Instead of dividing the squared partial sums by the number n of pixels in the
line, we multiply by 840/n, where 840=3*5*7*8. This not only avoids the
divisions, but it also makes the optimization exact as there is no more
rounding.

ntt-short1 resuts:
         MEDIUM (%) HIGH (%)
    PSNR -0.012070 -0.059644
 PSNRHVS -0.016845 -0.020871
    SSIM -0.026984 -0.031257
FASTSSIM -0.026078 0.414901

Change-Id: Ie553d5e3a545dee860a00879d724ecfc00f0a974

This fixes VP10/ActiveMapRefreshTest.Test/*.

Bring commits 575e81f7 and 3d6b8a66 to aom/vp10. These changes predate
the creation of the active map cyclic refresh test.

Change-Id: I3559b6933ffa5649926a4b214e45ed0fae523a25

Brings the following commit to aom:
b3ad8128 Port switch to 9-bit rate cost to vp10.

ntt-short1:

          MEDIUM (%)  HIGH (%)
    PSNR -0.078535   -0.398648
 PSNRHVS -0.416526   -0.468162
    SSIM -0.397539   -0.538516
FASTSSIM  0.008165    0.035524

subset1:

          RATE (%)  DSNR (dB)
    PSNR -0.00991   0.00063
 PSNRHVS  0.00115  -0.00009
    SSIM  0.01019  -0.00050
FASTSSIM  0.08721  -0.00279

Change-Id: I598ba7ee7efcde18439c31dfa96b86cbf297a580

Change-Id: Icfd4654d2fe43ab4851cd86dfedf9639cd4abab3

No change in code behaviour.

Change-Id: If612c1340b4ca30e290947607ecb64a0cb30e95e

No change to code behaviour.

Change-Id: Ia2d7d57fc2018689948d6ba9471a8e848e79aa15

This reduces the deringing worst-case number of "latency" lines from 10 to 8.

ntt-short1 results:
         MEDIUM (%) HIGH (%)
    PSNR -0.063681 -0.027661
 PSNRHVS -0.063628 -0.007841
    SSIM -0.007828 -0.048415
FASTSSIM  0.139420 -0.183910

Change-Id: I006544ab748b76c0334e406300b3eda20e125846

ntt-short1 results:
         MEDIUM (%) HIGH (%)
    PSNR  0.022148 -0.038920
 PSNRHVS  0.047667 -0.053941
    SSIM  0.014602 -0.046534
FASTSSIM -0.450261 -0.690958

Change-Id: Iaa8844d4d84c6265b488a4ade5a223bc3af99707

ntt_short1:
         MEDIUM (%) HIGH (%)
    PSNR -0.002561  0.044074
 PSNRHVS  0.021798  0.082885
    SSIM  0.005793  0.082526
FASTSSIM -0.054113 -0.130136

Change-Id: Icbc38383d27acb4abd3aa30c11d7fb4e51c8ce4c

Change-Id: I9fd4052f0b3f280496882f4475c97b1b5a75934f

The Daala implementation of OD_ILOG(x) computes the number of bits
 necessary to represent the unsigned integer x in two's compliment.
We adjust the computation in od_dering.c to better approximate the
 log base 2 and match the previous OD_ILOG() definition.
This commit makes no change to the bitstream.

Change-Id: Ib75c2f47dca7027b3cfc40b110f2d493f9690c9c

Nathan Egge authored 9 years ago and Nathan Egge committed 9 years ago

Change-Id: I4616af430ebbe43ced6f91ccf67a4fc0f40ba3bf

Change-Id: I7b02377c816599fb4cd8ce8144e0efbfab3874af

The deringing filter documentation is in:
J.-M. Valin, The Daala Directional Deringing Filter, arXiv:1602.05975 [cs.MM],
2016.
http://arxiv.org/pdf/1602.05975.pdf

ntt-short1 results:
          MEDIUM (%) HIGH (%)
    PSNR  -2.488088 -2.003235
 PSNRHVS  -1.588932 -0.797850
    SSIM  -1.522767 -1.287861
FASTSSIM   4.307822  3.983496

subset1 improvement is around 2-3% on PSNR (but the rate is mostly outside
the AWCY testing range)

Change-Id: Ic02344ce9faa509f5c3a50a0fb7a7b84a7977e72

Change-Id: I2ff4d994db5430c8b12be0cc4ead4907a4692628

Change-Id: I25911ea7666465ff19821b07e20c26586e88401c

Restore tests speed_test and end_to_end test since
quant matrices are not on by default.

Change-Id: I0baa649a91e83d62acaca099670671e0406a46ab

Change-Id: I080c7449c269054cf441c251acd8930431e94b46

Change-Id: If740ec7e549d8bcd387f92c43fd8105c528e86ed

Change-Id: If0ba62428216fa343b9a37a3b349edba4103c00a

This commit adds the constrained low-pass filter. For details, please
see document here: https://goo.gl/VUzEED

Change-Id: Idc21d62052071e6c939c91c0cc8246eed768e6b0

Change-Id: I5bd5d01308c391cd3855844a17336262d12241ee

Change-Id: Ic7a57957ad47e78300e73b85fa26e56b594a3a90

Change-Id: I93054d7a4bf9f66883d6cd895cc97b17da1e3eca

high-bitdepth version
d207e, d63e, d45e are only used with CONFIG_MISC_FIXES

cherry-picked from libvpx:
cdf1077d intrapred: protect functions w/CONFIG check x2

Change-Id: I77292e11f51fd76d4127fd0027f876866bcf8675

d207e, d63e, d45e are only used with CONFIG_MISC_FIXES

cherry-picked from libvpx:
fea27ccc intrapred: protect functions w/CONFIG check

Change-Id: If13946e483c4d0ccaa3e1d60dc14216c06d5a219

dead code since:
5d3327e8 Remove palette from VP10

cherry-picked from libvpx:
ebf25868 vp10: remove unused (read|write)_uniform

Change-Id: I4a36575706ea6fffefe5bc778595112ef6ff37d8

This commit takes several upstream commits in libvpx, which make a few
assemblly optimized inverse tranform functions to work for builds when
configured with --enable-vpx-highbitdepth.

Change-Id: Ibc1f18de196c1401a60faf851e5bee18ed616e69

Change-Id: I0d6763957d27a586088cdcd50bd46982b8a686f3

Change-Id: Ief21aab06cf6749be3b502a083554ad9e2d6e588

Change-Id: I4c80319cc7c4594cd6b94cb93fad2628d558c6da

cherry-picked from libvpx:
4f780e94 vp9/10: fix encoder crash on flush

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.

Change-Id: Ie0f6f7ab394f1ed5595aa9d0a49317c1dae0effc

clang-format authored 9 years ago and James Zern committed 9 years ago

Change-Id: I31e383b0e3312d6a07efa104c799ff9aa42d433f