Ahmad Sharif [Thu, 8 Nov 2012 23:14:44 +0000 (15:14 -0800)]
ads2gas.pl: convert push/pop and whole keywords.
This change converts push to stmdb and pop to ldmia. In addition word boundaries
are obeyed using \b avoiding substituion where not appropriate.
Patch provided by ihf@chromium.org.
TEST=Used on many Daisy assembly files.
BUG=None.
Change-Id: Ie5b197b158edd0467294551d0b640c8db6530d95
Scott LaVarnway [Wed, 7 Nov 2012 01:35:02 +0000 (17:35 -0800)]
Incorrect parameter passed into vp8_stuff_mb
Should be MACROBLOCK, not MACROBLOCKD
Change-Id: I0353bbdf085a181ae2ac1f0b96849b38566e9f6a
Scott LaVarnway [Wed, 7 Nov 2012 00:28:54 +0000 (16:28 -0800)]
Merge "Removed mb_norm_activity_map"
Scott LaVarnway [Tue, 6 Nov 2012 17:47:45 +0000 (09:47 -0800)]
Removed mb_norm_activity_map
Not used.
Change-Id: I07ba929e3ff82b4b0c86f731769cdf5434c7a816
Scott LaVarnway [Tue, 6 Nov 2012 01:30:49 +0000 (17:30 -0800)]
Moving _error counts to macroblock struct
Change-Id: I28ac1519d1594801fef9a623cb64598d3d751eb0
Scott LaVarnway [Mon, 5 Nov 2012 20:41:24 +0000 (12:41 -0800)]
Moving MVcount to macroblock struct
Change-Id: Ie22841d096f3c86694b95bd06fc3a8ce1f032a10
Scott LaVarnway [Mon, 5 Nov 2012 20:25:18 +0000 (12:25 -0800)]
Moving ymode_count, uv_mode_count to macroblock struct
Change-Id: Ib73c7b2bee4cb2eb2528fa6b381fffe9503079a0
Scott LaVarnway [Sat, 3 Nov 2012 01:03:03 +0000 (18:03 -0700)]
Moved skip_true_count to macroblock struct
Change-Id: Ie9a26be7c9baa54a0e43a63ed6c77f2746477a9c
Scott LaVarnway [Fri, 2 Nov 2012 23:04:52 +0000 (16:04 -0700)]
Moving coef_counts to macroblock struct
Change-Id: I289564a5a27f0d03ddc6f19c7838542ff22719be
Scott LaVarnway [Mon, 5 Nov 2012 17:51:40 +0000 (09:51 -0800)]
Merge "Removed unnecessary VP8_COMP *cpi parameters"
Scott LaVarnway [Fri, 2 Nov 2012 19:10:07 +0000 (12:10 -0700)]
Removed unnecessary VP8_COMP *cpi parameters
Code cleanup
Change-Id: I82f9d787a2f511d39895fd8dfd5347a1676d9dbc
Johann [Fri, 2 Nov 2012 16:26:18 +0000 (09:26 -0700)]
Merge "ads2gas.pl: various enhancements to work with flash."
John Koleszar [Fri, 2 Nov 2012 16:16:41 +0000 (09:16 -0700)]
Merge "vpx_scale: sync from experimental"
John Koleszar [Fri, 2 Nov 2012 02:14:42 +0000 (19:14 -0700)]
vpx_scale: sync from experimental
Import changes made on the experimental branch in preparation for
merging the two branches.
Change-Id: I7b5b8fb4fca155cb1d72e7ba13eef18e6a94a298
Ahmad Sharif [Thu, 1 Nov 2012 20:01:14 +0000 (13:01 -0700)]
ads2gas.pl: various enhancements to work with flash.
TEST=Ran it on different asm files.
Change-Id: Ief2a009366787954d0eb5c356c64acaef350cf84
Scott LaVarnway [Wed, 31 Oct 2012 20:35:13 +0000 (13:35 -0700)]
vp8dx_receive_compressed_data() cleanup
Change-Id: I6edf0626d00ae177c7c04eec64e1ec0dd861dfbe
Jim Bankoski [Tue, 30 Oct 2012 02:54:06 +0000 (19:54 -0700)]
add unit test for datarate control
Adds some basic datarate control tests..
