James Zern [Wed, 29 Mar 2017 02:39:21 +0000 (02:39 +0000)]
Merge changes from topic 'sync-highbd-intrapred'
* changes:
intrapred: sync highbd_d135_predictor w/d135_
intrapred: specialize highbd 4x4 predictors
intrapred: rename d63f to d63e
remove CONFIG_MISC_FIXES
James Zern [Wed, 29 Mar 2017 01:56:57 +0000 (01:56 +0000)]
Merge ".mailmap: add an additional entry for Yaowu Xu"
Johann Koenig [Tue, 28 Mar 2017 21:50:00 +0000 (21:50 +0000)]
Merge "vpx_comp_avg_pred: add test"
Johann [Thu, 23 Mar 2017 20:04:28 +0000 (13:04 -0700)]
vpx_comp_avg_pred: add test
BUG=webm:1389
Change-Id: I23cd65f1939db026958ccb5d70b8c5cc9aa5bc51
Marco [Fri, 24 Mar 2017 18:31:19 +0000 (11:31 -0700)]
vp9: Speed >= 8: avoid chrome check under some condition.
For non-rd variance partition, avoid the chrome check
unless y_sad is below some threshold.
Small decrease in avgPSNR (~0.3) on RTC set.
Small/negligible decrease on RTC_derf.
Change-Id: I7af44235af514058ccf9a4f10bb737da9d720866
Marco [Thu, 23 Mar 2017 21:37:47 +0000 (14:37 -0700)]
vp9: 1 pass: Move source sad computation into encodeframe loop.
Refactor to split the 1 passs source sad computation into scene
detection (currently used for VBR and screen-content mode), and
superblock based source sad computation (used in non-rd CBR mode).
This allows the source sad computation for CBR mode to be
multi-threaded.
No change in compression.
Change-Id: I112f2918613ccbd37c1771d852606d3af18c1388
James Zern [Tue, 21 Mar 2017 03:24:30 +0000 (20:24 -0700)]
intrapred: sync highbd_d135_predictor w/d135_
previously:
05437805f intrapred/d135: flatten border results before storing
BUG=webm:1316
Change-Id: I3b8bd89117ad7f2f4560b57f7c148da781e86f85
James Zern [Sat, 25 Mar 2017 02:24:39 +0000 (19:24 -0700)]
intrapred: specialize highbd 4x4 predictors
d207/d63/d45/d117/d135/d153
~9-45% better depending on the predictor on 32-bit ARM, similar range on
x86-64
this matches the non-highbitdepth implementation
BUG=webm:1316
Change-Id: Iddebdf7c58c6f31c47cae04da95c6e5318200e4c
James Zern [Sat, 25 Mar 2017 02:22:41 +0000 (19:22 -0700)]
intrapred: rename d63f to d63e
this is consistent with he/ve/d45e
Change-Id: I75641ae5667430b0ecd370db86fff6e666cb577d
James Zern [Sat, 25 Mar 2017 01:56:15 +0000 (18:56 -0700)]
remove CONFIG_MISC_FIXES
this belonged to vp10 with the changes now migrated to av1.
Change-Id: Ie30ead3e7b71f465bc14136e1b6f156ea978c43f
Marco [Thu, 23 Mar 2017 21:32:43 +0000 (14:32 -0700)]
vp9: Fix to condition on using source_sad for 1 pass real-time.
Make the source_sad feature work properly for cases of VBR or
screen_content with SVC.
Added unittest for SVC with screen-content on.
Change-Id: Iba5254fd8833fb11da521e00cc1317ec81d3f89b
Johann Koenig [Thu, 23 Mar 2017 18:38:35 +0000 (18:38 +0000)]
Merge "vp9 temporal filter: additional test"
Alex Converse [Thu, 23 Mar 2017 18:05:13 +0000 (18:05 +0000)]
Merge changes Ie989e60c,Ifc110b12
* changes:
Backport "Optimize the use case of token_cost table" to VP9
Drop vp9_get_token_extracost
Marco Paniconi [Thu, 23 Mar 2017 14:17:48 +0000 (14:17 +0000)]
Merge "vp9: Non-rd partition: avoid unneeded call to chrome_check"
Kaustubh Raste [Thu, 23 Mar 2017 07:44:16 +0000 (07:44 +0000)]
Merge "Fix mips msa fwd xform mismatch"
Marco [Thu, 23 Mar 2017 05:36:19 +0000 (22:36 -0700)]
vp9: Non-rd partition: avoid unneeded call to chrome_check
Since y_sad is not computed yet (on the early exit due to source_sad),
no need to check for setting color_sensitiviy.
