Commit Graph

328 Commits

Author SHA1 Message Date
Yunqing Wang
5ef694cfb8 Remove unused file
Removed vp9_idctllm_mmx.asm

Change-Id: I7152756f23a5a09ed69e8fb40edb2ab3237290fe
2013-02-27 11:00:58 -08:00
Ronald S. Bultje
e8c74e2b70 Move eob from BLOCKD to MACROBLOCKD.
Consistent with VP8.

Change-Id: I8c316ee49f072e15abbb033a80e9c36617891f07
2013-02-27 11:00:55 -08:00
John Koleszar
7ad8dbe417 Remove unused vp9_copy32xn
This function was part of an optimization used in VP8 that required
caching two macroblocks. This is unused in VP9, and might not
survive refactoring to support superblocks, so removing it for now.

Change-Id: I744e585206ccc1ef9a402665c33863fc9fb46f0d
2013-02-27 10:24:56 -08:00
John Koleszar
d8e68bd14b Merge changes I922f8602,I0ac3343d into experimental
* changes:
  Use 256-byte aligned filter tables
  Set scale factors consistently for SPLITMV
2013-02-27 10:08:53 -08:00
Jan Kratochvil
82ed3f9a41 Fix --as=nasm compatibility for new asm code.
s/movd/movq/

Change-Id: Id1a56de91551f8dc796f14f1056c565dfc1ba626
2013-02-27 09:55:38 -08:00
John Koleszar
350ba5f30e Merge "Combined motion compensation with scaled predictors" into experimental 2013-02-27 09:46:12 -08:00
John Koleszar
6fd7dd1a70 Use 256-byte aligned filter tables
This avoids duplicating all the filters twice. Includes fixups to the
convolve routines and associated tests to make this work.

Change-Id: I922f86021594e55072ddb63b42b2313605db6e00
2013-02-27 08:22:39 -08:00
John Koleszar
77f88e97fa Combined motion compensation with scaled predictors
This patch extends the previous support for using references of a
different resolution in ZEROMV mode to all inter prediction modes.
Subpixel based best-mv scoring is disabled when the reference frame
differs in resolution from the current frame.

Change-Id: Id4dc3e5e6692de98d9857fd56bfad3ac57e944ac
2013-02-27 08:22:39 -08:00
John Koleszar
472eeaf082 Set scale factors consistently for SPLITMV
This commit updates the 4x4 prediction to consistently use the
build_2x1_inter_predictor() method. That function is updated to
calculate the scale offset, rather than relying on the caller
to calculate it. In the case that the 2x1 prediction can not
be used, the scale offset is recalculated for each 1x1 block.
The idea here is that the offsets are calculated before each
call to vp9_build_scaled_inter_predictor().

Change-Id: I0ac3343dd54e2846efa3c4195fcd328b709ca04d
2013-02-27 08:22:39 -08:00
Yaowu Xu
858b60e8d0 Merge "Improve 32x32 forward dct" into experimental 2013-02-27 07:56:42 -08:00
John Koleszar
eb939f45b8 Spatial resamping of ZEROMV predictors
This patch allows coding frames using references of different
resolution, in ZEROMV mode. For compound prediction, either
reference may be scaled.

To test, I use the resize_test and enable WRITE_RECON_BUFFER
in vp9_onyxd_if.c. It's also useful to apply this patch to
test/i420_video_source.h:

  --- a/test/i420_video_source.h
  +++ b/test/i420_video_source.h
  @@ -93,6 +93,7 @@ class I420VideoSource : public VideoSource {

     virtual void FillFrame() {
       // Read a frame from input_file.
  +    if (frame_ != 3)
       if (fread(img_->img_data, raw_sz_, 1, input_file_) == 0) {
         limit_ = frame_;
       }

This forces the frame that the resolution changes on to be coded
with no motion, only scaling, and improves the quality of the
result.

Change-Id: I1ee75d19a437ff801192f767fd02a36bcbd1d496
2013-02-26 23:54:23 -08:00
Yunqing Wang
35bc02c6eb Optimize vp9_dc_only_idct_add_c function
Wrote SSE2 version of vp9_dc_only_idct_add_c function. In order to
improve performance, clipped the absolute diff values to [0, 255].
This allowed us to keep the additions/subtractions in 8 bits.
Test showed an over 2% decoder performance increase.

