Fritz Koenig [Mon, 22 Aug 2011 19:36:28 +0000 (12:36 -0700)]
Reclassify optimized ssim calculations as SSE2.
Calculations were incorrectly classified as either
SSE3 or SSSE3. Only using SSE2 instructions.
Cleanup function names and make non-RTCD code work
as well.
Change-Id: I48ad0218af0cc51c5078070a08511dee43ecfe09
Fritz Koenig [Mon, 22 Aug 2011 19:32:12 +0000 (12:32 -0700)]
Merge "Revert "Reclasify optimized ssim calculations as SSE2.""
Fritz Koenig [Mon, 22 Aug 2011 18:31:12 +0000 (11:31 -0700)]
Revert "Reclasify optimized ssim calculations as SSE2."
This reverts commit
01376858cd184d820ff4c2d8390361a8679c0e87
Fritz Koenig [Mon, 22 Aug 2011 16:20:33 +0000 (09:20 -0700)]
Merge "Reclasify optimized ssim calculations as SSE2."
John Koleszar [Fri, 19 Aug 2011 16:04:04 +0000 (12:04 -0400)]
tools: author_first_release.sh
First version of a simple tool to get a list of the version a user first
contributed to.
Change-Id: I8f1b1fef5343de269c4b6209632c9cedc2cf1a37
Fritz Koenig [Fri, 19 Aug 2011 15:51:27 +0000 (08:51 -0700)]
Reclasify optimized ssim calculations as SSE2.
Calculations were incorrectly classified as either
SSE3 or SSSE3. Only using SSE2 instructions.
Cleanup function names and make non-RTCD code work
as well.
Change-Id: I29f5c2ead342b2086a468029c15e2c1d948b5d97
John Koleszar [Fri, 19 Aug 2011 14:31:00 +0000 (07:31 -0700)]
Merge "Copy less when active map is in use"
Alpha Lam [Tue, 9 Aug 2011 19:59:45 +0000 (20:59 +0100)]
Copy less when active map is in use
When active map is specified and the current frame is not a key frame,
golden frame nor a altref frame then copy only those active regions.
This significantly reduces encoding time by as much as 19% on the test
system where realtime encoding is used. This is particularly useful
when the frame size is large (e.g. 2560x1600) and there's only a few
action macroblocks.
Change-Id: If394a813ec2df5a0201745d1348dbde4278f7ad4
Paul Wilkins [Wed, 17 Aug 2011 13:14:23 +0000 (14:14 +0100)]
Small boost to every other frame.
Instead of a single mid GF boost apply a few extra bits to
every other frame. This gives a very small average metrics
improvement on both derf and YT sets.
Also use min GF interval as min KF interval.
Change-Id: Iee238b8cae0ffaed850a5a944ac825cee18da485
John Koleszar [Mon, 15 Aug 2011 21:08:40 +0000 (17:08 -0400)]
Merge v0.9.7-p1 release int 'origin/master'
Change-Id: I93388d2f8846615ad1e26b975308c5e96b9b1918
John Koleszar [Mon, 15 Aug 2011 21:02:45 +0000 (17:02 -0400)]
Update CHANGELOG for v0.9.7-p1
Change-Id: I5490a9cad2d6752832b6bf4ec1835c06a45eeb9b
Stefan Holmer [Mon, 15 Aug 2011 07:28:41 +0000 (09:28 +0200)]
Don't set the bmi mode when doing error concealment
Since the block will be interpreted as an inter block, the mode will
be interpreted as a motion vector, resulting in bad concealment.
Change-Id: Ifcc685ae1cc883492bce6dbd61e418d91a89b053
Stefan Holmer [Mon, 15 Aug 2011 07:28:41 +0000 (09:28 +0200)]
Don't set the bmi mode when doing error concealment
Since the block will be interpreted as an inter block, the mode will
be interpreted as a motion vector, resulting in bad concealment.