Only affects speed >=8. No change in behavior.
Change-Id: I3a6f2d20fed38d8b8ec51b75bcacf9a21f2db916
James Zern [Thu, 23 Mar 2017 00:53:21 +0000 (00:53 +0000)]
Merge "vp9_rdopt: correct size to vpx_sum_squares_2d_i16"
Marco Paniconi [Wed, 22 Mar 2017 22:56:17 +0000 (22:56 +0000)]
Merge "vp9: Adjust some speed settings for speed 8."
Marco [Wed, 22 Mar 2017 19:15:06 +0000 (12:15 -0700)]
vp9: Adjust some speed settings for speed 8.
Allow for simple_block_rd for VGA resoln, and reduce
adaptive_rd_thresh to 1.
On average no loss on RTC set, ~4% speedup on mac.
Change-Id: Ib549c4061c853776062b5e34040f839d470fbebc
Jerome Jiang [Wed, 22 Mar 2017 22:02:24 +0000 (22:02 +0000)]
Merge "vp9: Enable adaptive_rd_threshold for row mt for realtime speed 8."
Johann [Wed, 22 Mar 2017 18:04:20 +0000 (11:04 -0700)]
vp9 temporal filter: additional test
Change tests to reflect use. Input sizes will be 8 or 16 (but not
necessarily square).
filter_weight is capped at 2 and filter_strength at 6
Speed test, disabled by default.
Change-Id: Idfde9d6c4b7d93aaf0e641b0f4862c15e2a2af7a
Johann Koenig [Wed, 22 Mar 2017 19:36:40 +0000 (19:36 +0000)]
Merge "vp9 temporal filter: add const to function prototype"
James Zern [Wed, 22 Mar 2017 19:30:11 +0000 (19:30 +0000)]
Merge "idct_neon: prefix non-static functions w/'vpx_'"
James Zern [Wed, 22 Mar 2017 06:29:12 +0000 (23:29 -0700)]
vp9_rdopt: correct size to vpx_sum_squares_2d_i16
the current implementations expect pixel size, not the block type
BUG=webm:1392
Change-Id: Ib91e9f30a1f56e13566b1fb76f089dae9bb50cdc
James Zern [Tue, 21 Mar 2017 05:46:53 +0000 (22:46 -0700)]
idct_neon: prefix non-static functions w/'vpx_'
Change-Id: I94fcdeae18468e6ef0cb7119b8142d982a048031
Johann [Wed, 22 Mar 2017 17:15:22 +0000 (10:15 -0700)]
vp9 temporal filter: add const to function prototype
The input frames are not modified.
Change-Id: Ideb810e3c5afeb4dbdc4c7d54024c43a8129ad39
Kaustubh Raste [Wed, 22 Mar 2017 08:31:03 +0000 (14:01 +0530)]
Fix mips msa fwd xform mismatch
Change-Id: I32a6df11463144aa1a562256ee7d57a41fd678d6
Jerome Jiang [Tue, 14 Mar 2017 20:45:20 +0000 (13:45 -0700)]
vp9: Enable adaptive_rd_threshold for row mt for realtime speed 8.
Change it to row based array to avoid the slow down cause by sync.
row-mt on, speed 8, 2 threads: ~4% speedup for VGA on ARM benefited
from adaptive_rd_threshold.
Change-Id: I887e65a53af20a6c4f48d293daaee09dab3512cf
Marco Paniconi [Tue, 21 Mar 2017 23:44:05 +0000 (23:44 +0000)]
Merge "vp9: Modify datarate tests to cover denoising with multi-threading."
Jerome Jiang [Tue, 21 Mar 2017 23:27:48 +0000 (23:27 +0000)]
Merge "Fix the data race caused by vp9 denoiser."
Marco [Tue, 21 Mar 2017 05:15:13 +0000 (22:15 -0700)]
vp9: Modify datarate tests to cover denoising with multi-threading.