Change-Id: Ie1a236d23d207e4ffcd1fc9f3d77462a9c7fe09d
2013-02-26 17:16:13 -08:00
Dmitry Kovalev
971ff2679f Removing redundant 'extern' keyword from function declarations.
Change-Id: I893fa36297b9bd9cff93d082f1736f6860b15c0d
2013-02-26 15:52:05 -08:00
John Koleszar
25686fc22d Merge "Refactor inter recon functions to support scaling" into experimental 2013-02-26 11:45:28 -08:00
John Koleszar
6a4f708c25 Refactor inter recon functions to support scaling
Ensure that all inter prediction goes through a common code path
that takes scaling into account. Removes a bunch of duplicate
1st/2nd predictor code. Also introduces a 16x8 mode for 8x8
MVs, similar to the 8x4 trick we were doing before. This has an
unexpected effect with EIGHTTAP_SMOOTH, so it's disabled in that
case for now.

Change-Id: Ia053e823a8bc616a988a0af30452e1e75a739cba
2013-02-26 10:03:29 -08:00
Yaowu Xu
66d94ac13c Improve 32x32 forward dct
The commit improves the 32x32 forward dct implementation:
1. change to use same constants and rounding as other forward dcts
2. select rounding to specifically minimize the roundtrip error, which
improved average 19/block to .77/block using 100000 random input.

Test showed a small but consistent gain on all test sets, about .15%

Change-Id: If0afd6a71880a522f60c1c234be0462092c2eb53
2013-02-26 09:23:01 -08:00
Dmitry Kovalev
9bf3f75168 Changing pitch value meaning for fht and iht transforms.
Pitch now means the number of elements, not the number of bytes.

Change-Id: Idb9f2f012e39b09d596a3cc1802305a80b7c13af
2013-02-25 18:19:55 -08:00
Dmitry Kovalev
9770d564f4 Code cleanup.
Removing switch statements for inverse hybrid transforms. Making code style
consistent for all similar transform implementations. Renaming shortpitch
and short_pitch variables to half_pitch.

Change-Id: I875f7a82aae4e8063a58777bf1cc3f1e67b48582
2013-02-25 15:14:01 -08:00
Dmitry Kovalev
20b0cb599b Code cleanup.
Removing redundant parentheses, better code formatting, introducing
ROUND_POWER_OF_TWO macro to replace repeated expression.

Change-Id: I91aad7a53ed03482428b2419de4bb99fd92c6771
2013-02-25 13:38:18 -08:00
Jingning Han
77a3becf92 clean up forward and inverse hybrid transform
Rebased.

Remove the old matrix multiplication transform computation. The 16x16
ADST/DCT can be switched on/off and evaluated by setting ACTIVE_HT16
300/0 in vp9/common/vp9_blockd.h.

Change-Id: Icab2dbd18538987e1dc4e88c45abfc4cfc6e133f
2013-02-25 09:16:12 -08:00
Ronald S. Bultje
0c9e2e9a1d Split coefficient token tables intra vs. inter.
Change-Id: I5416455f8f129ca0f450d00e48358d2012605072
2013-02-23 07:33:46 -08:00
Paul Wilkins
c17672a33d Further changes to coefficient contexts.
This patch alters the balance of context between the
coefficient bands (reflecting the position of coefficients
within a transform blocks) and the energy of the previous
token (or tokens) within a block.

In this case the number of coefficient bands is reduced
but more previous token energy bands are supported.

Some initial rebalancing of the default tables has been
by running multiple derf clips at multiple data rates using
the ENTOPY_STATS macro. Further balancing needs to be
done using larger image formatsd especially in regard to
the bigger transform sizes which are not as well represented
in encodings of smaller image formats.

Change-Id: If9736e95c391e711b04aef6393d26f60f36e1f8a
2013-02-23 07:29:09 -08:00
James Zern
e5fb6321a1 give vp9 variance struct a unique name
variance_vtable clashed with vp8/common/variance.h

Change-Id: I09c1de44d5519f1bd13f58c01144c0de4706de6f
2013-02-22 16:25:13 -08:00
John Koleszar
606a2561d6 Merge "Code cleanup." into experimental 2013-02-22 11:20:20 -08:00
Dmitry Kovalev
548b4dd5f2 Code cleanup.
Removing redundant 'extern' keywords and parentheses, fixing indentation,
making variable names lower case, using short expressions x *= c
instead of x = x * c, minor code simplifications.