Change-Id: I6eff18aa58ef7cfd14886c6b1e048247cd1ad247
Jim Bankoski [Sat, 27 Oct 2012 02:49:44 +0000 (19:49 -0700)]
Minor tweaks to video source frameworks
Fix the video source to close if file is open already and add a limit.
Change-Id: I36ada4c609d027b6eaa9b447fe9ad4115532edc1
James Zern [Fri, 26 Oct 2012 01:31:35 +0000 (18:31 -0700)]
vpxenc: add -quiet option
suppresses encoder progress
Change-Id: Iacc04d98de970f13697c002363ee3a40a03f021b
Johann [Fri, 26 Oct 2012 00:04:03 +0000 (17:04 -0700)]
Use vp8_clear_system_state helper function
Change-Id: I4dd192c3297e31b8eadd108ca8d0934e58f502f7
Johann [Thu, 25 Oct 2012 22:51:33 +0000 (15:51 -0700)]
variance.h uses defines in vpx_config.h
Change-Id: Ibd3e4425a3bf8aae8aa33b70da2b84e1cda8c07f
John Koleszar [Tue, 23 Oct 2012 17:59:05 +0000 (10:59 -0700)]
postproc_sse2: avoid reading off the end of the limits array
Rather than unconditionally reading in the next MB's limits, test
the loop exit condition first.
Change-Id: I105d1e92698fb5561aa87160816787604aed03a2
John Koleszar [Mon, 22 Oct 2012 22:09:33 +0000 (15:09 -0700)]
Merge "postproc: allocate enough memory for limits buffer"
John Koleszar [Mon, 22 Oct 2012 21:21:59 +0000 (14:21 -0700)]
postproc: allocate enough memory for limits buffer
The vp8_post_proc_down_and_across_mb_row_sse2() needs space for an
even number of macroblocks, as they are read two at a time for the
chroma planes. Round up the width during the allocation of
pp_limits_buffer to support this.
Change-Id: Ibfc10c7be290d961ab23ac3dde12a7bb96c12af0
Yunqing Wang [Mon, 22 Oct 2012 21:26:53 +0000 (14:26 -0700)]
Merge "Add unit test for decoder test_vector_test"
Yunqing Wang [Thu, 4 Oct 2012 19:59:36 +0000 (12:59 -0700)]
Add unit test for decoder test_vector_test
Got 61 test vectors from vp8-test-vectors.git
(http://git.chromium.org/gitweb/?p=webm/vp8-test-vectors.git)
Added decoder test vectors downloading in unit tests. Uploaded
the test vectors and their md5 files to WebM website.
$ gsutil cp *.* gs://downloads.webmproject.org/test_data/libvpx
Added their sha1sum to the test/test-data.sha1 file.
In unit tests, download the test vectors to LIBVPX_TEST_DATA_PATH.
Test_vector_test goes through the test vectors, decodes them, and
compute the md5 checksums. The checksums are compared with the
expected md5 checksums to tell if the decoder decodes correctly.
Change-Id: Ia1e84f9347ddf1d4a02e056c0fee7d28dccfae15
Marco Paniconi [Tue, 16 Oct 2012 01:41:32 +0000 (18:41 -0700)]
Fix to rd cost computation for mv bias.
This prevents the rd cost from wrapping around and going negative.
Change-Id: Id61390537d2ad47762e410918bb4e18f6a0912d4
John Koleszar [Thu, 11 Oct 2012 16:55:06 +0000 (09:55 -0700)]
Merge "Add option to disable documentation"
Yunqing Wang [Thu, 11 Oct 2012 15:46:23 +0000 (08:46 -0700)]
Clean up error return code in alloccommon.c
Clean up the duplicate code as Pascal suggested.
Change-Id: I685fcbb488502e969f6cb73a46db3ea79b90910d
Yunqing Wang [Wed, 10 Oct 2012 18:27:11 +0000 (11:27 -0700)]
post-proc: fix 0 or negative threshold handling
If the threshold(limits) <= 0, skipped filtering and copied the
frame directly. Also, fixed memory allocation checking.
Change-Id: If3d79d5b2bcb71b9777e6eb5cba1384585131e22
Johann [Wed, 10 Oct 2012 16:16:37 +0000 (09:16 -0700)]
Add option to disable documentation
Documentation is typically auto-detected by checking for php and
doxygen. Add an option to explicitly disable it.