Change-Id: I6ed48a630edf9923c25a05deaca50e0afec43918
Jerome Jiang [Tue, 21 Mar 2017 22:33:42 +0000 (15:33 -0700)]
Fix the data race caused by vp9 denoiser.
BUG=webm:1391
Change-Id: I9669ae62fe9c695d4c6f9973094cb0f39bed51c7
Yi Luo [Tue, 21 Mar 2017 22:32:20 +0000 (22:32 +0000)]
Merge "Make butterfly_self() signature consistent with butterfly()"
Yunqing Wang [Tue, 21 Mar 2017 16:47:55 +0000 (09:47 -0700)]
Code refactoring in the partition search
Computed the partition search early termination score in a separate
function.
Change-Id: I1894b517ff179a38b1c05e054d373ac4b7f4cbb4
Yi Luo [Tue, 21 Mar 2017 00:18:10 +0000 (17:18 -0700)]
Make butterfly_self() signature consistent with butterfly()
- Refer to patch:
48fca113d inv_txfm_ssse3,butterfly: fix win32 abi
compatibility.
- Change four butterfly() calls to butterfly_self(), to simplify the
operations.
Change-Id: Ib2a8cfe6cddcaf0a59e6e6270d8380055ea42ef3
James Zern [Tue, 21 Mar 2017 05:39:40 +0000 (22:39 -0700)]
.mailmap: add an additional entry for Yaowu Xu
Change-Id: I26b928848a7e72ff5ca25001ed6991c95f5992a5
James Zern [Tue, 21 Mar 2017 03:27:35 +0000 (03:27 +0000)]
Merge "Add vpx_highbd_idct32x32_1024_add_neon()"
James Zern [Tue, 21 Mar 2017 03:02:50 +0000 (03:02 +0000)]
Merge "Add vpx_highbd_idct32x32_34_add_neon()"
Marco Paniconi [Tue, 21 Mar 2017 00:17:35 +0000 (00:17 +0000)]
Merge "vp9: Nonrd variance partition: improve split to 16x16."
Yunqing Wang [Mon, 20 Mar 2017 23:20:12 +0000 (23:20 +0000)]
Merge "Record the sum of tx block eobs in the partition block"
Marco [Mon, 20 Mar 2017 16:16:23 +0000 (09:16 -0700)]
vp9: Nonrd variance partition: improve split to 16x16.
Add additional condition to split to 16x16, for resolutions <= 360p,
reduces dragging artifact near moving boundary.
Small/no change on RTC metrics.
Change-Id: I314694f2166435d918f74e7ab42f002b07f40dae
Marco Paniconi [Mon, 20 Mar 2017 21:35:11 +0000 (21:35 +0000)]
Merge "vp9: Use sb content measure to bias against golden."
Marco [Thu, 16 Mar 2017 22:55:33 +0000 (15:55 -0700)]
vp9: Use sb content measure to bias against golden.
For each superblock, keep track of how far from current frame
was the last significant content change, and use that (along
with GF distance), to turnoff GF search in non-rd pickmode.
Only enabled for speed >= 8.
avgPNSR on RTC/RTC_derf down by ~0.9/1.2.
Speedup on mac: ~3-5%.
Speedup on arm: 3.6% for VGA and 4.4% for HD.
Change-Id: Ic3f3d6a2af650aca6ba0064d2b1db8d48c035ac7
Johann Koenig [Mon, 20 Mar 2017 19:05:54 +0000 (19:05 +0000)]
Merge "temporal filter test: update types"
Yunqing Wang [Thu, 16 Mar 2017 22:45:07 +0000 (15:45 -0700)]
Record the sum of tx block eobs in the partition block
The sum of tx bloxk eobs is needed in the machine learning based partition
early termination. The eobs are first accumulated during tx search, and
then the value associated with the best tx_size is copied to ctx for later
use.
After the sum of eobs are calculated correctly, re-enabled
ml_partition_search_early_termination speed feature.
Re-did the quality/speed test to check the impact of the fix.
1. Borg test BDRATE result:
4k set: PSNR: +0.183%; SSIM: +0.100%;
hdres set: PSNR: +0.168%; SSIM: +0.256%;
midres set: PSNR: +0.186%; SSIM: +0.326%;
2.Average speed gain result:
4k clips: 21%;
hd clips: 26%;
midres clips: 15%.
