generic-library/vpx
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
17,183 Commits 55 Branches 16 Tags
Commit Graph

17183 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Scott LaVarnway
185836116c Merge "VP9: Eliminate up_available and left_available" 2016-03-30 23:21:07 +00:00
Geza Lore
9d288bf698 Use write_modes_b_wrapper throughout. 
Change-Id: Ifbef3aa6e6b0dbc3701a9ef91b8b685a918d84f4
 2016-03-30 22:26:54 +00:00
Angie Chiang
c7c40d2329 Generalize txfm scale in highbd quantizer 
Change-Id: I359aa49c09b244e0d44ebd09442e365a3d22556c
 2016-03-30 15:25:26 -07:00
Angie Chiang
25520d8dc3 change vp10_fwd_txfm2d_#x#_sse2 to vp10_fwd_txfm2d_#x#_sse4_1 
The speed performance for running 20k times  is as follows

Notice that the vp10_highbd_fdct#x#_sse2 version is
16-bit version plus range check

The rest are 32-bit version

vp10_fwd_txfm2d_4x4_c (2 ms)
vp10_fwd_txfm2d_8x8_c (9 ms)
vp10_fwd_txfm2d_16x16_c (45 ms)
vp10_fwd_txfm2d_32x32_c (233 ms)

vp10_fwd_txfm2d_4x4_sse4_1 (2 ms)
vp10_fwd_txfm2d_8x8_sse4_1 (3 ms)
vp10_fwd_txfm2d_16x16_sse4_1 (16 ms)
vp10_fwd_txfm2d_32x32_sse4_1 (80 ms)

vp10_highbd_fdct4x4_c (1 ms)
vp10_highbd_fdct8x8_c (3 ms)
vp10_highbd_fdct16x16_c (17 ms)
highbd_fdct32x32_c (160 ms)

vp10_highbd_fdct4x4_sse2 (0 ms)
vp10_highbd_fdct8x8_sse2 (2 ms)
vp10_highbd_fdct16x16_sse2 (8 ms)
highbd_fdct32x32_sse2 (105 ms)

Change-Id: I24daf1e0d4d66e91e4ce61ef71cefa7b70ee90ce
 2016-03-30 15:25:26 -07:00
Angie Chiang
c75f64780b Remove redundant code from vp10_fwd_txfm2d.c 
Change-Id: I87ae5e93957616c0f5160a4f679e42f77092c33f
 2016-03-30 15:25:26 -07:00
Angie Chiang
f2b311f580 Simplify rounding in vp10_[fwd/inv]_txfm[1/2]d_#x# 
Change-Id: I24ce46e157dc5b9c0d75000a1a48e9c136ed4ee1
 2016-03-30 15:25:26 -07:00
Angie Chiang
11d2bb5429 Add vp10_fwd_txfm2d_sse2 
Change-Id: Idfbe3c7f5a7eb799c03968171006f21bf3d96091
 2016-03-30 15:25:26 -07:00
Angie Chiang
64413a6ca7 Parameterize transform scale for quantizer 
This is to facilitate changing transform scale later

Change-Id: Ic8ca5afba57d2489ebd191ccc40c1b31605a0d8c
 2016-03-30 15:25:26 -07:00
James Zern
0269df41c1 vpx_fdct32x32_1_c: fix accumulator overflow 
tran_low_t is only 16-bits in non-high-bitdepth mode

Change-Id: Ifc06110c95e86e6d790c44250d52a538b2e9713b
 2016-03-30 15:20:20 -07:00
hui su
cbb8be769d Set block size upper bound for Palette mode 
Avoid buffer overflow in case of such new experiments as
128 x 128 superblock size.

Change-Id: Ib775f3925a85fc87227c0ddd9b6a6110a12ef196
 2016-03-30 14:39:44 -07:00
Debargha Mukherjee
8d3a4aa891 Some fixes/speed-ups on inter-intra part of ext-inter 
Fixes an issue with rectangular inter-intra blocks.
Includes various other refactoring and cleanups to enable fast mixing
of inter and intra predictors.
Uses only the best single inter reference so far for the inter-intra
search.

About 30% speed-up with a 0.1% hit in performance.

This is part one of overhauling on the ext-inter experiment. To be
continued in subsequent patches.

