1. 17 Aug, 2009 1 commit
    • Fiona Glaser's avatar
      Lookahead VBV · 30a82c75
      Fiona Glaser authored
      Use the large-scale lookahead capability introduced in MB-tree for ratecontrol purposes.
      (Does not require MB-tree, however.)
      Greatly improved quality and compliance in 1-pass VBV mode, especially in CBR; +2db OPSNR or more in some cases.
      Fix some other bugs in VBV, which should improve non-lookahead mode as well.
      Change the tolerance algorithm in row VBV to allow for more significant mispredictions when buffer is nearly full.
      Note that due to the fixing of an extremely long-standing bug (>1 year), bitrates may change by nontrivial amounts in CRF without MB-tree.
      30a82c75
  2. 14 Aug, 2009 2 commits
    • Fiona Glaser's avatar
      Fix bug in b-adapt 1 · 50f7afcd
      Fiona Glaser authored
      B-adapt 1 didn't use more than MAX(1,bframes-1) B-frames when MB-tree was off.
      50f7afcd
    • Fiona Glaser's avatar
      Fix a potential failure in VBV · e586d699
      Fiona Glaser authored
      If VBV does underflow, ratecontrol could be permanently broken for the rest of the clip.
      Revert part of the previous VBV changes to fix this.
      e586d699
  3. 13 Aug, 2009 4 commits
  4. 09 Aug, 2009 3 commits
  5. 08 Aug, 2009 3 commits
  6. 07 Aug, 2009 4 commits
    • Anton Mitrofanov's avatar
      Fix a potential infinite loop in QPfile parsing on Windows · 7dec1a15
      Anton Mitrofanov authored
      ftell doesn't seem to work properly on Windows in text mode.
      7dec1a15
    • Fiona Glaser's avatar
      Fix delay calculation with multiple threads · 3667fbf9
      Fiona Glaser authored
      Delay frames for threading don't actually count as part of lookahead.
      3667fbf9
    • Fiona Glaser's avatar
      Add "veryslow" preset · 07178d3c
      Fiona Glaser authored
      Apparently some people are actually *using* placebo, so I've added this preset to bridge the gap.
      07178d3c
    • Fiona Glaser's avatar
      Macroblock-tree ratecontrol · 835ccc3c
      Fiona Glaser authored
      On by default; can be turned off with --no-mbtree.
      Uses a large lookahead to track temporal propagation of data and weight quality accordingly.
      Requires a very large separate statsfile (2 bytes per macroblock) in multi-pass mode.
      Doesn't work with b-pyramid yet.
      Note that MB-tree inherently measures quality different from the standard qcomp method, so bitrates produced by CRF may change somewhat.
      This makes the "medium" preset a bit slower.  Accordingly, make "fast" slower as well, and introduce a new preset "faster" between "fast" and "veryfast".
      All presets "fast" and above will have MB-tree on.
      Add a new option, --rc-lookahead, to control the distance MB tree looks ahead to perform propagation analysis.
      Default is 40; larger values will be slower and require more memory but give more accurate results.
      This value will be used in the future to control ratecontrol lookahead (VBV).
      Add a new option, --no-psy, to disable all psy optimizations that don't improve PSNR or SSIM.
      This disables psy-RD/trellis, but also other more subtle internal psy optimizations that can't be controlled directly via external parameters.
      Quality improvement from MB-tree is about 2-70% depending on content.
      Strength of MB-tree adjustments can be tweaked using qcompress; higher values mean lower MB-tree strength.
      Note that MB-tree may perform slightly suboptimally on fades; this will be fixed by weighted prediction, which is coming soon.
      835ccc3c
  7. 05 Aug, 2009 1 commit
    • Fiona Glaser's avatar
      Various 1-pass VBV tweaks · 93cc2893
      Fiona Glaser authored
      Make predictors have an offset in addition to a multiplier.
      This primarily fixes issues in sources with lots of extremely static scenes, such as anime and CGI.