The result is in line with the original result.
Change-Id: I4209a95c89be03b4cbfb6a95b16885f89feddbda
Jingning Han [Fri, 17 Mar 2017 19:34:54 +0000 (12:34 -0700)]
Backport "Optimize the use case of token_cost table" to VP9
cherry picked from nextgenv2
90ea281f29df747282e56d3068a3ddbdde30cdd0
Change-Id: Ie989e60c6479ac3251cadaac9c7e795ccba52f4e
Alex Converse [Fri, 17 Mar 2017 23:24:45 +0000 (16:24 -0700)]
Drop vp9_get_token_extracost
vp9_get_token_cost does the same thing with one fewer lookup.
Change-Id: Ifc110b12403cb1a04a3f91357ab435c67b4815d6
James Zern [Fri, 17 Mar 2017 21:48:20 +0000 (21:48 +0000)]
Merge "inv_txfm_sse2: clear conversion warning in hbd build"
Johann [Fri, 17 Mar 2017 20:19:45 +0000 (13:19 -0700)]
temporal filter test: update types
Use 'int' for w/h since it is that way everywhere else.
Pass Buffer pointers
Change-Id: I9eef6890af657baba171c6bcfcc85fc976173399
Johann Koenig [Fri, 17 Mar 2017 18:18:05 +0000 (18:18 +0000)]
Merge "test: add vp9_temporal_filter_apply test"
Alex Converse [Fri, 17 Mar 2017 16:18:20 +0000 (16:18 +0000)]
Merge "vp9_optimize_b: Combine extrabits cost with token lookup"
James Zern [Fri, 17 Mar 2017 08:16:38 +0000 (01:16 -0700)]
inv_txfm_sse2: clear conversion warning in hbd build
tran_high -> tran_low in return from dct_const_round_shift()
Change-Id: I2fe06c4b604823b1d1fe40a487017c3c2819a440
Linfeng Zhang [Wed, 15 Mar 2017 18:31:35 +0000 (11:31 -0700)]
Add vpx_highbd_idct32x32_1024_add_neon()
BUG=webm:1301
Change-Id: Ib90af0c1712e56b301d0e981dbe9a641e15e36ca
Linfeng Zhang [Tue, 14 Mar 2017 21:07:25 +0000 (14:07 -0700)]
Add vpx_highbd_idct32x32_34_add_neon()
BUG=webm:1301
Change-Id: I74dd16c6c64e7bb71aa991cedccddf0663ef5e06
James Zern [Fri, 17 Mar 2017 07:26:52 +0000 (07:26 +0000)]
Merge "Add vpx_highbd_idct32x32_135_add_neon()"
Linfeng Zhang [Tue, 14 Mar 2017 17:16:35 +0000 (10:16 -0700)]
Add vpx_highbd_idct32x32_135_add_neon()
BUG=webm:1301
Change-Id: I58c2d65d385080711c3666d6d8f9d241dac7b21a
James Zern [Fri, 17 Mar 2017 05:24:57 +0000 (05:24 +0000)]
Merge "Clean vpx_idct32x32_1024_add_neon()"
Marco [Fri, 17 Mar 2017 00:05:42 +0000 (17:05 -0700)]
vp9: Fix speed 8 condition for enabling copy_partition.
Change-Id: I2c090e6ba853a30fef1957b620853315f9471753
Alex Converse [Thu, 16 Mar 2017 23:34:26 +0000 (16:34 -0700)]
vp9_optimize_b: Combine extrabits cost with token lookup
About 0.6% fewer cycles spent in vp9_optimize_b.
Change-Id: I2ae62a78374c594ed81d4e3100a5848e2f6f2c4e
Gabriel Marin [Wed, 14 Dec 2016 20:07:34 +0000 (12:07 -0800)]
Add a vector form of routine vp9_model_rd_from_var_lapndz
Add routine vp9_model_rd_from_var_lapndz_vec and call it from model_rd_for_sb
to model the rate and distortion for MAX_MB_PLANE Laplacian sources in
parallel. The caller ensures that all sources have non-zero variance.