Change-Id: Ifcc685ae1cc883492bce6dbd61e418d91a89b053
John Koleszar [Fri, 12 Aug 2011 18:59:10 +0000 (14:59 -0400)]
Generate libvpx_srcs.txt from current configuration
To get a list of files that the libvpx library depends on in the current
configuration, run:
$ make target=libs libvpx_srcs.txt
Change-Id: I68a69648ecf212f0fe29c325297728ac2a9393d9
John Koleszar [Fri, 12 Aug 2011 18:51:36 +0000 (14:51 -0400)]
Revert "Improved 1-pass CBR rate control"
This reverts commit
b5ea2fbc2c1554769848774c836aad262af95072. Further
testing showed noticable keyframe popping in some cases, reverting this
for now to give time for a proper fix.
Conflicts:
vp8/encoder/onyx_if.c
vp8/encoder/ratectrl.c
Change-Id: I159f53d1bf0e24c035754ab3ded8ccfd58fd04af
John Koleszar [Fri, 12 Aug 2011 15:30:54 +0000 (11:30 -0400)]
Propagate macroblock MV to subblocks for error concealment
EC expects the subblock MVs to be populated, but
f1d6cc79e43f0066632f19c1854ca365086b712b removed this code. This
commit restores it, protected by CONFIG_ERROR_CONCEALMENT. May move this
to the EC code more directly in the future.
Change-Id: I44f8f985720cb9a1bf222e59143f9e69abf56ad2
Stefan Holmer [Mon, 8 Aug 2011 08:56:20 +0000 (10:56 +0200)]
Disable error concealment until first key frame is decoded
When error concealment is enabled the first key frame must
be successfully received before error concealment is activated.
Error concealment will be activated when the delta following
delta frame is received.
Also fixed a couple of bugs related to error tracking in
multi-threading. And avoiding decoding corrupt residual
when we have multiple non-resilient partitions.
Change-Id: I45c4bb296e2f05f57624aef500a874faf431a60d
John Koleszar [Fri, 5 Aug 2011 16:27:25 +0000 (12:27 -0400)]
Fix potential OOB read with Error Concealment
This patch fixes an OOB read when error concealment is enabled and the
partition sizes are corrupt. The partition size read from the bitstream
was not being validated in EC mode.
Change-Id: Ia81dfd4bce1ab29ee78e42320abe52cee8318974
John Koleszar [Fri, 12 Aug 2011 18:45:26 +0000 (11:45 -0700)]
Merge "Disable error concealment until first key frame is decoded"
John Koleszar [Fri, 12 Aug 2011 15:30:54 +0000 (11:30 -0400)]
Propagate macroblock MV to subblocks for error concealment
EC expects the subblock MVs to be populated, but
f1d6cc79e43f0066632f19c1854ca365086b712b removed this code. This
commit restores it, protected by CONFIG_ERROR_CONCEALMENT. May move this
to the EC code more directly in the future.
Change-Id: I44f8f985720cb9a1bf222e59143f9e69abf56ad2
Stefan Holmer [Mon, 8 Aug 2011 08:56:20 +0000 (10:56 +0200)]
Disable error concealment until first key frame is decoded
When error concealment is enabled the first key frame must
be successfully received before error concealment is activated.
Error concealment will be activated when the delta following
delta frame is received.
Also fixed a couple of bugs related to error tracking in
multi-threading. And avoiding decoding corrupt residual
when we have multiple non-resilient partitions.
Change-Id: I45c4bb296e2f05f57624aef500a874faf431a60d
John Koleszar [Fri, 5 Aug 2011 16:27:25 +0000 (12:27 -0400)]
Fix potential OOB read with Error Concealment
This patch fixes an OOB read when error concealment is enabled and the
partition sizes are corrupt. The partition size read from the bitstream
was not being validated in EC mode.