Remove toggle keywords from libraries, examples, documentation and
unit tests. They were not consistent with the default status.
Change-Id: I21049675ccfd8e58ac612cd058641b197db5c0eb
John Koleszar [Mon, 8 Oct 2012 22:58:04 +0000 (15:58 -0700)]
multi-res: disable intra on forced ref frames
If a reference frame is forced because of low dissimilarity, then
shut off the search of intra modes. This change has mixed results. On
one clip (QVGA), it hurt quality by ~1.5% with negligible speed impact.
On another (VGA) it had negligible affect on quality, but a ~0.2% speed
impact.
Change-Id: Ic8b07648979d732f489de5f094957e140f84d2eb
John Koleszar [Tue, 9 Oct 2012 17:24:37 +0000 (10:24 -0700)]
Merge "multi-res: add parent_ref_valid flag"
John Koleszar [Mon, 8 Oct 2012 22:54:24 +0000 (15:54 -0700)]
multi-res: add parent_ref_valid flag
Rather than overloading the parent_ref_frame value to shut off the
search in some cases, add a new validity flag. This cleans up some
of the duplicated mr_encoder_id && mr_low_res_mv_avail checks as
well, for readability.
Change-Id: Iddad93a27066c3d85ff2f25a361ac113b288ab7b
Yunqing Wang [Tue, 9 Oct 2012 16:26:36 +0000 (09:26 -0700)]
Merge "post-proc: deblock filter optimization"
John Koleszar [Mon, 8 Oct 2012 22:48:25 +0000 (15:48 -0700)]
Merge "multi-res: work around reference mismatch"
Yunqing Wang [Fri, 28 Sep 2012 17:13:07 +0000 (10:13 -0700)]
post-proc: deblock filter optimization
1. Algorithm modification:
Instead of having same filter threshold for a whole frame, now we
allow the thresholds to be adjusted for each macroblock. In current
implementation, to avoid excessive blur on background as reported
in issue480(http://code.google.com/p/webm/issues/detail?id=480), we
reduce the thresholds for skipped macroblocks.
2. SSE2 optimization:
As started in issue479(http://code.google.com/p/webm/issues/detail?id=479),
the filter calculation was adjusted for better performance. The c
code was also modified accordingly. This made the deblock filter
2x faster, and the decoder was 1.2x faster overall.
Next, the demacroblock filter will be modified similarly.
Change-Id: I05e54c3f580ccd427487d085096b3174f2ab7e86
John Koleszar [Mon, 8 Oct 2012 15:39:47 +0000 (08:39 -0700)]
multi-res: work around reference mismatch
In some situations, believed to be an interaction between temporal
scalability and dropped frames, the references available to an
encoder may not be the same references available to its parent.
Previously, the code tried to force the reference frame chosen by
the parent to be used on this frame, even if it was disabled. This
was preventing the pick mode loop from running even once, which led
to a crash.
Attempts to reproduce this bug locally were unsuccessful, so it is
still undetermined what the underlying cause of this issue is. In
the specific case that was failing, the application did not set
any flags which influenced the reference selection on that frame.
ref_frame_flags indicated that the golden frame was disabled,
believed to be because the last frame updated the last and golden
frames, so golden was shut off by default. It's not clear why this
wouldn't have also been true in the lower res encoder, ie, why the
lower res encoder decided to use and/or was allowed to use the
golden frame. We weren't able to debug into the non-crashing
lower res encoder as the crash couldn't be reproduced locally.
Change-Id: Ifb265253d26963ac2afde0e20cf6792788be6af7
Adrian Grange [Fri, 5 Oct 2012 00:33:55 +0000 (17:33 -0700)]
Merge "Unit Test for Error Resilience Mode"
Ronald S. Bultje [Thu, 4 Oct 2012 23:38:21 +0000 (16:38 -0700)]
Add a unit test for CQ mode.
Change-Id: I66c391987eabc5ea0159bf4a2a4fd8e8e163872f
Adrian Grange [Thu, 4 Oct 2012 17:38:35 +0000 (10:38 -0700)]
Merge "Added handler for PSNR packets to EncoderTest class"
Adrian Grange [Thu, 4 Oct 2012 17:37:38 +0000 (10:37 -0700)]
Merge "Add initialization and per frame flag members"
Adrian Grange [Tue, 2 Oct 2012 19:16:27 +0000 (12:16 -0700)]
Unit Test for Error Resilience Mode
This unit test compares the difference in quality with
error resilience enabled and disabled. The test runs
for all of the one-pass encoding modes.