Measured a 18% to 25% reduction in retired instructions, and 17% to 24%
reduction in instruction execution cost with different compilers for the
Laplacian modeling.
No change in behavior.
TEST=Verified that encoded files match bit for bit, with and without this
change.
BUG=b/
33678225
Change-Id: I6b76947f21c659a349adb896e13e99f6e3f951e6
Marco Paniconi [Thu, 16 Mar 2017 21:53:38 +0000 (21:53 +0000)]
Merge "vp9: Fixes in non-rd pickmode for denoising with SVC."
Johann Koenig [Thu, 16 Mar 2017 21:53:17 +0000 (21:53 +0000)]
Merge "Remove ppc-linux-gcc target"
Johann Koenig [Thu, 16 Mar 2017 21:52:15 +0000 (21:52 +0000)]
Merge "Add Hadamard for Power8"
Marco [Thu, 16 Mar 2017 19:47:44 +0000 (12:47 -0700)]
vp9: Fixes in non-rd pickmode for denoising with SVC.
Don't denoise spatial layer frames whose base layer is a key frame.
Disallow golden reference for SVC with denoising on frames
that will be denoised (highest layer), as this removes bad artifact.
Will re-enable when issue is resolved.
Change-Id: I87a6597812330500966458172acfce54af65f70f
Marco [Tue, 14 Mar 2017 17:38:50 +0000 (10:38 -0700)]
vpx_codec.h: include vpx/*.h -> ./*.h
This matches the other includes and also fixes a compile issue in
chromium.
Change-Id: I45e00a1454f7ed948aa3b96b04cc5946b1d02985
Jerome Jiang [Thu, 16 Mar 2017 16:43:41 +0000 (16:43 +0000)]
Merge "Refactor: Change cpi->resize_state to enum values."
Marco Paniconi [Thu, 16 Mar 2017 05:13:38 +0000 (05:13 +0000)]
Merge "vp8: Fix compiler warning in vp8 pickinter.c"
Rafael de Lucena Valle [Thu, 20 Oct 2016 00:21:09 +0000 (22:21 -0200)]
Add Hadamard for Power8
Change-Id: I3b4b043c1402b4100653ace4869847e030861b18
Signed-off-by: Rafael de Lucena Valle <rafaeldelucena@gmail.com>
Marco Paniconi [Thu, 16 Mar 2017 02:42:55 +0000 (02:42 +0000)]
Merge "vp9: Fix some issues with denoiser and SVC."
Marco [Wed, 15 Mar 2017 23:51:34 +0000 (16:51 -0700)]
vp9: Fix some issues with denoiser and SVC.
Fix the update of the denoiser buffer when the base
spatial layer is a key frame. And allow for better/lower
QP on high spatial layers when their base layer is key frame.
Change-Id: I96b2426f1eaa43b8b8d4c31a68b0c6d68c3024a2
Jerome Jiang [Mon, 13 Mar 2017 21:08:32 +0000 (14:08 -0700)]
Refactor: Change cpi->resize_state to enum values.
Change-Id: Iab1409b0fc1175bc5a14afc4749a08c536c98c41
Marco [Wed, 15 Mar 2017 20:44:26 +0000 (13:44 -0700)]
vp9: Turn off ml_partition_search_early_termination.
Fails on nightly ubsan, valgrind tests.
Enabled on commit:6701014
Change-Id: Ied3f5cb38e39cba54ac134f4514107cdfdfce159
Marco [Wed, 15 Mar 2017 18:44:07 +0000 (11:44 -0700)]
vp8: Fix compiler warning in vp8 pickinter.c
Change-Id: I0e5714538fe53d885a2201d808846901ae8fc288
Linfeng Zhang [Tue, 14 Mar 2017 22:14:34 +0000 (15:14 -0700)]
Clean vpx_idct32x32_1024_add_neon()
Change-Id: I05921e16d6a3e4e7e5b00a90624735050a186636
Yi Luo [Wed, 15 Mar 2017 02:32:52 +0000 (02:32 +0000)]
Merge "Improve idct32x32_1024_add SSSE3 intrinsics performance"
Linfeng Zhang [Wed, 15 Mar 2017 00:38:17 +0000 (00:38 +0000)]
Merge "Fix overflow issue in 32x32 idct NEON intrinsics"
Jerome Jiang [Wed, 15 Mar 2017 00:03:52 +0000 (00:03 +0000)]
Merge "vp9: Using source sad for speedup for dynamic resizing."