Change-Id: Ia81dfd4bce1ab29ee78e42320abe52cee8318974
Yunqing Wang [Fri, 5 Aug 2011 19:15:32 +0000 (12:15 -0700)]
Merge "Adjust half-pixel only search"
John Koleszar [Thu, 4 Aug 2011 21:05:55 +0000 (17:05 -0400)]
(Re)Merge branch 'master' into cayuga
Corrected the merge direction this time, so that running
`git describe` on the master branch finds v0.9.7 as the most recent
tag.
Change-Id: I9e7b5d473c26e670c6d9a76f5c03fa617690651d
John Koleszar [Thu, 4 Aug 2011 17:30:15 +0000 (10:30 -0700)]
Merge cayuga release into master branch
Change-Id: I54ab3c22d281efe69af9c9108fcbfadd3cc747ee
John Koleszar [Thu, 4 Aug 2011 17:30:15 +0000 (10:30 -0700)]
Merge "Update CHANGELOG for Cayuga release" into cayuga
John Koleszar [Wed, 3 Aug 2011 20:45:24 +0000 (13:45 -0700)]
Merge changes Ic7725e27,Ib3d54bfa into cayuga
* changes:
Update AUTHORS
Update .mailmap entry for Ralph Giles
John Koleszar [Wed, 3 Aug 2011 20:44:32 +0000 (13:44 -0700)]
Merge changes I585167e1,Ia07602bd into cayuga
* changes:
Fix building of static libs on universal-darwin
Fix asm offsets generation for universal-darwin builds
John Koleszar [Wed, 3 Aug 2011 20:12:12 +0000 (16:12 -0400)]
Fix source buffer selection
This patch fixes a bug in the interaction between the recode loop and
spatial resampling. If the codec was in a spatial resampling state,
and a subsequent iteration of the recode loop disables resampling,
then the source buffer must be reset to the unscaled source.
Change-Id: I4e4cd47b943f6cd26a47449dc7f4255b38e27c77
Yunqing Wang [Wed, 3 Aug 2011 15:51:07 +0000 (11:51 -0400)]
Adjust half-pixel only search
Changed motion search in vp8_find_best_half_pixel_step() to be the
same as in vp8_find_best_sub_pixel_step(), which checks 5 points
instead of 8 points. This only affects real-time mode with
cpu-used >=9. Tests showed it gives 2% encoding speedup with
a quality loss(psnr) of up to 0.5%.
Change-Id: I16049cad1535002346d46cfdfad345bfc3dc5146
John Koleszar [Wed, 3 Aug 2011 13:20:37 +0000 (09:20 -0400)]
Fix building of static libs on universal-darwin
The static libs should not be built from sources during the top level
of a universal build. This regression was introduced in commit
495b241fa6b03345baf2b2f39aa8c06c735fccc2, which made the static
libs selectable under CONFIG_STATIC.
Change-Id: I585167e17459877e0fa7fa19e1046c3703d91c97
John Koleszar [Wed, 3 Aug 2011 12:43:43 +0000 (08:43 -0400)]
Fix asm offsets generation for universal-darwin builds
Added BUILD_PFX to correct dependencies.
Change-Id: Ia07602bd98ef2253242b1bd66ef05e3b1e64ba7d
Johann [Tue, 2 Aug 2011 23:20:27 +0000 (19:20 -0400)]
update extend frame borders
the neon code made several assumptions which were broken by a recent
change: https://review.webmproject.org/2676
update the code with new assumptions and guard them with a compile time
assert
Change-Id: I32a8378030759966068f34618d7b4b1b02e101a0
John Koleszar [Tue, 2 Aug 2011 18:41:32 +0000 (14:41 -0400)]
Update CHANGELOG for Cayuga release
Change-Id: If6f20553159105c05f9a684cb7c8f3778c7894a1
James Berry [Tue, 2 Aug 2011 17:42:03 +0000 (13:42 -0400)]
include asm_com/dec_offsets for make dist
Change-Id: Ia1ad66066a24c01915cd9e3ff75c7e070cc984c8
John Koleszar [Tue, 2 Aug 2011 14:09:59 +0000 (10:09 -0400)]
Update AUTHORS
Change-Id: Ic7725e279d2263515e5312c152c58e1644eb2495
John Koleszar [Tue, 2 Aug 2011 14:09:36 +0000 (10:09 -0400)]
Update .mailmap entry for Ralph Giles
Change-Id: Ib3d54bfa81720a0b2877837d7149cd12d26e75e4
Lou Quillio [Mon, 1 Aug 2011 19:34:48 +0000 (15:34 -0400)]
Sync `vpxenc --timebase` usage wording with docs change.