Change-Id: If6a25fcf306d1db26e90d27e3c24a32735c607de
2013-02-22 11:03:14 -08:00
Jingning Han
babbd5d170 Forward butterfly hybrid transform
This patch includes 4x4, 8x8, and 16x16 forward butterfly ADST/DCT
hybrid transform. The kernel of 4x4 ADST is sin((2k+1)*(n+1)/(2N+1)).
The kernel of 8x8/16x16 ADST is of the form sin((2k+1)*(2n+1)/4N).

Change-Id: I8f1ab3843ce32eb287ab766f92e0611e1c5cb4c1
2013-02-21 18:24:28 -08:00
Ronald S. Bultje
35524e2231 Remove "eobs" array in MACROBLOCKD.
The information is a duplicate of "eob" in BLOCKD.

Change-Id: Ia6416273bd004611da801e4bfa6e2d328d6f02a3
2013-02-21 10:07:36 -08:00
Deb Mukherjee
048f593703 Merge "Refactoring of switchable filter search for speed" into experimental 2013-02-21 09:23:50 -08:00
John Koleszar
138ffb6ea9 Merge "Avoid division in intra prediction" into experimental 2013-02-21 08:33:17 -08:00
Deb Mukherjee
28b1db9278 Refactoring of switchable filter search for speed
Refactors the switchable filter search in the rd loop to
improve encode speed.

Uses a piecewise approximation to a closed form expression to estimate
rd cost for a Laplacian source with a given variance and quantization
step-size.

About 40% encode time reduction is achieved.

Results (on a feb 12 baseline) show a slight drop:

derf: -0.019%
yt: +0.010%
std-hd: -0.162%
hd: -0.050%

Change-Id: Ie861badf5bba1e3b1052e29a0ef1b7e256edbcd0
2013-02-20 18:34:42 -08:00
Dmitry Kovalev
e6c89a1f9b Merge "Code cleanup." into experimental 2013-02-20 12:47:54 -08:00
Dmitry Kovalev
eb6aee50a4 Code cleanup.
Change-Id: I7c6e3bebd94856b24dbe2aded7f9e04ef8bb8c08
2013-02-20 11:36:31 -08:00
Yaowu Xu
d262e26cc7 Merge lossless experiment
Change-Id: I7b7b8d4fda3a23699e0c920d727f8c15d37d43aa
2013-02-20 07:54:28 -08:00
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
Jingning Han
cd907b1601 16x16 butterfly inverse ADST/DCT hybrid transform
rebased.

This patch includes 16x16 butterfly inverse ADST/DCT hybrid
transform. It uses the variant ADST of kernel
    sin((2k+1)*(2n+1)/4N),
which allows a butterfly implementation.

The coding gains as compared to DCT 16x16 are about 0.1% for
both derf and std-hd. It is noteworthy that for std-hd sets
many sequences gains about 0.5%, some 0.2%. There are also few
points that provides -1% to -3% performance. Hence the average
goes to about 0.1%.

Change-Id: Ie80ac84cf403390f6e5d282caa58723739e5ec17
2013-02-19 09:07:00 -08:00
Yaowu Xu
93d6b86cfd Use lossless for Q0
The commit changes the coding mode to lossless whenever the lowest
quantizer is choosen.

As expected, test results showed no difference for cif and std-hd
set where Q0 is rarely used. For yt and yt-hd set, Q0 is used for
a number of clips, where this commit helped a lot in the high end.

Average over all clips in the sets:
yt: 2.391% 1.017% 1.066%
hd: 1.937%  .764%  .787%

Change-Id: I9fa9df8646fd70cb09ffe9e4202b86b67da16765
2013-02-19 06:18:42 -08:00
Ronald S. Bultje
3af36ea8cc Remove Y2 and Y-no-DC token types from the bitstream.
Change-Id: I7a5314daca993d46b8666ba1ec2ff3766c1e5042
2013-02-15 14:06:30 -08:00
Ronald S. Bultje
48598e30b1 Remove y2dc/ac Q delta values from the bitstream.
Since there is no Y2, these values are always zero. This changes the
bitstream results slightly, hence a separate commit.

Change-Id: I2f838f184341868f35113ec77ca89da53c4644e0
2013-02-15 14:06:30 -08:00
Ronald S. Bultje
46dff5d233 Remove some Y2-related code.
Change-Id: I4f46d142c2a8d1e8a880cfac63702dcbfb999b78
2013-02-15 14:06:25 -08:00
Scott LaVarnway
7755657ea7 Merge "WIP: ssse3 version of convolve avg functions" into experimental 2013-02-15 07:54:21 -08:00
Scott LaVarnway
ae886d6bff Moved vp9_get_coef_band to header file
allowing the compiler to inline.