Linfeng Zhang [Tue, 14 Mar 2017 16:31:52 +0000 (09:31 -0700)]
Fix overflow issue in 32x32 idct NEON intrinsics
Similar issue as Change bc1c18e.
The PartialIDctTest.ResultsMatch test on vpx_idct32x32_135_add_neon()
in high bit-depth mode exposes 16-bit overflow in final stage of pass
2, when changing the test number from 1,000 to 1,000,000.
Change to use saturating add/sub for vpx_idct32x32_34_add_neon(),
vpx_idct32x32_135_add_neon and vpx_idct32x32_1024_add_neon() in high
bit-depth mode.
Change-Id: Iaec0e9aeab41a3fdb4e170d7e9b3ad1fda922f6f
Jerome Jiang [Tue, 14 Mar 2017 23:29:46 +0000 (23:29 +0000)]
Merge "vp9: Enable row multithreading for SVC in real-time mode."
Jerome Jiang [Mon, 13 Mar 2017 22:27:02 +0000 (15:27 -0700)]
vp9: Using source sad for speedup for dynamic resizing.
Only for speed >= 7.
Change-Id: I3ac85fbb4023cf7e6f8333806b345b0174382a09
Yi Luo [Mon, 6 Mar 2017 23:11:49 +0000 (15:11 -0800)]
Improve idct32x32_1024_add SSSE3 intrinsics performance
- Function level speed improves ~12%.
Change-Id: I9b7dbddabf08c7d0f6b25264e6074d5ccbe39290
James Zern [Tue, 14 Mar 2017 19:21:42 +0000 (19:21 +0000)]
Merge "vp9/encoder: fix segfault on win32 using vs < 2015"
Yunqing Wang [Tue, 14 Mar 2017 18:07:05 +0000 (18:07 +0000)]
Merge "Apply machine learning-based early termination in VP9 partition search"
Marco Paniconi [Tue, 14 Mar 2017 17:50:17 +0000 (17:50 +0000)]
Merge "vp9: Speed >= 8: Enable simple_block_yrd speed feature."
Marco [Tue, 14 Mar 2017 16:17:06 +0000 (09:17 -0700)]
vp9: Adjust copy partition threshold, for speed 8.
Reduce it from 5 to 4, small/no change in metrics or speed.
Small reduction in dragging artifact near moving head.
Change-Id: Ic3bc5ca67c70bf0c89fc2ed14454840a28ae5b6a
Marco [Mon, 13 Mar 2017 05:38:52 +0000 (22:38 -0700)]
vp9: Speed >= 8: Enable simple_block_yrd speed feature.
Enable speed feature for resolutions > VGA.
avgPSNR on RTC down by ~1.7%.
Speedup on ARM: ~5%.
Change-Id: I7a3fe5f7425aa8df3f4a2eced1afa355bc0d4c95
Johann [Mon, 13 Mar 2017 21:54:38 +0000 (14:54 -0700)]
test: add vp9_temporal_filter_apply test
Add an independent implementation of the filter.
BUG=webm:1379
Change-Id: I309c459b493c3011273b78b127a786bb23c59f9c
Marco Paniconi [Mon, 13 Mar 2017 19:18:30 +0000 (19:18 +0000)]
Merge "vp9: Fix to source_sad feature for SVC."
Linfeng Zhang [Mon, 13 Mar 2017 18:49:01 +0000 (18:49 +0000)]
Merge "Add vpx_highbd_idct32x32_135_add_c()"
Marco [Wed, 8 Mar 2017 18:57:48 +0000 (10:57 -0800)]
vp9: Fix to source_sad feature for SVC.
Allow speed feature sf->use_source_sad to be used
on highest spatial layer for SVC.
Change-Id: I260eb0478902764f49f83e43b17024fe86ff3b22
Yunqing Wang [Mon, 27 Feb 2017 22:26:15 +0000 (14:26 -0800)]
Apply machine learning-based early termination in VP9 partition search
This patch was based on Yang Xian's intern project code. Further modifications
were done.