The test ensures that the effect of turning on error
resilience makes less than a 10% difference in PSNR.
Further cases should be added to do a more comprehensive
test.
Change-Id: I1fc747fc78c9459bc6c74494f4b38308dbed0c32
John Koleszar [Wed, 3 Oct 2012 21:52:56 +0000 (14:52 -0700)]
fix uninitialized value in multi-res encoding
If a parent mb is available but is intra coded, then parent_ref_mv is
invalid. Check that the parent is inter coded before trying to access
the parent_ref_mv. Previously the parent_ref_mv was being read from
an uninitialized stack allocation, causing potential OOB reads and
other undefined behavior.
Change-Id: I0c93cd412a19c3a184bcf6decaa145b3a036a6c0
Adrian Grange [Tue, 2 Oct 2012 18:27:29 +0000 (11:27 -0700)]
Added handler for PSNR packets to EncoderTest class
Added a virtual function to handle PSNR packets.
Change-Id: Id2a6372c691a14f19bbeed217a93a9df03e81e75
Adrian Grange [Tue, 2 Oct 2012 18:03:09 +0000 (11:03 -0700)]
Add initialization and per frame flag members
Modified EncoderTest class to have separate member variables
for initialization time and per-frame.
Change-Id: I08a1901f8f3ec16e45f96297e08e7f6df0f4aa0b
Adrian Grange [Wed, 3 Oct 2012 20:45:46 +0000 (13:45 -0700)]
Merge "Added Reset method to TwopassStatsStore"
John Koleszar [Wed, 3 Oct 2012 19:51:29 +0000 (12:51 -0700)]
Merge "rtcd/win32: use InitializeCriticalSection explicitly"
Adrian Grange [Tue, 2 Oct 2012 16:36:41 +0000 (09:36 -0700)]
Added Reset method to TwopassStatsStore
The stats buffer needs to be reset between runs of the
encoder. I added a Reset() function to TwopassStatsStore
and called it at the beginning of each encode.
This enables us to run multiple encodes which was
previously not possible since there was no way to reset
the stats between runs.
Change-Id: Iebb18dab83ba9331f009f764cc858609738a27f9
John Koleszar [Mon, 24 Sep 2012 21:52:18 +0000 (14:52 -0700)]
rtcd/win32: use InitializeCriticalSection explicitly
Protect the call to {Initialize,Delete}CriticalSection() with an
Interlocked{Inc,Dec}rement() pair, rather than the previous static
initialization. This should play better with AppVerifier, and fix issue
http://code.google.com/p/webm/issues/detail?id=467
Change-Id: I06eadbfac1b3b4414adb8eac862ef9bd14bbe4ad
John Koleszar [Wed, 3 Oct 2012 15:17:16 +0000 (08:17 -0700)]
Merge "Added unit test for subtract functions"
Jim Bankoski [Mon, 1 Oct 2012 21:17:43 +0000 (14:17 -0700)]
Disable keyframe in real time that's placed one frame after a cut.
The codec as it stood placed a keyframe one frame after a
real cut scene - and ignored datarate and other considerations.
TODO: Its possible that we should detect a keyframe and recode
the frame ( in certain circumstances) to improve quality.
Change-Id: Ia1fd6d90103f4da4d21ca5ab62897d22e0b888a8
Marco Paniconi [Thu, 27 Sep 2012 21:30:40 +0000 (14:30 -0700)]
Resetting of the cyclic_refresh_mode_index:
Reset the cyclie refresh mode index in alloc_compressor_data().
This is needed to handle both cases of internal and
external spatial resizing.
Change-Id: I2697e12d45135eae2e8f0d45161811f24722312a
Marco Paniconi [Wed, 26 Sep 2012 22:42:36 +0000 (15:42 -0700)]
Reset the cyclic_refresh_mode_index to 0 on resize.
On an internal spatial resize, this mode index was not reset to 0,
and therefore could exceed dimensions of seg_map or cyclic_refresh_map.