Change-Id: I66e5caf5e7fefa68a223ff0603aa3f9e11e35dbb
2013-02-14 12:27:25 -08:00
Paul Wilkins
45712dc8c8 Merge "Abstract selection of coef band." into experimental 2013-02-14 03:23:31 -08:00
Ronald S. Bultje
51afedbe28 Merge "Remove 2nd-order transform for first-order DC coefficients." into experimental 2013-02-13 13:58:02 -08:00
Ronald S. Bultje
89a206ef2f Add support for tile rows.
These allow sending partial bitstream packets over the network before
encoding a complete frame is completed, thus lowering end-to-end
latency. The tile-rows are not independent.

Change-Id: I99986595cbcbff9153e2a14f49b4aa7dee4768e2
2013-02-13 12:31:00 -08:00
Ronald S. Bultje
42d6be8080 Remove 2nd-order transform for first-order DC coefficients.
Since addition of the larger-scale transforms (16x16, 32x32), these
don't give a benefit at macroblock-sizes anymore. At superblock-sizes,
2nd-order transform was never used over the larger transforms. Future
work should test whether there is a benefit for that use case.

Change-Id: I90cadfc42befaf201de3eb0c4f7330c56e33330a
2013-02-13 12:28:19 -08:00
Paul Wilkins
9255ad107f Abstract selection of coef band.
This patch abstracts the selection of the coefficient band
context into a function as a precursor to further experiments
with the coefficient context.

It also removes the large per TX size coefficient band structures
and uses a single matrix for all block sizes within the test function.

This may have an impact on quality (results to follow) but is only an
intermediate step in the process of redefining the context. Also the
quality impact will be larger initially because the default tables will
be out of step with the new banding.

In particular the 4x4 will in this case only use 7 bands. If needed we
can add back block size dependency localized within the function, but
this can follow on after the other changes to the definition of the
context.

Change-Id: Id7009c2f4f9bb1d02b861af85fd8223d4285bde5
2013-02-13 19:01:25 +00:00
Paul Wilkins
0d284ffed1 Abstract the selection of coefficient context.
This is an initial step to facilitate experimentation
with changes to the prior token context used to code
coefficients to take better account of the energy of
preceding tokens.

This patch merely abstracts the selection of context into
two functions and does not alter the output.

Change-Id: I117fff0b49c61da83aed641e36620442f86def86
2013-02-13 18:56:30 +00:00
Paul Wilkins
afa57bfc97 Merge "Remove NEWCOEFCONTEXT experiment." into experimental 2013-02-13 10:41:13 -08:00
Yaowu Xu
f01b08c96c Merge "enable bitstream lossless support" into experimental 2013-02-13 10:26:58 -08:00
Yaowu Xu
d3de97794f Merge "fix the lossless experiment" into experimental 2013-02-13 09:54:35 -08:00
Yaowu Xu
17db5d00be enable bitstream lossless support
1. Added a bit in frame header to  to indicate if a frame is encoded
in lossless mode, so decoder does not make the decision based on Q0
2. Minor changes to make sure that lossy coding works same as when
the lossless experiment is not enabled.
3. Renamed function pointers for transforms to be consistent, using
prefix fwd_txm and inv_txm for forward and inverse respectively

To encode in lossless mode, using "--lossless=1 --min-q=0 --max-q=0"
with vpxenc.

Change-Id: Ifae53b26d2ffbe378d707e29d96817b8a5e6c068
2013-02-13 09:24:39 -08:00
Yaowu Xu
16f25f9dc8 fix the lossless experiment
Change-Id: I95acfc1417634b52d344586ab97f0abaa9a4b256
2013-02-13 09:20:26 -08:00
Scott LaVarnway
30f866f44b WIP: ssse3 version of convolve avg functions
Initial ssse3 convolve avg functions and is one step closer
to using x86inc.asm.  The decoder performance improved by 8% for
the test clip used.  This should be revisited later to see if
averaging outside the loop is better than having many similar
filter functions.

Change-Id: Ice3fafb423b02710b0448ffca18b296bcac649e9
2013-02-13 09:15:38 -08:00
Paul Wilkins
6a9f0c61a4 Remove NEWCOEFCONTEXT experiment.
Removal of the  NEWCOEFCONTEXT experiment to
reduce code clutter and make it easier to experiment with
some other changes to the coefficient coding context.

