pass VBR mode selects a Q range based on a
      moving average of recent Q values. This calculation
      should have been excluding arf overlay frames as these
      are usually coded at the highest allowed value. Their
      inclusion skews the average and can cause it to drift
      upwards even when the clip as a whole is undershooting.
      As such it can undermine correct adaptation of the allowed
      Q range especially for easy content.
      Change-Id: I7d10fe4227262376aa2dc2a7aec0f1fd82bf11f9