Scott LaVarnway [Fri, 23 Mar 2012 19:50:08 +0000 (15:50 -0400)]
New vp8_decode_mb_tokens()
This new vp8_decode_mb_tokens() uses a modified version of
WebP's GetCoeffs function. For now, the dequant does not
occur in GetCoeffs.
Tests showed performance improvements up to 2.5% depending
on material.
Deb Mukherjee [Thu, 22 Mar 2012 16:13:18 +0000 (09:13 -0700)]
Miscellaneous changes in mfqe and postproc modules
Adds logic to disable mfqe for the first frame after a configuration
change such as change in resolution. Also adds some missing
if CONFIG_POSTPROC macro checks.
Yunqing Wang [Wed, 14 Mar 2012 14:03:39 +0000 (10:03 -0400)]
Add motion search skipping in first pass
This change added a motion search skipping mechanism similar
to what we did in second pass. For a macroblock that is very
similar to the macroblock at same location on last frame,
we can set its mv to be zero, and skip motion search. This
improves first-pass performance for slide shows and video
conferencing clips with a slight PSNR loss.
John Koleszar [Tue, 21 Feb 2012 22:41:39 +0000 (14:41 -0800)]
Update XCode SDK search paths
Newer XCodes have moved the SDK path from /Developer/SDKs
Use a suggestion from jorgenisaksson@gmail.com to locate it
osx_sdk_dir is not required to be set. Apple now offers a set
command line tools which do not require this. isysroot is also
not required in newer versions of XCode so only set it when we
are confident in the location.
There remain issues with the iOS configure steps which will be
addressed later
James Zern [Thu, 15 Mar 2012 23:51:51 +0000 (16:51 -0700)]
doxy: fix conditional usage, ref warnings
doxygen < 1.7.? seems to have been more tolerant of single line
\if/\endif
This change fixes warnings such as:
mainpage.dox:13: warning: unable to resolve reference to `vp8_encoder-'
for \ref command
vpx_decoder.h:193: warning: explicit link request to 'n' could not be
resolved
Johann [Fri, 27 Jan 2012 18:23:52 +0000 (10:23 -0800)]
RFC: Reorganize MFQE loops
Break MFQE code into it's own file.
It is currently only valid for 16x16 and 8x8 Y blocks. It also filters
4x4 U/V blocks.
Refactor filtering and add associated assembly. Limited test cases show
--mfqe introduces a penalty of ~20% with HD content. The assembly
reduces the penalty to ~15%
Johann [Tue, 6 Mar 2012 00:36:23 +0000 (16:36 -0800)]
include CHANGELOG in CODEC_SRCS
build/make/version.sh requires CHANGELOG to generate vpx_version.h
The file is already included when building the documentation. However,
documentation is not build if doxygen/php are not present.
This is necessary when using '--enable-install-srcs --enable-codec-srcs'
and 'make dist'
Johann [Fri, 2 Mar 2012 00:12:53 +0000 (16:12 -0800)]
Fix encoder debug setting
Propagate debug setting to the EBML struct. When writing the application
name, this allows us to strip the version code and keep the output
metadata static.
Yunqing Wang [Wed, 29 Feb 2012 13:24:53 +0000 (08:24 -0500)]
vpxenc: fix time and fps calculation in 2-pass encoding
When we do 2-pass encoding, elapsed time is accumulated through
whole 2-pass process, which gives incorrect time and fps results
for second pass. This change fixed that by resetting the time
accumulator for second pass.
Attila Nagy [Thu, 9 Feb 2012 10:37:03 +0000 (12:37 +0200)]
Packing bitstream on-the-fly with delayed context updates
Produce the token partitions on-the-fly, while processing each MB.
Context is updated at the beginning of each frame based on the
previoud frame's counters. Optimally encoder outputs partitions in
separate buffers. For frame based output, partitions are concatenated
internally.
Limitations:
- enabled just in combination with realtime-only mode
- number of encoding threads has to be equal or less than the
number of token partitions. For this reason, by default the encoder
will do 8 token partitions.
- vpxenc supports partition output (-P) just in combination with
IVF output format (--ivf)
Performance:
- Realtime encoder can be up to 13% faster (ARM) depending on the number
of threads and bitrate settings. Constant gain over the 5-16 speed
range.
- Token buffer reduced from one frame to 8 MBs
Quality:
- quality is affected by the delayed context updates. This again
dependents on input material, speed and bitrate settings. For VC
style input the loss seen is up to 0.2dB. If error-resilient=2
mode is used than the effect of this change is negligible.