Change-Id: Icd17b421384c354df6117cc714747647c5eb7e98
2013-02-13 15:12:17 +00:00
Paul Wilkins
649be94cf0 Removal of Hybrid DWT/DCT experiment.
Removal of experiment to simplify code base for other
changes.

Change-Id: If0a33952504558511926ad212bc311fc2bffb19a
2013-02-13 15:08:48 +00:00
John Koleszar
1d60b6bcb5 Merge "Replace as_mv struct with array" into experimental 2013-02-12 13:59:04 -08:00
Ronald S. Bultje
f496f601fb Add tile column size limits (256 pixels min, 4096 pixels max).
This is after discussion with the hardware team. Update the unit test
to take these sizes into account. Split out some duplicate code into
a separate file so it can be shared.

Change-Id: I8311d11b0191d8bb37e8eb4ac962beb217e1bff5
2013-02-12 10:33:34 -08:00
Scott LaVarnway
ff024f812b Merge "Bug fix: ssse3 version of subpixel did not match C code" into experimental 2013-02-12 08:45:24 -08:00
Christian Duvivier
094e2572df Faster convolve8_avg.
Implement convolve8_avg using common functions which are already optimized
instead of using more obscure ones which have only C versions. Encoder
overall speed-up of about 12%.

Change-Id: I8c57aa76936c8a48f22b115f19f61d9f2ae1e4b6
2013-02-11 16:53:11 -08:00
Jingning Han
57e995ff9c butterfly inverse 4x4 ADST
fixed format issues.

Implement the inverse 4x4 ADST using 9 multiplications. For this
particular dimension, the original ADST transform can be
factorized into simpler operations, hence is retained.

Change-Id: Ie5d9749942468df299ab74e90d92cd899569e960
2013-02-11 10:42:39 -08:00
Ronald S. Bultje
5f2e8449b7 Merge "Port sadNxNx4d functions to x86inc.asm." into experimental 2013-02-11 08:20:12 -08:00
Scott LaVarnway
eda30b410e Bug fix: ssse3 version of subpixel did not match C code
A 16 bit overflow condition occurs when using the EIGHTTAP_SMOOTH filters.
(vp9_sub_pel_filters_8lp)  Changed the order of the adds to fix this problem.
Also added ssse3 support for 4x4 subpixel filtering.

Change-Id: I475eaadae920794c2de5e01e9735c059a856518e
2013-02-09 15:15:14 -08:00
Paul Wilkins
e4f949b55a Merge "Nearest / Zero Mv default entropy tweak." into experimental 2013-02-09 04:21:08 -08:00
John Koleszar
7ca517f755 Replace as_mv struct with array
Replace as_mv.{first, second} with a two element array, so that they
can easily be processed with an index variable.

Change-Id: I1e429155544d2a94a5b72a5b467c53d8b8728190
2013-02-08 20:23:35 -08:00
John Koleszar
dc836109e4 Merge "Pass macroblock index to pick inter functions" into experimental 2013-02-08 20:20:37 -08:00
Ronald S. Bultje
c0ce2ab349 Port sadNxNx4d functions to x86inc.asm.
Change-Id: Ic639f5742f7a007753d7a3fa5c66235172eb31d8
2013-02-08 17:59:32 -08:00
Ronald S. Bultje
02ff360b33 Add sad64x64 and sad32x32 SSE2 versions.
Also port the 4x4, 16x16, 8x16 and 16x8 versions to x86inc.asm; this
makes them all slightly faster, particularly on x86-64. Remove SSE3
sad16x16 version, since the SSE2 version is now faster.

About 1.5% overall encoding speedup.

Change-Id: Id4011a78cce7839f554b301d0800d5ca021af797
2013-02-08 16:32:25 -08:00
John Koleszar
6125a1ed81 Pass macroblock index to pick inter functions
Pass the current mb row and column around rather than the
recon_yoffset and recon_uvoffset, since those offsets will
change from predictor to predictor, based on the reference
frame selection.