Change-Id: Id10ee100c78c6e00009a3a4f930a4435ef403a95
 2016-03-30 14:39:29 -07:00
Debargha Mukherjee
91707ac79e Merge "Extend superblock size fo 128x128 pixels." into nextgenv2 2016-03-30 20:55:32 +00:00
Geza Lore
552d5cd715 Extend superblock size fo 128x128 pixels. 
If --enable-ext-partition is used at build time, the superblock size
(sometimes also referred to as coding unit (CU) size) is extended to
128x128 pixels.

Change-Id: Ie09cec6b7e8d765b7555ff5d80974aab60803f3a
 2016-03-30 18:23:06 +01:00
Debargha Mukherjee
e467627f33 Merge "Fix for ext_interp experiment" into nextgenv2 2016-03-30 14:44:39 +00:00
Debargha Mukherjee
cd1d01b96a Merge "Add new config flag for global motion experiment" into nextgenv2 2016-03-30 14:36:45 +00:00
Scott LaVarnway
ba962a5f37 VP9: Eliminate up_available and left_available 
Use above_mi and left_mi instead.

Change-Id: I0b50e232c31d11da30aa2fb6f91a695aaf725e0c
 2016-03-30 04:47:39 -07:00
James Zern
8d8ee1f644 Merge "vp8_peek_si_internal: quiet static analysis warning" 2016-03-30 05:38:59 +00:00
Johann Koenig
904cb53302 Merge "Properly propagate out of memory errors." 2016-03-29 22:36:42 +00:00
Jingning Han
b6238b413e Refactor the sub8x8 block motion search control 
Change-Id: Ia340e66e0a61403070adf8e4f18f00eab143f8f7
 2016-03-29 09:53:55 -07:00
Hui Su
aa6f5724ec Merge "Palette mode: record selected transform type" into nextgenv2 2016-03-29 16:23:07 +00:00
Alex Converse
21ce8b9671 Merge "Force the VPX boolcoder trees in the ANS test." into nextgenv2 2016-03-29 16:23:00 +00:00
Yaowu Xu
37241e6f95 Merge "Merge branch 'masterbase' into nextgenv2" into nextgenv2 2016-03-29 16:05:53 +00:00
Julia Robson
068e799459 Fix for ext_interp experiment 
Amends previous commit to also handle subsampling correctly.
Change ID of prev commit: I6b07e6cf9b287ba4b5bd6599af4a7412e50b3bdc

Was causing occassional failures for 422 streams due to accessing
elements beyond the extent of the bmi array.

Change-Id: I37ebabf4c01ca84bcd1851428172bdf753805d98
 2016-03-29 16:09:49 +01:00
Paul Wilkins
aea8d97a74 Merge "Increase min-max q range for normal inter frames." 2016-03-29 14:02:22 +00:00
hui su
4ab00912c4 Palette mode: record selected transform type 
Change-Id: I4c3d3224571176ac924d79ddfaba56990fc4000e
 2016-03-28 20:43:59 -07:00
Jingning Han
78ee83125b Merge "Fix a rdcost computation issue in sub8x8 block mode search" into nextgenv2 2016-03-29 00:51:01 +00:00
Yaowu Xu
c810740c36 Merge branch 'masterbase' into nextgenv2 
Conflicts:
	vp9/encoder/vp9_encoder.c
	vpx_dsp/x86/convolve.h

Change-Id: I60c3532936bedd796a75dfe78245a95ec21e2e55
 2016-03-28 17:44:28 -07:00
Jingning Han
7279a4748f Merge "Rename run_rd_check to run_mv_search" into nextgenv2 2016-03-28 23:10:48 +00:00
Jingning Han
b534987110 Merge "Rework the predicted motion vector for sub8x8 block" into nextgenv2 2016-03-28 23:10:35 +00:00
Jingning Han
d133524e7c Fix a rdcost computation issue in sub8x8 block mode search 
Compute the rate-distortion cost for sub8x8 blocks with integer
motion vectors.

Change-Id: I7dc034fcc4bec3850f26d1f9ae0595c91df1137e
 2016-03-28 23:09:53 +00:00
Jingning Han
59d45d603b Rename run_rd_check to run_mv_search 
Improve the readability in the related rate-distortion optimization
search control function of sub8x8 blocks.

Change-Id: I7f7456bf40a98aa5146abfe0488cda745b84d899
 2016-03-28 21:59:10 +00:00
Jingning Han
0586460938 Rework the predicted motion vector for sub8x8 block 
This commit makes the sub8x8 block to use its nearest neighbor's
motion vector as predicted motion vector for NEWMV mode. It improves
the coding performance by 0.12%.