Change-Id: Ia406272a97806c0194435bb7f24e24d353ef5cc6
John Koleszar [Tue, 2 Aug 2011 12:21:10 +0000 (05:21 -0700)]
Merge "include the arm header files in make dist" into cayuga
John Koleszar [Tue, 2 Aug 2011 12:19:11 +0000 (05:19 -0700)]
Merge "Fix building with --disable-postproc" into cayuga
John Koleszar [Mon, 1 Aug 2011 20:42:14 +0000 (16:42 -0400)]
Fix building with --disable-postproc
Change-Id: I7e6bc28e7974a376da747300744e0dd5dc1d21e9
Johann [Mon, 1 Aug 2011 21:16:10 +0000 (17:16 -0400)]
include the arm header files in make dist
Change-Id: Ibcf5b4b14153f65ce1b53c3bfba87ad2feb17bbd
John Koleszar [Mon, 1 Aug 2011 20:56:32 +0000 (13:56 -0700)]
Merge "build error fix - obj_int_extract.bat" into cayuga
James Berry [Mon, 1 Aug 2011 20:10:41 +0000 (16:10 -0400)]
build error fix - obj_int_extract.bat
obj_int_extract.bat was not being copied
correctly for make dist. It now is.
Change-Id: I976479f90bbfa4798f241db1055e1e3b04ca2830
John Koleszar [Fri, 29 Jul 2011 18:23:32 +0000 (14:23 -0400)]
Disable FORTIFY_SOURCE on glibc targets
Improve binary distributions by defeating longjmp interception. See
http://code.google.com/p/webm/issues/detail?id=166 for more information.
Change-Id: I5ac731ec3f3570088597201d0f411473e2dffa4f
John Koleszar [Fri, 29 Jul 2011 20:56:43 +0000 (16:56 -0400)]
install asm_offsets.h
Ensure vpx_ports/asm_offsets.h is installed with make dist
Change-Id: If9f32273fff975d60de1583b039dbbce8a7ccd27
John Koleszar [Fri, 29 Jul 2011 18:57:48 +0000 (11:57 -0700)]
Merge "Convert rc_max_intra_bitrate_pct to control"
John Koleszar [Fri, 29 Jul 2011 16:26:55 +0000 (12:26 -0400)]
Correctly track sharpness in vp8cx_pick_filter_level_fast
Make sure to update last_sharpness_level from the current
sharpness_level whenever it changes.
Change-Id: I0258d2f5b11a407abf6176a8d4c4994d925943f0
John Koleszar [Thu, 28 Jul 2011 14:08:35 +0000 (07:08 -0700)]
Merge "configure: add --enable-static option"
John Koleszar [Thu, 28 Jul 2011 13:17:32 +0000 (09:17 -0400)]
Convert rc_max_intra_bitrate_pct to control
Since this is the only ABI incompatible change since the last release,
convert it to use the control interface instead. The member of the
configuration struct is replaced with the VP8E_SET_MAX_INTRA_BITRATE_PCT
control.
More significant API changes were expected to be forthcoming when this
control was first introduced, and while they continue to be expected,
it's not worth breaking compatibility for only this change.