Scott LaVarnway [Tue, 28 Feb 2012 19:12:30 +0000 (14:12 -0500)]
Eliminated reconintra_mt.c
Reworked the code to use vp8_build_intra_predictors_mby_s,
vp8_intra_prediction_down_copy, and vp8_intra4x4_predict_d_c
functions instead. vp8_intra4x4_predict_d_c is a decoder-only
version of vp8_intra4x4_predict. Future commits will fix this
code duplication.
Yunqing Wang [Mon, 27 Feb 2012 22:38:53 +0000 (17:38 -0500)]
Only do uv intra-mode evaluation when intra mode is checked
When we encode slide-show clips, for the majority of the time,
only ZEROMV mode is checked, and all other modes are skipped.
This change delayed uv intra-mode evaluation until intra mode is
actually checked. This gave big performance gain for slide-show
video encoding (2nd pass gain: 18% to 28%). But, this change
doesn't help other types of videos.
Also, zbin_mode_boost is adjusted in mode-checking loop, which
causes bitstream mismatch before/after this change when --best
or --good with --cpu-used=0 are used.
Marco Paniconi [Wed, 22 Feb 2012 22:39:17 +0000 (14:39 -0800)]
Remove the frame rate factor for key frame size.
When temporal layers is used (i.e., number_of_layers > 1),
we don't use the frame rate boost for setting the key
frame target size. The factor was forcing the target size to be
always at its minimum (2* per_frame_bandwidth) for low frame rates
(i.e., base layer frame rate).
Generally we should modify or remove this frame rate factor;
for now we turn if off for number_of_layers > 1.
Yunqing Wang [Fri, 17 Feb 2012 14:15:08 +0000 (09:15 -0500)]
Fix incorrect use of uv eobs in intra modes
In vp8_rd_pick_inter_mode(), if total of eobs is zero, rate needs
to be adjusted since there are no non-zero coefficients for
transmission. The uv intra eobs calculated in
rd_pick_intra_mbuv_mode() need to be saved before they are
overwritten by inter-mode eobs.
Attila Nagy [Fri, 17 Feb 2012 09:50:33 +0000 (11:50 +0200)]
Update encoder mb_skip_coeff and prob_skip_false calculation
mode_info_context->mbmi.mb_skip_coeff has to always reflect the
existence or not of coeffs for a certain MB. The loopfilter needs this
info.
mb_skip_coeff is either set by the vp8_tokenize_mb or has to be set to
1 when the MB is skipped by mode selection. This has to be done
regardless of the mb_no_coeff_skip value.
prob_skip_false is needed just when mb_no_coeff_skip is 1. No need to
keep count of both skip_false and skip_true as they are complementary
(skip_true+skip_false = total_mbs)
John Koleszar [Wed, 15 Feb 2012 20:39:38 +0000 (12:39 -0800)]
vpxenc: initial implementation of multistream support
Add the ability to specify multiple output streams on the command line.
Streams are delimited by --, and most parameters inherit from previous
streams.
In this implementation, resizing streams is still not supported. It
does not make use of the new multistream support in the encoder either.
Two pass support runs all streams independently, though it's
theoretically possible that we could combine firstpass runs in the
future. The logic required for this is too tricky to do as part of this
initial implementation. This is mostly an effort to get the parameter
passing and independent streams working from the application's
perspective, and a later commit will add the rescaling and
multiresolution support.
The text of the bug is somewhat misleading as I initially read it to
imply the bug was present in v0.9.7-p1 (Cayuga), but note the text
"master", which indicates this was something subsequent. This issue
bisects back to v0.9.7-p1-84-ga99c20c, so unfortunately it was broken
during the Duclair release.
Thanks to Alexei Leonenko for investigating the root cause.
Attila Nagy [Fri, 16 Sep 2011 10:54:06 +0000 (13:54 +0300)]
Multithreaded encoder, late sync loopfilter
Second shot at this...
Sync with loopfilter thread as late as possible, usually just at the
beginning of next frame encoding. This returns control to application
faster and allows a better multicore scaling.
When PSNR packets are generated the final filtered frame is needed
imediatly so we cannot delay the sync. Same has to be done when
internal frame is previewed.
Johann [Thu, 9 Feb 2012 20:38:31 +0000 (12:38 -0800)]
Fix variance overflow
In the variance calculations the difference is summed and later squared.
When the sum exceeds sqrt(2^31) the value is treated as a negative when
it is shifted which gives incorrect results.
To fix this we cast the result of the multiplication as unsigned.
The alternative fix is to shift sum down by 4 before multiplying.
However that will reduce precision.
For 16x16 blocks the maximum sum is 65280 and sqrt(2^31) is 46340 (and
change).
projects / libvpx / log