generic-library/vpx
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
vpx/vp9/common
History
Tero Rintaluoma 56e6c66b49 Avoid division in intra prediction 
- Using multiplication and shifting instead of division in
  intra prediction.
- Maximum absolute difference is 1 for division statements
  in d45, d27, d63 prediction modes. However, errors can
  cumulate for large block sizes when using already predicted
  values.
- Maximum number of non-matching result values in loops using
  division are:
  4x4        0/16
  8x8        0/64
  16x16     10/256
  32x32     13/1024
  64x64    122/4096

  Overall PSNR
  derf:     0.005
  yt:      -0.022
  std-hd:   0.021
  hd:      -0.006

Change-Id: I3979a02eb6351636442c1af1e23d6c4e6ec1d01d
2013-02-20 10:37:36 +02:00
..
generic
Convert subpixel filters to use convolve framework
2013-02-05 14:23:17 -08:00
ppc
Convert subpixel filters to use convolve framework
2013-02-05 14:23:17 -08:00
x86
WIP: ssse3 version of convolve avg functions
2013-02-13 09:15:38 -08:00
vp9_alloccommon.c
Remove NEWCOEFCONTEXT experiment.
2013-02-13 15:12:17 +00:00
vp9_alloccommon.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_asm_com_offsets.c
Remove ARM optimizations from VP9
2012-12-05 08:59:25 -08:00
vp9_blockd.c
Remove some Y2-related code.
2013-02-15 14:06:25 -08:00
vp9_blockd.h
16x16 butterfly inverse ADST/DCT hybrid transform
2013-02-19 09:07:00 -08:00
vp9_coefupdateprobs.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_common.h
Use configure checks for various inline keywords.
2013-02-06 16:12:56 -08:00
vp9_context.c
fixed includes to be fully specified
2012-11-28 10:53:17 -08:00
vp9_convolve.c
Faster convolve8_avg.
2013-02-11 16:53:11 -08:00
vp9_convolve.h
Add 8-tap generic convolver
2013-02-05 14:19:28 -08:00
vp9_debugmodes.c
Replace as_mv struct with array
2013-02-08 20:23:35 -08:00
vp9_default_coef_probs.h
Remove Y2 and Y-no-DC token types from the bitstream.
2013-02-15 14:06:30 -08:00
vp9_entropy.c
Remove Y2 and Y-no-DC token types from the bitstream.
2013-02-15 14:06:30 -08:00
vp9_entropy.h
Remove Y2 and Y-no-DC token types from the bitstream.
2013-02-15 14:06:30 -08:00
vp9_entropymode.c
Adds an error-resilient mode with test
2013-01-23 21:56:15 -08:00
vp9_entropymode.h
Adds an error-resilient mode with test
2013-01-23 21:56:15 -08:00
vp9_entropymv.c
Consistently use get_prob(), clip_prob() and newly added clip_pixel().
2012-12-12 10:01:19 -08:00
vp9_entropymv.h
Added update-able mv-ref probabilities.
2013-01-02 14:22:11 +00:00
vp9_extend.c
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_extend.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_filter.c
Convert subpixel filters to use convolve framework
2013-02-05 14:23:17 -08:00
vp9_filter.h
Convert subpixel filters to use convolve framework
2013-02-05 14:23:17 -08:00
vp9_findnearmv.c
Merge changes Ife0d8147,I7d469716,Ic9a5615f into experimental
2013-02-08 13:19:47 -08:00
vp9_findnearmv.h
Replace as_mv struct with array
2013-02-08 20:23:35 -08:00
vp9_header.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_idct.h
butterfly inverse 4x4 ADST
2013-02-11 10:42:39 -08:00
vp9_idctllm.c
16x16 butterfly inverse ADST/DCT hybrid transform
2013-02-19 09:07:00 -08:00
vp9_implicit_segmentation.c
Use INT64_MAX instead of LLONG_MAX
2013-01-14 15:57:21 -08:00
vp9_invtrans.c
16x16 butterfly inverse ADST/DCT hybrid transform
2013-02-19 09:07:00 -08:00
vp9_invtrans.h
Merge tx32x32 experiment.