Change-Id: I799d8dbe24c8bc9c241e0b7743b2b64f81327d59
Yunqing Wang [Fri, 22 Jul 2011 20:01:11 +0000 (16:01 -0400)]
Preload reference area in sub-pixel motion search (real-time mode)
This change implemented same idea in change "Preload reference area
to an intermediate buffer in sub-pixel motion search." The changes
were made to vp8_find_best_sub_pixel_step() and vp8_find_best_half
_pixel_step() functions which are called when speed >= 5. Test
result (using tulip clip):
1. On Core2 Quad machine(Linux)
rt mode, speed (-5 ~ -8), encoding speed gain: 2% ~ 3%
rt mode, speed (-9 ~ -11), encoding speed gain: 1% ~ 2%
rt mode, speed (-12 ~ -14), no noticeable encoding speed gain
2. On Xeon machine(Linux)
Test on speed (-5 ~ -14) didn't show noticeable speed change.
Change-Id: I21bec2d6e7fbe541fcc0f4c0366bbdf3e2076aa2
Yunqing Wang [Wed, 27 Jul 2011 16:34:13 +0000 (09:34 -0700)]
Merge "Fix range checks in motion search"
Yunqing Wang [Wed, 27 Jul 2011 14:37:33 +0000 (10:37 -0400)]
Fix range checks in motion search
There were some situations that the start motion vectors were
out of range. This fix adjusted range checks to make sure they
are checked and clamped.
Change-Id: Ife83b7fed0882bba6d1fa559b6e63c054fd5065d
James Zern [Mon, 25 Jul 2011 20:26:18 +0000 (13:26 -0700)]
vpxenc: cosmetics: timebase help update / spelling
The timebase update fixes Issue #61.
Change-Id: I425158da7ea639464f61e6dd604ac9e6c72b7266
John Koleszar [Tue, 26 Jul 2011 19:57:43 +0000 (12:57 -0700)]
Merge "cosmetics: consistently use [u]int64_t"
James Zern [Tue, 26 Jul 2011 01:44:59 +0000 (18:44 -0700)]
cosmetics: consistently use [u]int64_t
Removes mixed usage of (unsigned) long long and INT64.
Fixes Issue #208.
Change-Id: I220d3ed5ce4bb1280cd38bb3715f208ce23cf83a
Johann [Tue, 26 Jul 2011 13:34:31 +0000 (06:34 -0700)]
Merge ""Eliminated TOKENEXTRABITS" broke the windows build."
Scott LaVarnway [Tue, 26 Jul 2011 13:33:16 +0000 (09:33 -0400)]
"Eliminated TOKENEXTRABITS" broke the windows build.
Fixed.
Change-Id: I3348e8dbcaee6ace263af413701101d77636e5df
James Zern [Mon, 25 Jul 2011 22:40:36 +0000 (15:40 -0700)]
configure: add --enable-static option
Fixes issue #62.
Change-Id: I0567cf7897c0942666c19b3231c8c3b8e9c3e7cc
Scott LaVarnway [Mon, 25 Jul 2011 21:35:58 +0000 (14:35 -0700)]
Merge "Eliminated TOKENEXTRABITS"
Scott LaVarnway [Mon, 25 Jul 2011 21:11:24 +0000 (17:11 -0400)]
Eliminated TOKENEXTRABITS
Noticed small performance gains, depending on material.
Change-Id: I334369f6312bc19aa73481fc3f790ab181e11867
Yunqing Wang [Mon, 25 Jul 2011 19:40:50 +0000 (12:40 -0700)]
Merge "Use CONFIG_FAST_UNALIGNED consistently in codec"
Yunqing Wang [Mon, 25 Jul 2011 15:30:45 +0000 (11:30 -0400)]
Specify size for argument pushed to stack
The change fixes building error on Win64.
Change-Id: I63d25b26220c4da8a98ca2e36530cbb802468e6b
Yunqing Wang [Mon, 25 Jul 2011 14:11:24 +0000 (10:11 -0400)]
Use CONFIG_FAST_UNALIGNED consistently in codec
CONFIG_FAST_UNALIGNED is enabled by default. Disable it if it is
not supported by hardware.