Change-Id: I99e56715b327573ce7e8a26e3515a4984dadfd98
 2016-03-28 14:58:17 -07:00
Jean-Yves Avenard
6f51672c4e Properly propagate out of memory errors. 
It would otherwise result in an infinite loop.

Change-Id: Ic03fb220cc048538bd62dee599653187f2093079
 2016-03-28 14:14:49 -07:00
Angie Chiang
4144a11552 Merge "Use vp10_[fwd/inv]_txfm2d_add_32x32 for bd 10" into nextgenv2 2016-03-28 19:20:48 +00:00
Alex Converse
7b22c1d433 Force the VPX boolcoder trees in the ANS test. 
Change-Id: I282f958c35aabcdfaf1077f8909c56c999420937
 2016-03-28 12:10:08 -07:00
Yunqing Wang
89a8174fec Merge "Make set_reference control API work in VP9 and VP10" into nextgenv2 2016-03-28 18:33:03 +00:00
Hui Su
14f2d03b4b Merge "Fix assertion fail in build_intra_predictors" into nextgenv2 2016-03-28 18:14:47 +00:00
Angie Chiang
33833aefdd Merge "Use vp10_[fwd/inv]_txfm2d_add_#x# for bd 10" into nextgenv2 2016-03-28 18:11:47 +00:00
Angie Chiang
46b234478f Use vp10_[fwd/inv]_txfm2d_add_32x32 for bd 10 
Change-Id: I996c48a90d7d71b52594a91a35cb8712c7fc212e
 2016-03-28 11:08:40 -07:00
Marco
48e7294e48 vp9-denoiser: Incorporate consec_zeromv in skin detection for denoising. 
Change-Id: Ibe4e1207c0db1779e1358f4566da67c9a07bdb15
 2016-03-28 10:57:54 -07:00
Alex Converse
72e29c3a73 Merge changes I3c72a2d8,I9905f3a8 into nextgenv2 
* changes:
  Add pluggable bitwriters.
  Add pluggable bitreaders.
 2016-03-28 16:59:18 +00:00
Yunqing Wang
9aaa3c933c Make set_reference control API work in VP9 and VP10 
Moved the API patch from NextGen to NextGenv2 and also added this
API to VP10. An example was included. To try it, for example, run
the following command:
$ examples/vpx_cx_set_ref vp10 352 288 in.yuv out.ivf 4 30

Change-Id: Ib56bc3d365e530cfc8d859a13ddbf4c007907b81
 2016-03-28 09:55:24 -07:00
Hui Su
9ab6e589b0 Merge "Fixes for Palette mode" into nextgenv2 2016-03-28 16:44:34 +00:00
hui su
f24b91c9e1 Fix assertion fail in build_intra_predictors 
Change-Id: Id6683b9593b52aa0d159f8f013782d9e0bd07206
 2016-03-28 09:37:54 -07:00
Marco Paniconi
86fb8130bf Merge "vp8-denoiser: Use the same skin detection for model=1 as in vp9." 2016-03-28 16:05:47 +00:00
Yi Luo
80bc1d1d90 Merge "8x8/16x16 HT types V_DCT to H_FLIPADST SSE2 optimization" into nextgenv2 2016-03-28 15:39:24 +00:00
James Zern
ee1bd86d9c Revert "third_party: Roll libwebm snapshot." 
This reverts commit 80ddeb281a0d4b3d9a6d39a7ef91370f52845b39.

breaks Android build and causes valgrind failures in the examples

Change-Id: Ie13ceed2f9f8970af87339b807032a52083f27fe
 2016-03-26 12:56:21 -07:00
Tom Finegan
aa1132d1de Merge "third_party: Roll libwebm snapshot." 2016-03-26 01:41:49 +00:00
hui su
8a128c2a72 Fixes for Palette mode 
This patch fixes 2 issues in Palette mode:
1. More memory is needed in PALETTE_BUFFER for 444 video format.
2. A merge issue caused by
https://chromium-review.googlesource.com/#/c/333940/7

Change-Id: I2aedc7dfdfb6b66fbd600189ec6e1e2cc6120d40
 2016-03-25 18:16:44 -07:00
Alex Converse
9859dde47b Merge "Use speed 2 on superframe test." into nextgenv2 2016-03-26 00:49:06 +00:00