Change-Id: If3f9df059e00f5048ca729d3d083ff428e1859c1
2013-02-08 14:25:40 -08:00
John Koleszar
3de8ee6ba1 Merge changes Ife0d8147,I7d469716,Ic9a5615f into experimental
* changes:
  Restore SSSE3 subpixel filters in new convolve framework
  Convert subpixel filters to use convolve framework
  Add 8-tap generic convolver
2013-02-08 13:19:47 -08:00
John Koleszar
29d47ac80e Restore SSSE3 subpixel filters in new convolve framework
This commit adds the 8 tap SSSE3 subpixel filters back into the code
underneath the convolve API. The C code is still called for 4x4
blocks, as well as compound prediction modes. This restores the
encode performance to be within about 8% of the baseline.

Change-Id: Ife0d81477075ae33c05b53c65003951efdc8b09c
2013-02-08 12:18:14 -08:00
Paul Wilkins
bbede82f24 Nearest / Zero Mv default entropy tweak.
Tweak to default mode context to account for the fact
that when there are no non zero motion candidates
Nearest is now the preferred mode for coding a 0,0
vector.

Also resolve duplicate function name and typos.

Change-Id: I76802788d46c84e3d1c771be216a537ab7b12817
2013-02-08 10:16:13 +00:00
Yaowu Xu
e6ad9ab02c move dct/idct constants to a header file
also removed some un-unsed functions.

Change-Id: Ie363bcc8d94441d054137d2ef7c4fe59f56027e5
2013-02-07 13:51:45 -08:00
Jingning Han
d15e1da494 Butterfly ADST based hybrid transform
Refactor the 8x8 inverse hybrid transform. It is now consistent
with the new inverse DCT. Overall performance loss (due to the
use of this variant ADST, and the rounding errors in the butterfly
implementation) for std-hd is -0.02.

Fixed BUILD warning.

Devise a variant of the original ADST, which allows butterfly
computation structure. This new transform has kernel of the
form: sin((2k+1)*(2n+1) / (4N)). One of its butterfly structures
using floating-point multiplications was reported in Z. Wang,
"Fast algorithms for the discrete W transform and for the discrete
Fourier transform", IEEE Trans. on ASSP, 1984.

This patch includes the butterfly implementation of the inverse
ADST/DCT hybrid transform of dimension 8x8.

Change-Id: I3533cb715f749343a80b9087ce34b3e776d1581d
2013-02-07 10:07:46 -08:00
Ronald S. Bultje
aac73df1a7 Use configure checks for various inline keywords.
Change-Id: I8508f1a3d3430f998bb9295f849e88e626a52a24
2013-02-06 16:12:56 -08:00
Ronald S. Bultje
a788e0fe63 Add sse2 versions of sub_pixel_variance{32x32,64x64}.
7.5% faster overall encoding.

Change-Id: Ie9bb7f9fdf93659eda106404cb342525df1ba02f
2013-02-06 11:20:59 -08:00
Paul Wilkins
8b4e9c5925 Merge "Change definition of NearestMV." into experimental 2013-02-06 04:06:31 -08:00
Ronald S. Bultje
1407bdc243 [WIP] Add column-based tiling.
This patch adds column-based tiling. The idea is to make each tile
independently decodable (after reading the common frame header) and
also independendly encodable (minus within-frame cost adjustments in
the RD loop) to speed-up hardware & software en/decoders if they used
multi-threading. Column-based tiling has the added advantage (over
other tiling methods) that it minimizes realtime use-case latency,
since all threads can start encoding data as soon as the first SB-row
worth of data is available to the encoder.

There is some test code that does random tile ordering in the decoder,
to confirm that each tile is indeed independently decodable from other
tiles in the same frame. At tile edges, all contexts assume default
values (i.e. 0, 0 motion vector, no coefficients, DC intra4x4 mode),
and motion vector search and ordering do not cross tiles in the same
frame.
t log

Tile independence is not maintained between frames ATM, i.e. tile 0 of
frame 1 is free to use motion vectors that point into any tile of frame
0. We support 1 (i.e. no tiling), 2 or 4 column-tiles.

The loopfilter crosses tile boundaries. I discussed this briefly with Aki
and he says that's OK. An in-loop loopfilter would need to do some sync
between tile threads, but that shouldn't be a big issue.

Resuls: with tiling disabled, we go up slightly because of improved edge
use in the intra4x4 prediction. With 2 tiles, we lose about ~1% on derf,
~0.35% on HD and ~0.55% on STD/HD. With 4 tiles, we lose another ~1.5%
on derf ~0.77% on HD and ~0.85% on STD/HD. Most of this loss is
concentrated in the low-bitrate end of clips, and most of it is because
of the loss of edges at tile boundaries and the resulting loss of intra
predictors.