Change-Id: I7d6905ed79fed918bca074bd62820b0c929d81ab
Johann [Fri, 22 Jul 2011 16:34:55 +0000 (09:34 -0700)]
Merge "fix sharpness bug and clean up"
Johann [Wed, 20 Jul 2011 19:53:42 +0000 (15:53 -0400)]
fix sharpness bug and clean up
sharpness was not recalculated in vp8cx_pick_filter_level_fast
remove last_filter_type. all values are calculated, don't need to update
the lfi data when it changes.
always use cm->sharpness_level. the extra indirection was annoying.
don't track last frame_type or sharpness_level manually. frame type
only matters for motion search and sharpness_level is taken care of in
frame_init
move function declarations to their proper header
Change-Id: I7ef037bd4bf8cf5e37d2d36bd03b5e22a2ad91db
Yunqing Wang [Fri, 22 Jul 2011 13:56:15 +0000 (06:56 -0700)]
Merge "Preload reference area to an intermediate buffer in sub-pixel motion search"
Yunqing Wang [Tue, 28 Jun 2011 13:14:13 +0000 (09:14 -0400)]
Preload reference area to an intermediate buffer in sub-pixel motion search
In sub-pixel motion search, the search range is small(+/- 3 pixels).
Preload whole search area from reference buffer into a 32-byte
aligned buffer. Then in search, load reference data from this buffer
instead. This keeps data in cache, and reduces the crossing cache-
line penalty. For tulip clip, tests on Intel Core2 Quad machine(linux)
showed encoder speed improvement:
3.4% at --rt --cpu-used =-4
2.8% at --rt --cpu-used =-3
2.3% at --rt --cpu-used =-2
2.2% at --rt --cpu-used =-1
Test on Atom notebook showed only 1.1% speed improvement(speed=-4).
Test on Xeon machine also showed less improvement, since unaligned
data access latency is greatly reduced in newer cores.
Next, I will apply similar idea to other 2 sub-pixel search functions
for encoding speed > 4.
Make this change exclusively for x86 platforms.
Change-Id: Ia7bb9f56169eac0f01009fe2b2f2ab5b61d2eb2f
Johann [Thu, 21 Jul 2011 19:56:59 +0000 (12:56 -0700)]
Merge "Add .size directive to ARM asm functions."
Johann [Thu, 21 Jul 2011 19:20:00 +0000 (12:20 -0700)]
Merge "Mark ARM asm objects as allowing a non-executable stack."
Timothy B. Terriberry [Tue, 19 Jul 2011 19:13:18 +0000 (12:13 -0700)]
Add .size directive to ARM asm functions.
This makes them show up properly in debugging tools like gdb and
valgrind.
Change-Id: I0c72548a1090de88ba226314e5efe63360b7e07f
Timothy B. Terriberry [Tue, 19 Jul 2011 20:09:22 +0000 (13:09 -0700)]
Mark ARM asm objects as allowing a non-executable stack.
This adds the magic .note.GNU-stack section at the end of each ARM
asm file (when built with gas), indicating that a non-executable
stack is allowed.
Without this section, the linker will assume the object requires an
executable stack by default, forcing an executable stack for the
entire program.
Change-Id: Ie86de6a449b52d392b9e5e0479833ed8c508ee65
John Koleszar [Thu, 21 Jul 2011 14:32:39 +0000 (07:32 -0700)]
Merge "Increase chrow row alignment to 16 bytes."
Yunqing Wang [Thu, 21 Jul 2011 14:27:08 +0000 (07:27 -0700)]
Merge "Add improvements made in good-quality mode to real-time mode"
Timothy B. Terriberry [Wed, 20 Jul 2011 17:20:31 +0000 (10:20 -0700)]
Increase chrow row alignment to 16 bytes.
This is done by expanding luma row to 32-byte alignment, since
there is currently a bunch of code that assumes that
uv_stride == y_stride/2 (see, for example, vp8/common/postproc.c,
common/reconinter.c, common/arm/neon/recon16x16mb_neon.asm,
encoder/temporal_filter.c, and possibly others; I haven't done a
full audit).