Change-Id: I6fe85dbd2765eb0207a9d9f71fda8d8b8c34f075
Scott LaVarnway [Wed, 26 Sep 2012 20:45:53 +0000 (13:45 -0700)]
Added unit test for subtract functions
Patch Set 1: gain familiarity with unit tests... added simple
4x4 subtract test
Patch Set 2: fixed mistakes, parameterized as suggested
Patch Set 3: randomized the source/predictor data
Change-Id: I33432bdf7c9f2a9b8c2533a37106382c2a8209ee
Signed-off-by: Scott LaVarnway <slavarnway@google.com>
John Koleszar [Tue, 25 Sep 2012 18:32:11 +0000 (11:32 -0700)]
Merge "check for x32 targets"
John Koleszar [Tue, 25 Sep 2012 18:25:53 +0000 (11:25 -0700)]
Merge "update gitignores"
Mike Frysinger [Tue, 25 Sep 2012 18:11:13 +0000 (14:11 -0400)]
update gitignores
Change-Id: I67bf154e35fa02b407a9b2c312a536d6bb2aaaf2
Jim Bankoski [Tue, 25 Sep 2012 17:49:25 +0000 (10:49 -0700)]
Merge "disable segmentation on enhancement layers"
Mike Frysinger [Tue, 25 Sep 2012 04:37:11 +0000 (00:37 -0400)]
check for x32 targets
Add configure detection of the new x32 ABI as well as support in asm.
Change-Id: Ic66a069599adeb81062090e3f11b71ee1fb97cb8
Scott LaVarnway [Mon, 24 Sep 2012 23:15:54 +0000 (16:15 -0700)]
Merge "Move frame allocations out of vp8_decode_frame()"
Scott LaVarnway [Mon, 24 Sep 2012 20:50:30 +0000 (13:50 -0700)]
Merge "Removed bc and bc2 vp8_readers from VP8D_COMP"
Scott LaVarnway [Mon, 24 Sep 2012 19:44:45 +0000 (12:44 -0700)]
Move frame allocations out of vp8_decode_frame()
in order to make it easier to implement the frame-base multithreading.
Change-Id: Iea2fd05be34fa704848fdc7669bf167f2ed229c5
John Koleszar [Mon, 24 Sep 2012 18:47:35 +0000 (11:47 -0700)]
rtcd/win32: prefer win32 primatives to pthreads
Fixes some build issues for people building for win32 who have a
pthreads emulation layer installed.
Change-Id: I0e0003fa01f65020f6ced35d961dcb1130db37a8
Jim Bankoski [Mon, 24 Sep 2012 14:53:26 +0000 (07:53 -0700)]
disable segmentation on enhancement layers
This should avoid problems with blocks gettings high quality
improvement despite having recently moved:
Change-Id: Ic0af0de2d6577807fa3c553f47b55d547ef36359
Marco Paniconi [Mon, 24 Sep 2012 03:56:49 +0000 (20:56 -0700)]
Update to cyclic refresh:
Set the seg map to 0 for key frame.
In previous commit on cyclic refresh, the seg map for key frame
was not reset, and instead used the seg map from last frame.
Change-Id: I848eb2face420dfcd2f7daca6f070b9127ca938b
Marco Paniconi [Sat, 22 Sep 2012 01:54:11 +0000 (18:54 -0700)]
Update to cyclic refresh.
-Increase the amount of mbs to be refreshed.
-Replace the delta qp with a fixed and reduced delta.
-Change to the mb update loop to try to always update same amount of mbs.
Change-Id: I93ac88002fd8dc677d2337f77998ff93f64e4ff9
Scott LaVarnway [Fri, 21 Sep 2012 18:00:29 +0000 (11:00 -0700)]
Removed bc and bc2 vp8_readers from VP8D_COMP
Used mbc instead.
Change-Id: If22136b4fee89b68ca2246265f3c27853701136c
Scott LaVarnway [Wed, 19 Sep 2012 19:30:44 +0000 (12:30 -0700)]
Moved vp8dx_get_raw_frame() call to vp8_get_frame()
This change is necessary for the frame-based multithreading
implementation.
Since the postproc occurs in this call, vpxdec was modified to time around
vpx_codec_get_frame()
Change-Id: I389acf78b6003cd35e41becc16c893f7d3028523
Scott LaVarnway [Mon, 17 Sep 2012 23:07:08 +0000 (16:07 -0700)]
Merge "Changed setup intra recon to be row based"
Adrian Grange [Mon, 17 Sep 2012 20:40:07 +0000 (13:40 -0700)]
Merge "Update to 2 layer case in vp8_scalable_patterns."