1. Moved machine-learning related parameters into the context structure.
2. Corrected the calculation of sum_eobs.
3. Removed unused parameters and calculations.
4. Made it work with multiple tiles.
5. Added a speed feature for the machine-learning based partition search
early termination.
6. Re-organized the code.
The patch was rebased to the top-of-tree.
Borg test BDRATE result:
4k set: PSNR: +0.144%; SSIM: +0.043%;
hdres set: PSNR: +0.149%; SSIM: +0.269%;
midres set: PSNR: +0.127%; SSIM: +0.257%;
Average speed gain result:
4k clips: 22%;
hd clips: 23%;
midres clips: 15%.
Change-Id: I0220e93a8277e6a7ea4b2c34b605966e3b1584ac
Marco Paniconi [Mon, 13 Mar 2017 06:11:12 +0000 (06:11 +0000)]
Merge "vp9: Fix condition for intra search in non-rd pickmode."
Marco [Sat, 11 Mar 2017 06:50:43 +0000 (22:50 -0800)]
vp9: Fix condition for intra search in non-rd pickmode.
Fixes an issue when the LAST and golden is not used as a reference,
in which case its possible no encoding mode is set (since intra may be
skipped under certain codtions). Fix is to make sure intra is searched
if no inter mode is checked.
Issue can happen for temporal layer pattern#7 in vpx_temporal_svc_encoder.c
Change-Id: I5ab4999b2f9dbd739044888e0916b5ec491d966b
James Zern [Fri, 10 Mar 2017 07:29:54 +0000 (23:29 -0800)]
inv_txfm_ssse3,butterfly: fix win32 abi compatibility
only the first 3 parameters can be aligned to 16 as required by __m128i,
make them all pointers for consistency.
since:
07c48ccfe Improve idct32x32_34_add SSSE3 intrinsics performance
BUG=webm:1384
Change-Id: I0324f701e723a27cb470036a180693ba8829d01d
James Zern [Fri, 10 Mar 2017 07:36:11 +0000 (23:36 -0800)]
vp9/encoder: fix segfault on win32 using vs < 2015
shift the bsse[] member of the macroblock struct to the front to avoid
an incorrect offset (0) to the upper half of bsse[0] which leads to a
negative resulting in a crash. restrict this to visual studio versions
before 2015 (the bug was observed with 2013, fixed in 2015) to avoid any
potential cache impact on other platforms.
https://connect.microsoft.com/VisualStudio/feedback/details/2396360/bad-structure-offset-in-32-bit-code
BUG=webm:1054
Change-Id: I40f68a1d421ccc503cc712192263bab4f7dde076
Marco Paniconi [Fri, 10 Mar 2017 18:26:06 +0000 (18:26 +0000)]
Merge "vp9: Sample encoder vpx_temporal_svc_encoder: enable row-mt"
Marco [Fri, 10 Mar 2017 16:46:23 +0000 (08:46 -0800)]
vp9: Sample encoder vpx_temporal_svc_encoder: enable row-mt
Enable row-mt in the sample encoder vpx_temporal_svc_encoder.c,
under certain condiitons.
Change-Id: Ic103ee81a9d80be5bf6e5778cc21fc3199db909d
Yi Luo [Fri, 10 Mar 2017 17:14:30 +0000 (17:14 +0000)]
Merge "Improve idct32x32_135_add SSSE3 intrinsics performance"
Marco [Tue, 7 Mar 2017 22:32:30 +0000 (14:32 -0800)]
vp9: Enable row multithreading for SVC in real-time mode.
Enable row-mt for SVC for real-time mode, speed >=5.
Add the controls to the sample encoders, but keep it off for now.
Add the control and enable it for the 1 pass CBR unittests.
For speed 7, 3 layer SVC, 2 threads, row-mt enabled gives about ~5% speedup.
Change-Id: Ie8e77323c17263e3e7a7b9858aec12a3a93ec0c1
Yi Luo [Fri, 3 Mar 2017 00:52:41 +0000 (16:52 -0800)]
Improve idct32x32_135_add SSSE3 intrinsics performance
- Split the inv txfm into three parts to avoid stack spillover.
- Function level speed improves ~12%.
- Use function and macro to remove some repeated code.
Change-Id: I14f5f072334fd766808cb52bf648df792e7379ee