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.
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.
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.
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.
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.
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.
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.
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:
Yunqing Wang [Thu, 7 Jul 2011 15:21:41 +0000 (11:21 -0400)]
Adjust full-pixel clamping and motion vector limit calculation
Do mvp clamping in full-pixel precision instead of 1/8-pixel
precision to avoid error caused by right shifting operation.
Also, further fixed the motion vector limit calculation in change: b7480454706a6b15bf091e659cd6227ab373c1a6
Attila Nagy [Fri, 10 Jun 2011 11:10:21 +0000 (14:10 +0300)]
New loop filter interface
Separate simple filter with reduced no. of parameters.
MB filter level picking based on precalculated table. Level table updated for
each frame. Inside and edge limits precalculated and updated just when
sharpness changes. HEV threshhold is constant.
ARM targets use scalars and others vectors.
Change works only with --target=generic-gnu
All other targets have to be updated!
Yunqing Wang [Thu, 30 Jun 2011 15:20:13 +0000 (11:20 -0400)]
Bug fix in motion vector limit calculation
Motion vector limits are calculated using right shifts, which
could give wrong results for negative numbers. James Berry's
test on one clip showed encoder produced some artifacts. This
change fixed that.
Paul Wilkins [Tue, 28 Jun 2011 16:29:47 +0000 (17:29 +0100)]
Change to arf boost calculation.
In this commit I have added an experimental function
that tests prediction quality either side of a central position
to calculate a suggested boost number for an ARF frame.
The function is passed an offset from the current position and
a number of frames to search forwards and backwards.
It returns a forward, backward and compound boost number.
The new code can be deactivated using #define NEW_BOOST 0
In its current default state the code searches forwards and backwards
from the proposed position of the next alt ref.
The the old code used a boost number calculated by scanning forward
from the previous GF up to the proposed alt ref frame position.
I have also added some code to try and prevent placement of a gf/arf
where there is a brief flash.
John Koleszar [Wed, 29 Jun 2011 13:53:12 +0000 (09:53 -0400)]
vpxenc: prevent wraparound in the --rate-hist ringbuffer
For clips that are near 60fps and have a lot of alt refs, it's possible
that the ring buffer holding the previous frames sizes/pts could wrap
around, leading to a division by zero.
In addition to checking for this condition in the ring buffer loop,
the buffer size is made dependent on the actual frame rate in use,
rather than defaulting to 60, which should improve accuracy at frame
rates >= ~60.
John Koleszar [Tue, 28 Jun 2011 21:03:47 +0000 (17:03 -0400)]
Use MAX_ENTROPY_TOKENS and ENTROPY_NODES more consistently
There were many instances in the code of vp8_coef_tokens and
vp8_coef_tokens-1, which was a preprocessor macro despite the naming
convention. Replace these with MAX_ENTROPY_TOKENS and ENTROPY_NODES,
respectively.
Stefan Holmer [Mon, 13 Jun 2011 14:42:27 +0000 (16:42 +0200)]
New ways of passing encoded data between encoder and decoder.
With this commit frames can be received partition-by-partition
from the encoder and passed partition-by-partition to the
decoder.
At the encoder-side this makes it easier to split encoded
frames at partition boundaries, useful when packetizing
frames. When VPX_CODEC_USE_OUTPUT_PARTITION is enabled,
several VPX_CODEC_CX_FRAME_PKT packets will be returned
from vpx_codec_get_cx_data(), containing one partition
each. The partition_id (starting at 0) specifies the decoding
order of the partitions. All partitions but the last has
the VPX_FRAME_IS_FRAGMENT flag set.
At the decoder this opens up the possibility of decoding partition
N even though partition N-1 was lost (given that independent
partitioning has been enabled in the encoder) if more info
about the missing parts of the stream is available through
external signaling.
Each partition is passed to the decoder through the
vpx_codec_decode() function, with the data pointer pointing
to the start of the partition, and with data_sz equal to the
size of the partition. Missing partitions can be signaled to
the decoder by setting data != NULL and data_sz = 0. When
all partitions have been given to the decoder "end of data"
should be signaled by calling vpx_codec_decode() with
data = NULL and data_sz = 0.
The first partition is the first partition according to the
VP8 bitstream + the uncompressed data chunk + DCT address
offsets if multiple residual partitions are used.
projects / libvpx / log