Scott LaVarnway [Mon, 17 Sep 2012 20:08:38 +0000 (13:08 -0700)]
Merge "Moved mb_no_coeff_skip flag read"
Scott LaVarnway [Mon, 17 Sep 2012 20:08:18 +0000 (13:08 -0700)]
Merge "Removed pre_mvc"
Marco Paniconi [Mon, 17 Sep 2012 18:57:51 +0000 (11:57 -0700)]
Update to 2 layer case in vp8_scalable_patterns.
Change-Id: I11aa90be41dd585c9784301f47a9d580fac09260
Marco Paniconi [Fri, 14 Sep 2012 19:07:02 +0000 (12:07 -0700)]
Update to vp8_scalable_patters.
Change-Id: I43ea65bf0a35ddb33a096a392e10b1869ea73f9a
Adrian Grange [Fri, 14 Sep 2012 16:49:06 +0000 (09:49 -0700)]
Merge "Updates to vp8_scalable_patterns.c."
Scott LaVarnway [Thu, 13 Sep 2012 20:37:24 +0000 (13:37 -0700)]
Moved mb_no_coeff_skip flag read
from vp8_decode_frame() to mb_mode_mv_init()
Change-Id: I267fd764a4c826d62e6799dceb1b8fb5f97fde6d
Scott LaVarnway [Thu, 13 Sep 2012 18:55:57 +0000 (11:55 -0700)]
Removed pre_mvc
Not used.
Change-Id: Ifbd97c9ad537d5184f7e305fa8acf6f027978b6b
Scott LaVarnway [Thu, 13 Sep 2012 18:51:27 +0000 (11:51 -0700)]
Changed setup intra recon to be row based
Initialize the top line at the beginning of each frame and
the left column at the beginning of each row.
Change-Id: I5412f7ea49ffc490215cf65a62715a6c5e3a5a29
Jim Bankoski [Wed, 12 Sep 2012 20:50:29 +0000 (13:50 -0700)]
Merge "valgrind found motion vectors which exceeded frame boundaries"
Jim Bankoski [Tue, 11 Sep 2012 23:30:19 +0000 (16:30 -0700)]
fix valgrind mem leak on garbage decode
Multiple decoders were getting allocated per frame.
If the decoder crashed we exitted with out freeing
them and the next time in we'd allocate over.
This fix removes the allocation and just has 8
boolcoders in the pbi structure
Change-Id: I638b5bda23b622b43b7992aec21dd7cf6f6278da
Marco Paniconi [Tue, 11 Sep 2012 20:15:51 +0000 (13:15 -0700)]
Updates to vp8_scalable_patterns.c.
Updated/added a few cases of temporal patterns,
and some changes to the encoder settings.
Change-Id: Idaa75cfc4a26e5b6321f273ef9450b5c85368a4d
Scott LaVarnway [Tue, 11 Sep 2012 16:14:02 +0000 (09:14 -0700)]
valgrind found motion vectors which exceeded frame boundaries
Bug introduced in I02d034c70cd97b65025d59dd67c695e1db529f0b
Change-Id: Ifb74ab659f3b3cf5f5b95046bacffb202d5b4d64
Jim Bankoski [Fri, 7 Sep 2012 23:38:15 +0000 (16:38 -0700)]
valgrind caught uninitialized cond
If the decoder crashes and returned an error before it set up
block offsets but after it set up frame buffers. We had a
problem decoding the next keyframe because the block offsets
were never set.
Change-Id: Ied2866e9770d80fc66241d5e0d978d4f5f9cdd89
Yaowu Xu [Thu, 6 Sep 2012 16:28:17 +0000 (09:28 -0700)]
Merge "added encode/decode matching validation to tests"
Deb Mukherjee [Tue, 4 Sep 2012 19:09:05 +0000 (12:09 -0700)]
Adjusting thresholds in mfqe post-processing
Adjusts some of the qualification thresholds in mfqe to eliminate
artifacts due to wrong decisions. Besides, a new qualification
criteria is used to disable mfqe if the quality of the previous
frame is itself not too good.