2013-01-10 08:23:59 -08:00
vp9_loopfilter_filters.c
Use configure checks for various inline keywords.
2013-02-06 16:12:56 -08:00
vp9_loopfilter.c
Merge "Intrinsic version of loopfilter now matches C code" into experimental
2013-01-24 08:45:22 -08:00
vp9_loopfilter.h
Initial sse2 version of the wide loopfilters
2013-01-11 14:54:14 -08:00
vp9_maskingmv.c
Add vp9_ prefix to all vp9 files
2012-11-27 14:12:30 -08:00
vp9_mbpitch.c
Remove some Y2-related code.
2013-02-15 14:06:25 -08:00
vp9_modecont.c
Nearest / Zero Mv default entropy tweak.
2013-02-08 10:16:13 +00:00
vp9_modecont.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_modecontext.c
fixed includes to be fully specified
2012-11-28 10:53:17 -08:00
vp9_mv.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_mvref_common.c
[WIP] Add column-based tiling.
2013-02-05 15:43:03 -08:00
vp9_mvref_common.h
[WIP] Add column-based tiling.
2013-02-05 15:43:03 -08:00
vp9_onyx.h
Add support for tile rows.
2013-02-13 12:31:00 -08:00
vp9_onyxc_int.h
Remove Y2 and Y-no-DC token types from the bitstream.
2013-02-15 14:06:30 -08:00
vp9_postproc.c
Merge vp9-preview changes into experimental branch
2013-01-08 10:19:59 -08:00
vp9_postproc.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_ppflags.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_pragmas.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_pred_common.c
[WIP] Add column-based tiling.
2013-02-05 15:43:03 -08:00
vp9_pred_common.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_quant_common.c
fixed includes to be fully specified
2012-11-28 10:53:17 -08:00
vp9_quant_common.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_recon.c
Merge tx32x32 experiment.
2013-01-10 08:23:59 -08:00
vp9_reconinter.c
Replace as_mv struct with array
2013-02-08 20:23:35 -08:00
vp9_reconinter.h
Convert subpixel filters to use convolve framework
2013-02-05 14:23:17 -08:00
vp9_reconintra4x4.c
[WIP] Add column-based tiling.
2013-02-05 15:43:03 -08:00
vp9_reconintra.c
Avoid division in intra prediction
2013-02-20 10:37:36 +02:00
vp9_reconintra.h
Merge superblocks (32x32) experiment.
2013-01-08 12:54:45 -08:00
vp9_rtcd_defs.sh
16x16 butterfly inverse ADST/DCT hybrid transform
2013-02-19 09:07:00 -08:00
vp9_rtcd.c
Add vp9_ prefix to all vp9 files
2012-11-27 14:12:30 -08:00
vp9_sadmxn.h
Use configure checks for various inline keywords.
2013-02-06 16:12:56 -08:00
vp9_seg_common.c
Segment Skip Flag
2013-01-28 17:28:04 +00:00
vp9_seg_common.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_setupintrarecon.c
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_setupintrarecon.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_subpelvar.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_swapyv12buffer.c
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_swapyv12buffer.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_systemdependent.h
Merge vp9-preview changes into experimental branch
2013-01-08 10:19:59 -08:00
vp9_tapify.py
Add vp9_ prefix to all vp9 files
2012-11-27 14:12:30 -08:00
vp9_textblit.c
Add vp9_ prefix to all vp9 files
2012-11-27 14:12:30 -08:00
vp9_textblit.h
Use standard integer types for pixel values and coefficients.
2012-12-18 15:31:19 -08:00
vp9_tile_common.c
Add support for tile rows.
2013-02-13 12:31:00 -08:00
vp9_tile_common.h
Add support for tile rows.
2013-02-13 12:31:00 -08:00
vp9_treecoder.c
Consistently use get_prob(), clip_prob() and newly added clip_pixel().
2012-12-12 10:01:19 -08:00
vp9_treecoder.h
Use configure checks for various inline keywords.
2013-02-06 16:12:56 -08:00