It also uses replaces the hardcoded border of 16 in a number of
encoder buffers with VP8BORDERINPIXELS (currently 32), as the
chroma rows start at an offset of border/2.
Together, these two changes have the nice advantage that simply
dumping the frame memory as a contiguous blob produces a valid,
if padded, image.
Change-Id: Iaf5ea722ae5c82d5daa50f6e2dade9de753f1003
Attila Nagy [Wed, 20 Jul 2011 11:09:42 +0000 (14:09 +0300)]
encoder: don't set the fragment bit for the last partition
Change-Id: Icb4e4f0d7c3074a8507852178be87541a1cb5bac
Scott LaVarnway [Tue, 19 Jul 2011 15:15:14 +0000 (08:15 -0700)]
Merge "Moved vp8_encode_bool into boolhuff.h"
John Koleszar [Tue, 19 Jul 2011 13:58:47 +0000 (09:58 -0400)]
Revert "Disable __longjmp_chk protection"
This reverts commit
b73a3693e581583e9ec676f4396d0c3d173e2462.
This version of the check doesn't work with generic-gnu, and figuring
out the correct symbol version at configure time is probably more work
than this is worth. May revisit in the future.
Change-Id: I6c75e88bd3bd82a4b21e09a25780fe53aacb7d70
Johann [Tue, 19 Jul 2011 13:20:38 +0000 (09:20 -0400)]
remove old armv5 code
armv5 dequantizer is not referenced
Change-Id: Id1cc617dcee35ebd6a406816ec6aaa26e8bbc8ad
Scott LaVarnway [Tue, 19 Jul 2011 13:17:25 +0000 (09:17 -0400)]
Moved vp8_encode_bool into boolhuff.h
allowing the compiler to inline this function. For real-time
encodes, this gave a boost of 1% to 2.5%, depending on the
speed setting.
Change-Id: I3929d176cca086b4261267b848419d5bcff21c02
John Koleszar [Wed, 29 Jun 2011 15:41:50 +0000 (11:41 -0400)]
Improved 1-pass CBR rate control
This patch attempts to improve the handling of CBR streams with
respect to the short term buffering requirements. The "buffer level"
is changed to be an average over the rc buffer, rather than a long
running average. Overshoot is also tracked over the same interval
and the golden frame targets suppressed accordingly to correct for
overly aggressive boosting.
Testing shows that this is fairly consistently positive in one
metric or another -- some clips that show significant decreases
in quality have better buffering characteristics, others show
improvenents in both.
Change-Id: I924c89aa9bdb210271f2e03311e63de3f1f8f920
John Koleszar [Mon, 18 Jul 2011 15:43:59 +0000 (08:43 -0700)]
Merge "Disable __longjmp_chk protection"
John Koleszar [Mon, 18 Jul 2011 13:07:50 +0000 (06:07 -0700)]
Merge "Fixed rate histogram calculation"
Tero Rintaluoma [Fri, 15 Jul 2011 11:47:44 +0000 (14:47 +0300)]
Fixed rate histogram calculation
Using small values for --buf-sz= in command line causes
floating point exception due to division by zero.
Change-Id: Ibfe2d44db922993a78ebc9a4a1087d9625de48ae
Scott LaVarnway [Fri, 15 Jul 2011 14:54:14 +0000 (07:54 -0700)]
Merge "Tokenize MB optimized"
Yunqing Wang [Fri, 15 Jul 2011 12:21:35 +0000 (05:21 -0700)]
Merge "Fix vpxenc encoding incorrect webm file header on big endian machines(Issue 331)"
Tero Rintaluoma [Mon, 11 Jul 2011 12:55:45 +0000 (15:55 +0300)]
Tokenize MB optimized
Optimized C-code of the following functions:
- vp8_tokenize_mb
- tokenize1st_order_b
- tokenize2nd_order_b
Gives ~1-5% speed-up for RT encoding on Cortex-A8/A9
depending on encoding parameters.