TODO:
- more tiles (perhaps allow row-based tiling also, and max. 8 tiles)?
- maybe optionally (for EC purposes), motion vectors themselves
  should not cross tile edges, or we should emulate such borders as
  if they were off-frame, to limit error propagation to within one
  tile only. This doesn't have to be the default behaviour but could
  be an optional bitstream flag.

Change-Id: I5951c3a0742a767b20bc9fb5af685d9892c2c96f
2013-02-05 15:43:03 -08:00
Ronald S. Bultje
822864131b Merge "Add SSE3 versions for sad{32x32,64x64}x4d functions." into experimental 2013-02-05 15:40:46 -08:00
Yaowu Xu
c9ae73b251 Merge "rewrite 4x4 idct and fdct" into experimental 2013-02-05 15:26:36 -08:00
Ronald S. Bultje
58c983d109 Add SSE3 versions for sad{32x32,64x64}x4d functions.
Overall encoding about 15% faster.

Change-Id: I176a775c704317509e32eee83739721804120ff2
2013-02-05 15:21:47 -08:00
John Koleszar
7a07eea13f Convert subpixel filters to use convolve framework
Update the code to call the new convolution functions to do subpixel
prediction rather than the existing functions. Remove the old C and
assembly code, since it is unused. This causes a 50% performance
reduction on the decoder, but that will be resolved when the asm for
the new functions is available.

There is no consensus for whether 6-tap or 2-tap predictors will be
supported in the final codec, so these filters are implemented in
terms of the 8-tap code, so that quality testing of these modes
can continue. Implementing the lower complexity algorithms is a
simple exercise, should it be necessary.

This code produces slightly better results in the EIGHTTAP_SMOOTH
case, since the filter is now applied in only one direction when
the subpel motion is only in one direction. Like the previous code,
the filtering is skipped entirely on full-pel MVs. This combination
seems to give the best quality gains, but this may be indicative of a
bug in the encoder's filter selection, since the encoder could
achieve the result of skipping the filtering on full-pel by selecting
one of the other filters. This should be revisited.

Quality gains on derf positive on almost all clips. The only clip
that seemed to be hurt at all datarates was football
(-0.115% PSNR average, -0.587% min). Overall averages 0.375% PSNR,
0.347% SSIM.

Change-Id: I7d469716091b1d89b4b08adde5863999319d69ff
2013-02-05 14:23:17 -08:00
John Koleszar
5ca6a3667f Add 8-tap generic convolver
This commit introduces a new convolution function which will be used to
replace the existing subpixel interpolation functions. It is much the
same as the existing functions, but allows for changing the filter
kernel on a per-pixel basis, and doesn't bake in knowledge of the
filter to be applied or the size of the resulting block into the
function name.

Replacing the existing subpel filters will come in a later commit.

Change-Id: Ic9a5615f2f456cb77f96741856fc650d6d78bb91
2013-02-05 14:19:28 -08:00
Yaowu Xu
fa36981ec8 rewrite 4x4 idct and fdct
This commit changes the 4x4 iDCT to use same algorithm & constants as
other iDCTs. The 4x4 fDCT is also changed to be based on the new iDCT.

Change-Id: Ib1a902693228af903862e1f5a08078c36f2089b0
2013-02-05 11:42:49 -08:00
Paul Wilkins
81043e8d62 Change definition of NearestMV.
This commit makes the NearestMV match the chosen
best reference MV. It can be a 0,0 or non zero vector
which means the the compound nearest mv mode can
combine a 0,0 and a non zero vector.

Change-Id: I2213d09996ae2916e53e6458d7d110350dcffd7a
2013-02-05 17:03:25 +00:00
Scott LaVarnway
5780c4cbd5 Added vp9_short_idct1_32x32_c
and called this function in vp9_dequant_idct_add_32x32_c when
eob == 1.  For the test clip used, the decoder performance improved
by 21+%.  Based on Yaowu's 16 point idct work.

Change-Id: Ib579a90fed531d45777980e04bf0c9b23c093c43
2013-02-04 16:49:17 -08:00
Yaowu Xu
1eb79dc1dc re-write 8 point idct
to be consistent with idct16 and idct32.