Change-Id: I4097c20b7fd4fcc60cc3003c1e33e8faae2ff066
Yunqing Wang [Fri, 31 Aug 2012 20:51:30 +0000 (13:51 -0700)]
Merge "Encoder denoiser performance improvement"
Yunqing Wang [Fri, 10 Aug 2012 19:35:55 +0000 (12:35 -0700)]
Encoder denoiser performance improvement
The denoiser function was modified to reduce the computational
complexity.
1. The denoiser c function modification:
The original implementation calculated pixel's filter_coefficient
based on the pixel value difference between current raw frame and last
denoised raw frame, and stored them in lookup tables. For each pixel c,
find its coefficient using
filter_coefficient[c] = LUT[abs_diff[c]];
and then apply filtering operation for the pixel.
The denoising filter costed about 12% of encoding time when it was
turned on, and half of the time was spent on finding coefficients in
lookup tables. In order to simplify the process, a short cut was taken.
The pixel adjustments vs. pixel diff value were calculated ahead of time.
adjustment = filtered_value - current_raw
= (filter_coefficient * diff + 128) >> 8
The adjustment vs. diff curve becomes flat very quick when diff increases.
This allowed us to use only several levels to get a close approximation
of the curve. Following the denoiser algorithm, the adjustments are
further modified according to how big the motion magnitude is.
2. The sse2 function was rewritten.
This change made denoiser filter function 3x faster, and improved the
encoder performance by 7% ~ 10% with the denoiser on.
Change-Id: I93a4308963b8e80c7307f96ffa8b8c667425bf50
Yaowu Xu [Thu, 30 Aug 2012 20:43:15 +0000 (13:43 -0700)]
added encode/decode matching validation to tests
This commit adds the ability of validating matched encoder and
decoder to unit tests.
Change-Id: Ie00d69a42477b6a69b324a6bd134939684f7300b
John Koleszar [Fri, 31 Aug 2012 17:50:01 +0000 (10:50 -0700)]
Merge "msvs/tests: fix data alignment for asm tests"
James Zern [Tue, 28 Aug 2012 00:13:09 +0000 (17:13 -0700)]
msvs/tests: fix data alignment for asm tests
Replace DECLARE_ALIGNED_ with vpx_memalign()
DECLARE_ALIGNED (__declspec(align())) does not work as intended when
used on class data members:
Data in classes or structures is aligned within the class or structure
at the minimum of its natural alignment and the current packing setting
(from #pragma pack or the /Zp compiler option)
Change-Id: I304aaa6c3716fbfae24675ecf192f4b40787e83e
Yaowu Xu [Tue, 28 Aug 2012 14:03:19 +0000 (07:03 -0700)]
Merge "silent one more compiler warning"
Yaowu Xu [Tue, 28 Aug 2012 13:33:24 +0000 (06:33 -0700)]
silent one more compiler warning
Change-Id: I621a2bbd5f07691fa76dac36b65f406ad4fed792
Scott LaVarnway [Thu, 23 Aug 2012 20:55:06 +0000 (13:55 -0700)]
Merge "Added error checking to vp8cx_create_encoder_threads()"
Scott LaVarnway [Thu, 23 Aug 2012 00:23:28 +0000 (17:23 -0700)]
Merge "Added row based extend borders"
Jim Bankoski [Tue, 21 Aug 2012 22:56:59 +0000 (15:56 -0700)]
Merge "Add biasing to ZEROMV for videos with static background"
Scott LaVarnway [Tue, 21 Aug 2012 22:20:14 +0000 (15:20 -0700)]
Merge "Removed unused dr from VP8D_COMP"
Scott LaVarnway [Tue, 21 Aug 2012 22:19:58 +0000 (15:19 -0700)]
Merge "Removed last_kf_gf_q"
Yunqing Wang [Tue, 21 Aug 2012 17:52:35 +0000 (10:52 -0700)]
Add biasing to ZEROMV for videos with static background
For videos with big static background(such as video conferencing
clips), the mode decision was biased to ZEROMV in order to
obtain a stable background. The percentage of ZEROMV on last
frame was used to predict if there is static area in current frame,
and checking already-encoded neighboring macroblocks' motion
vectors to make sure the local area has low motion.
Change-Id: I05b3241d3a56a0bda88b6681e5646c1c8baf2e57
James Zern [Tue, 21 Aug 2012 20:33:46 +0000 (13:33 -0700)]
Merge "sad_test: fix msvc build"