      We tried linear regressions, but they were very unreliable as predictors.
      Also allow VBV to be slightly more aggressive in raising QPs to avoid not having enough bits left in some situations.
      Up to 1db improvement on some clips.
      93cc2893
  8. 29 Jul, 2009 1 commit
    • Fiona Glaser's avatar
      Fix another 10L in QPRD · 1d735afb
      Fiona Glaser authored
      An entry in subpel_iterations was missing.
      I have no idea how QPRD was working at all without this change.
      1d735afb
  9. 28 Jul, 2009 3 commits
  10. 27 Jul, 2009 2 commits
  11. 26 Jul, 2009 4 commits
    • Fiona Glaser's avatar
      Fix 10l in QPRD · 11f50441
      Fiona Glaser authored
      Trellis used wrong lambda with trellis=1
      11f50441
    • Fiona Glaser's avatar
      Fix a nondeterminism with threads and subme>7 · fa3b8139
      Fiona Glaser authored
      Also add a few more checks to eliminate the need for spel_border.
      fa3b8139
    • Fiona Glaser's avatar
      Add QPRD support as subme=10 · 4304c427
      Fiona Glaser authored
      Refactor trellis lambda selection to be done in analyse_init instead of in trellis.
      This will allow for more easy adaption of lambda later on; for now it allows constant lambda across variable QPs.
      QPRD is only available with adaptive quantization enabled and generally improves SSIM and visual quality.
      Additionally, weight the SSD values from RD based on the relative QP offset for chroma; helps visually at high QPs where chroma has a lower QP than luma.
      This fixes some visual artifacts created by QPRD at high QPs.
      Note that this generally hurts PSNR and SSIM, and so is only on when psy-RD is on.
      4304c427
    • Fiona Glaser's avatar
      SSSE3 cachesplit workaround for avg2_w16 · d68f3b07
      Fiona Glaser authored
      Palignr-based solution for the most commonly used qpel function.
      1-1.5% faster overall on Core 2 chips.
      d68f3b07
  12. 22 Jul, 2009 1 commit
  13. 20 Jul, 2009 1 commit
    • Anton Mitrofanov's avatar
      New AQ algorithm option · 2e1db1f6
      Anton Mitrofanov authored
      "Auto-variance" uses log(var)^2 instead of log(var) and attempts to adapt strength per-frame.
      Generates significantly better SSIM; on by default with --tune ssim.
      Whether it generates visually better quality is still up for debate.
      Available as --aq-mode 2.
      2e1db1f6
  14. 17 Jul, 2009 1 commit
  15. 12 Jul, 2009 1 commit
  16. 10 Jul, 2009 1 commit
  17. 09 Jul, 2009 1 commit
  18. 07 Jul, 2009 3 commits
  19. 03 Jul, 2009 2 commits
    • Fiona Glaser's avatar
      Update Gabriel's email address in AUTHORS · 8878778c
      Fiona Glaser authored
      8878778c
    • Fiona Glaser's avatar
      Early termination for chroma encoding · 205a032c
      Fiona Glaser authored
      Faster chroma encoding by terminating early if heuristics indicate that the block will be DC-only.
      This works because the vast majority of inter chroma blocks have no coefficients at all, and those that do are almost always DC-only.
      Add two new helper DSP functions for this: dct_dc_8x8 and var2_8x8.  mmx/sse2/ssse3 versions of each.
      Early termination is disabled at very low QPs due to it not being useful there.
      Performance increase is ~1-2% without trellis, up to 5-6% with trellis=2.
      Increase is greater with lower bitrates.
      205a032c
  20. 26 Jun, 2009 1 commit
    • David Conrad's avatar
      Fix bug in checkasm · 8a96d510
      David Conrad authored
      frame_init_lowres_core check didn't check the C plane.
      However, all x86 and PPC assembly was correct regardless of the unit test being incorrect.
      8a96d510