Change-Id: Ie89dbd32b65c33274b7fecb4b41160fcf1962204
2013-02-04 07:31:25 -08:00
Yaowu Xu
ccaaeb4b5a a couple of minor fixes
fixed a function prototypes to prevent compiler warnings;
removed a function not in use;
un-capitialize "Refstride" to ref_stride

Change-Id: Ib4472b6084f357d96328c6a06e795b6813a9edba
2013-02-04 07:19:32 -08:00
Yaowu Xu
af4c9d2f88 Merge "Changes 16 point idct" into experimental 2013-02-01 08:22:20 -08:00
Yaowu Xu
91e0e80142 Changes 16 point idct
This commit changes the inverse 16 point dct to use the same algorithm
as the one for 32 point idct. In fact, now 16 point dct uses the exact
version of the souce code for even portion of the 32 point idct.

Tests showed current implementation has significant better accuracy
than the previous version. With this implementation and the minor bug
fix on forward 16 point dct, encoding tests showed about 0.2% better
compression of CIF set, test results on std-hd setting pending.

Change-Id: I68224b60c816ba03434e9f08bee147c7e344fb63
2013-01-31 19:52:18 -08:00
Yaowu Xu
c94e55add0 Merge "A fix point implementation of 32x32 idct" into experimental 2013-01-31 10:48:01 -08:00
Yaowu Xu
5149d7f7bd A fix point implementation of 32x32 idct
This commit changes the 32x32 idct to use integer only. The algorithm
was taken directly from "A Fast Computational Algorithm for the
Discrete Cosine Tranform" by W. Chen, et al., which was published in
IEEE Transaction on Communication Vol. Com.-25 No. 9, 1977. The signal
flow graph in the original paper is for a 32 point forward dct, the
current implementation of inverse DCT was done by follow the graph in
reversed direction.

With this implementation, the 32 point inverse dct contains a 16 point
inverse dct in its even portion, similarly the 16 point idct further
contains 8 point and 4 point inverse dcts.

As of patch 4, encoding tests showed there is no compression loss when
compared against the floating point baseline. Numbers even showed very
small postives. (cif: .01%, std-hd: .05%).

Change-Id: I2d2d17a424b0b04b42422ef33ec53f5802b0f378
2013-01-31 09:45:49 -08:00
Deb Mukherjee
a53be60904 Merge "Adding a frame parallel decoding mode" into experimental 2013-01-30 12:03:45 -08:00
Ronald S. Bultje
b499c24c2f Merge "don't code the branch for the predicted seg_id if that flag is false." into experimental 2013-01-30 10:02:51 -08:00
Ronald S. Bultje
3a4b18bc67 don't code the branch for the predicted seg_id if that flag is false.
Change-Id: Icb6e21dc0c2d9918faa33c8bf70943660df7ad88
2013-01-30 09:30:46 -08:00
Deb Mukherjee
d28750537e Merge "Further improvement on compound inter-intra expt" into experimental 2013-01-30 08:38:17 -08:00
Frank Galligan
0524f33108 libvpx: Fix warnings on windows.
Warnings found when tyring to build libvpx in Chromium.

Change-Id: I5824d9e2c06351e0cf46e9f5fa102cc8b04cf963
2013-01-29 13:57:09 -08:00
Ronald S. Bultje
64401f838f Merge "Fix overread/write reported by valgrind if (mb_cols) & 3 != 0." into experimental 2013-01-29 12:49:22 -08:00
Deb Mukherjee
3b04d467ac Further improvement on compound inter-intra expt
Adds a special combination mode specific to intra prediciton
mode D45.

Current results with the compound inter/intra experiment:
derf: 0.2%
yt: 0.55%
std-hd: 0.75%
hd: 0.74%

Change-Id: I8976bdf3b9b0b66ab8c5c628bbc62c14fc72ca86
2013-01-29 00:21:29 -08:00
Paul Wilkins
0ff9b033b0 Segment Skip Flag
First step in simplifying the segment mode and
segment EOB flags into a simpler segment skip
flag that implies 0,0 mv and EOB at position 0.

Change-Id: Ib750cac31a7a02dc21082580498efd9f7d8d72a5
2013-01-28 17:28:04 +00:00
Ronald S. Bultje
9dc9f07fb8 Fix overread/write reported by valgrind if (mb_cols) & 3 != 0.
We'd backup and restore all cols for a 64x64 SB, but the array wouldn't
be big enough to hold all that data.

Change-Id: Ic68ea721bf07e0b2f3937bd16b0b734bcc743ce1
2013-01-25 17:18:08 -08:00