1. 06 Dec, 2017 7 commits
    • Jerome Jiang's avatar
      14dbdd95
    • Jerome Jiang's avatar
      Add frame width & height to frame pkt. Add test. · bd1d995c
      Jerome Jiang authored
      Used to return correct frame width and height when dynamic resizing happens.
      
      BUG=webm:1474
      
      Change-Id: Ia2043f7e1635b3821848a67b9b134f47f14b0f3a
      bd1d995c
    • Marco's avatar
      vp9-svc: Set downsampling filter for VGA layer. · 3562d6b0
      Marco authored
      Downsampling filter for SVC was set to subsample (phase 0)
      for HD -> VGA, and bilinear averaging (phase 8) for VGA -> QVGA.
      This change makes it bilinear averaging for HD -> VGA.
      
      Given the recent commit 9f9d4f8d, quality is improved with
      this change: avgPSNR/SSIM up ~1-3% on HD clips in RTC set.
      Speed decrease of ~1% for 3 layer SVC.
      
      Change-Id: If834a320e372b8b922a6bf7cab4227703b1beae6
      3562d6b0
    • Marco Paniconi's avatar
      575c1933
    • Hui Su's avatar
    • Marco's avatar
      vp9: Nonrd-pickmode: move some early exits up. · 33953f31
      Marco authored
      Move the early exit checks on usable_ref_frame and
      skip_ref_find_pref up before the check on flag_svc_subpel.
      The code under flag_svc_subpel requires frame_mv to be set
      for the golden/spatial reference, which is only set if the
      both those exits don't pass.
      
      No change in behavior.
      
      Change-Id: Id304276c745eeb389ff85fa2dcf510d5976bc413
      33953f31
    • Marco's avatar
      vp9-svc: Allow for nonzero motion on spatial reference. · 9f9d4f8d
      Marco authored
      For nonrd pickmode on a given spatial layer, the spatial
      (golden) reference was always only using zeromv for prediction.
      In this patch if the downsampling filter used for generating
      the lower spatial layer is an averaging filter (nonzero phase),
      we allow for subpel motion on the spatial (golden) reference to
      compensate for the shift. This is done by forcing the testing of
      nonzero motion mode to compensate for spatial downsampling shift.
      
      Improvement for cases where the downsampling is averaging filter.
      In the current code this is only done for generating
      resolutions <= QVGA.
      
      Improvement for avgPSNR/SSIM on RTC set for speed 7: ~1.2%.
      Gain is larger (~2-3%) for VGA clips with 2 spatial layers.
      ~1% speed slowdown for 3 layer SVC on mac.
      
      Change-Id: I9ec4fa20a38947934fc650594596c25280c3b289
      9f9d4f8d
  2. 04 Dec, 2017 1 commit
  3. 02 Dec, 2017 1 commit
  4. 01 Dec, 2017 10 commits
  5. 30 Nov, 2017 5 commits
  6. 29 Nov, 2017 13 commits
  7. 27 Nov, 2017 2 commits
    • Johann's avatar
      quantize x86: dedup some parts · bd990cad
      Johann authored
      Change-Id: I9f95f47bc7ecbb7980f21cbc3a91f699624141af
      bd990cad
    • Marco's avatar
      vp9-svc: Fix to the layer buffer settings. · cbe62b9c
      Marco authored
      For the case when the number of temporal layers > 1,
      the buffer levels (starting/optimal_buffer_level,
      and maximum_buffer_size) were not scaled properly.
      
      In vp9_update_layer_context_change_config():
      when setting the layer-buffer levels, fix is to scale
      the layer-target_bandwidth by the target_bandwidth
      (which is the full stream bandwidth) instead of the
      spatial_layer_target.
      
      This is needed because prior to the call
      vp9_update_layer_context_change_config(), set_rc_buffer_sizes()
      is called which sets the buffer levels based on target bandwidth
      (which is the full bandwidth for the SVC stream).
      
      This fix properly sets the layer-buffer levels based on the
      layer-bandwidth, and leads to better rate targeting.
      
      Small/neutral change in avgPSNR/SSIM metrics on RTC set.
      
      Change-Id: Ic0f4f7f3487c37b9a9adb4781ae5edfed7140a57
      cbe62b9c
  8. 21 Nov, 2017 1 commit