Change-Id: I6be86104a589a06dcbc9ed3318e8bf264ef4176c
James Berry [Thu, 14 Jul 2011 19:53:50 +0000 (15:53 -0400)]
bug fix vpx_copy_and_extend_frame size issue
vpx_copy_and_extend_frame could incorrectly
resize uv frames which could result in a crash.
Change-Id: Ie96f7078b1e328b3907a06eebeee44ca39a2e898
John Koleszar [Thu, 14 Jul 2011 18:18:47 +0000 (14:18 -0400)]
Remove unused speed features
min_fs_radius, max_fs_radius, full_freq were set but never read.
Change-Id: I82657f4e7f2ba2acc3cbc3faa5ec0de5b9c6ec74
Fritz Koenig [Wed, 13 Jul 2011 21:18:11 +0000 (14:18 -0700)]
Merge "Better allocate yuv buffers."
Yunqing Wang [Wed, 13 Jul 2011 20:32:57 +0000 (13:32 -0700)]
Merge "Fix unnecessary casting of B_PREDICTION_MODE (issue 349)"
John Koleszar [Mon, 11 Jul 2011 15:25:25 +0000 (11:25 -0400)]
Disable __longjmp_chk protection
glibc implements some checking on longjmp() calls by replacing it with
an internal function __longjmp_chk(), when FORTIFY_SOURCE is defined.
This can be problematic when compiling the library under one version of
glibc and running it under another. Work around this issue for the one
symbol affected for now, before taking out the undef hammer.
Fixes http://code.google.com/p/webm/issues/detail?id=166
Change-Id: Ifc5e25cdec17915e394711f2185b3e9214572d10
Yunqing Wang [Tue, 12 Jul 2011 21:22:36 +0000 (17:22 -0400)]
Fix unnecessary casting of B_PREDICTION_MODE (issue 349)
Minor fix.
Change-Id: Iaf93f6e47e882a33c479e57c7a0d0bf321e291c0
Yunqing Wang [Wed, 13 Jul 2011 18:51:02 +0000 (14:51 -0400)]
Add improvements made in good-quality mode to real-time mode
Several improvements we made in good-quality mode can be added
into real-time mode to speed up encoding in speed 1, 2, and 3
with small quality loss. Tests using tulip clip showed:
--rt --cpu-used=-1
(before change)
PSNR: 38.028
time: 1m33.195s
(after change)
PSNR: 38.014
time: 1m20.851s
--rt --cpu-used=-2
(before change)
PSNR: 37.773
time: 0m57.650s
(after change)
PSNR: 37.759
time: 0m54.594s
--rt --cpu-used=-3
(before change)
PSNR: 37.392
time: 0m42.865s
(after change)
PSNR: 37.375
time: 0m41.949s
Change-Id: I76ab2a38d72bc5efc91f6fe20d332c472f6510c9
Fritz Koenig [Mon, 11 Jul 2011 18:42:28 +0000 (11:42 -0700)]
Better allocate yuv buffers.
Previously allocated more memory than necessary for yuv buffers.
This makes it harder to track bugs with reading uninitialized
data.
Change-Id: I510f7b298d3c647c869be6e5d51608becc63cce9
Fritz Koenig [Wed, 13 Jul 2011 17:07:30 +0000 (10:07 -0700)]
Merge "Reduce motion vector search on alt-ref frame."
John Koleszar [Wed, 13 Jul 2011 12:46:30 +0000 (05:46 -0700)]
Merge "Remove rotting NDS_NITRO code."
Johann [Wed, 13 Jul 2011 11:10:03 +0000 (04:10 -0700)]
Merge "update x86 asm for loopfilter"
Johann [Wed, 13 Jul 2011 11:09:55 +0000 (04:09 -0700)]
Merge "Update armv6 loopfilter to new interface"
Johann [Wed, 13 Jul 2011 11:09:42 +0000 (04:09 -0700)]
Merge "Update armv7 loopfilter to new interface"