1. 08 Mar, 2011 1 commit
    • John Koleszar's avatar
      correct zbin boost for splitmv mode · 5d1d9911
      John Koleszar authored
      Disable zbin boost in SPLITMV mode as intended. Was incorrectly looking
      at vp8_ref_frame_order instead of vp8_mode_order when comparing against
      SPLITMV. This condition should have always been false, as SPLITMV is
      not in the range of valid reference frames.
      Change-Id: I0408cc7595eff68f00efef6d008e79f5b60d14bf
  2. 07 Mar, 2011 2 commits
    • John Koleszar's avatar
      Merge "Fix format-string warning" · 1016b856
      John Koleszar authored
    • Ralph Giles's avatar
      Fix format-string warning · fe9a604b
      Ralph Giles authored
      Cast size_t to (unsigned long) and print it with the %lu format
      string, which is more portable than C99's explict %zu for size_t.
      This truncates on Windows x64 but otherwise works on 32 and 64 bit
      platforms. In practice the stats file is unlikely to be so large.
      Change-Id: I0432b3acf85fc6ba4ad50640942e1ca4614b21cb
  3. 04 Mar, 2011 11 commits
  4. 03 Mar, 2011 2 commits
    • Mikhal Shemer's avatar
      Fixing divide by zero · 1de99a2a
      Mikhal Shemer authored
      Change-Id: I9d8a98a2f7ed1e3116d0bae35164618c41998bac
    • John Koleszar's avatar
      Fix drastic undershoot in long form content · 36be4f7f
      John Koleszar authored
      When the modified_error_left accumulator exceeds INT_MAX, an incorrect
      cast to int resulted in a negative value, causing the rate control to
      allocate no bits to that keyframe group, leading to severe undershoot
      and subsequent poor quality.
      This error was exposed by the recent change to the rolling target and
      actual spend accumulators in commit 305be4e4 which fixed them to
      actually calculate the average value rather than be re-initialized
      on every frame to the average per-frame bitrate. When this bug was
      triggered, the target bitrate could be 0, so the rolling target
      becomes small, which causes the undershoot. The code prior to 305be4e4
      did not exhibit this behavior because the rolling target was always
      set to a reasonable value and was independent of the actual target
      bitrate. With this patch, the actual target bitrate is calculated
      correctly, and the rate control tracks as expected.
      This cast was likely added to silence a compiler warning on a comparison
      between a double (modified_error_left) and an int (0). Instead, this
      patch removes the cast and changes the comparison to be against 0.0,
      which should prevent the warning from reoccuring.
      This fixes issue #289. Special thanks to gnafu for his efforts in
      reporting and debugging this fix.
      Change-Id: Ie5cc1a7b516c578a76c3a50c892a6f04a11621fe
  5. 02 Mar, 2011 2 commits
  6. 01 Mar, 2011 2 commits
    • John Koleszar's avatar
      change CFLAGS for 64 bit icc builds · 06ce0d88
      John Koleszar authored
      AMD64 only implies SSE2, not SSE3. There aren't any known cases where
      icc was generating SSE3 instructions since all the vectorizable code
      is already in handwritten asm, so this fix is included mostly for
      correctness. Fixes issue #259.
      Change-Id: I993335a4740b68b559035305fb52ca725a6beaff
    • John Koleszar's avatar
      examples: use function to get iface pointers · 987ac894
      John Koleszar authored
      MSVC can't pass the address of global variables in a DLL correctly
      across DLL boundaries. This patch allows linking the examples to
      a libvpx dll build. Fixes issue #268.
      Change-Id: I1c52d076cfc68efb3efdfba019f12d53c5019f58
  7. 28 Feb, 2011 4 commits
  8. 25 Feb, 2011 6 commits
  9. 24 Feb, 2011 7 commits
  10. 23 Feb, 2011 2 commits
  11. 22 Feb, 2011 1 commit
    • Johann's avatar
      purge wince configuration · 418f4219
      Johann authored
      this has been broken since the initial release
      Change-Id: If0d4deb2de9f7d0c4c05641e2bbf9cc1bf11e171