Paul Wilkins [Thu, 3 Jan 2013 15:14:36 +0000 (15:14 +0000)]
Further change to mv reference search.
This experimental change reorders the search so
that all possible references that match the target
reference frame are tested first and these in order
of distance from the current block. These will usually
be the highest scoring candidates.
If we do not find enough good candidates this way
we try non matching cases. These will usually be lower
scoring candidates.
The change in order together with breakouts when
we have found enough candidates should reduce
the computational cost and especially reduce the number
of sort operations.
Quality Results:
Std Hd +0.228%, Hd +0.074%, YT +0.046%, derf +0.137%
This effect is probably due to the fact that more distant
weak candidates are now less likely to get "promoted" over
near candidates even if they are repeated.
Adrian Grange [Thu, 20 Dec 2012 22:56:19 +0000 (14:56 -0800)]
New interpolation filter selection algorithm
Old Scheme:
When SWITCHABLE filter selection is enabled the encoder
evaluates the use of each interpolation filter type and
selects the best one to use at the MB level. A frame-
level flag can be set to force the use of a particular
filter type for all MBs in a frame if it is more efficient
to encode that way. The logic here involved a Q dependent
threshold that assumed that the second 8-tap filter was
a high-pass filter. However, this requires a trip around
the recode loop. If the frame-level flag indicates use
of a particular filter, the other filters are not
evaluated in the pick_mode loop.
New Scheme:
Each filter type is evaluated at the MB level and a record
of the best filter is kept, irrespective of what filter
is signaled at the frame-level. Once all MBs have been
encoded, a decision is made as to what frame-level mode
to set for the *next* frame. If one filter is used by 80%
or more of the MBs, then this filter is forced since it
is assumed that this will be more efficient if the
next frame has similar characteristics. i.e. there is a
one-frame lag between measuring the filter selection and
setting the frame-level mode to use.
Yaowu Xu [Thu, 3 Jan 2013 16:00:00 +0000 (08:00 -0800)]
Merge cost_coeffs_2x2() into cost_coeffs()
Remove special case function cost_coeffs_2x2() and change function
cost_coeffs() to handle 2nd order haar block as it is handle all
other block types already.
Yunqing Wang [Fri, 28 Dec 2012 00:04:44 +0000 (16:04 -0800)]
Skip finding best ref_mvs when the mode is ZEROMV
Read mode before calling vp9_find_best_ref_mvs(). If the mode is
ZEROMV, the best ref_mvs are not needed. Then, we can skip calling
vp9_find_best_ref_mvs().
Yunqing Wang [Thu, 27 Dec 2012 21:48:17 +0000 (13:48 -0800)]
Switch the order of calculating 2-D inverse transform
The 2-D inverse transform X = M1*Z*Transposed_M2 was calculated
in 2 steps from left to right:
1. Vertical transform: Y = M1*Z
2. Horizontal transform: X= Y*Transposed_M2
In SIMD, a transpose is needed in vertical transform.
Here, switched the calculation order to do it from right to left.
In this way, we could eliminate that transpose by writing the
intermediate results out to their transposed positions.
Deb Mukherjee [Tue, 27 Nov 2012 23:51:06 +0000 (15:51 -0800)]
New previous coef context experiment
Adds an experiment to derive the previous context of a coefficient
not just from the previous coefficient in the scan order but from a
combination of several neighboring coefficients previously encountered
in scan order. A precomputed table of neighbors for each location
for each scan type and block size is used. Currently 5 neighbors are
used.
Results are about 0.2% positive using a strategy where the max coef
magnitude from the 5 neigbors is used to derive the context.
John Koleszar [Wed, 19 Dec 2012 21:44:32 +0000 (13:44 -0800)]
make: fix dependency generation
Remove an extra level of escaping around the $@ variable to get valid output.
Prior to this change, modifying header files did not trigger a rebuild of
sources dependent on them.
John Koleszar [Fri, 16 Nov 2012 18:48:23 +0000 (10:48 -0800)]
Use boolcoder API instead of inlining
This patch changes the token packing to call the bool encoder API rather
than inlining it into the token packing function, and similarly removes
a special get_signed case from the detokenizer. This allows easier
experimentation with changing the bool coder as a whole.
Ronald S. Bultje [Tue, 18 Dec 2012 23:31:19 +0000 (15:31 -0800)]
Use standard integer types for pixel values and coefficients.
For coefficients, use int16_t (instead of short); for pixel values in
16-bit intermediates, use uint16_t (instead of unsigned short); for all
others, use uint8_t (instead of unsigned char).
Yaowu Xu [Tue, 18 Dec 2012 22:36:20 +0000 (14:36 -0800)]
Changed MAX_PSNR to 100
The MAX_PSNR was used to assign a "psnr" number when the mse is close
to zero. The direct assignment is used to prevent divide by zero in
computation. Changing it from 60 to 100 to be consistent against what
is being done in VP9
Deb Mukherjee [Wed, 12 Dec 2012 01:06:35 +0000 (17:06 -0800)]
Further improvements on the hybrid dwt/dct expt
Modifies the scanning pattern and uses a floating point 16x16
dct implementation for now to handle scaling better.
Also experiments are in progress with 2/6 and 9/7 wavelets.
Results have improved to within ~0.25% of 32x32 dct for std-hd
and about 0.03% for derf. This difference can probably be bridged by
re-optimizing the entropy stats for these transforms. Currently
the stats used are common between 32x32 dct and dwt/dct.
Experiments are in progress with various scan pattern - wavelet
combinations.
Ideally the subbands should be tokenized separately, and an
experiment will be condcuted next on that.
Ronald S. Bultje [Wed, 12 Dec 2012 18:25:58 +0000 (10:25 -0800)]
New default coefficient/band probabilities.
Gives 0.5-0.6% improvement on derf and stdhd, and 1.1% on hd. The
old tables basically derive from times that we had only 4x4 or
only 4x4 and 8x8 DCTs.
Note that some values are filled with 128, because e.g. ADST ever
only occurs as Y-with-DC, as does 32x32; 16x16 ever only occurs
as Y-with-DC or as UV (as complement of 32x32 Y); and 8x8 Y2 ever
only has 4 coefficients max. If preferred, I can add values of
other tables in their place (e.g. use 4x4 2nd order high-frequency
probabilities for 8x8 2nd order), so that they make at least some
sense if we ever implement a larger 2nd order transform for the
8x8 DCT (etc.), please let me know
Ronald S. Bultje [Mon, 10 Dec 2012 20:09:07 +0000 (12:09 -0800)]
Consistently use get_prob(), clip_prob() and newly added clip_pixel().
Add a function clip_pixel() to clip a pixel value to the [0,255] range
of allowed values, and use this where-ever appropriate (e.g. prediction,
reconstruction). Likewise, consistently use the recently added function
clip_prob(), which calculates a binary probability in the [1,255] range.
If possible, try to use get_prob() or its sister get_binary_prob() to
calculate binary probabilities, for consistency.
Since in some places, this means that binary probability calculations
are changed (we use {255,256}*count0/(total) in a range of places,
and all of these are now changed to use 256*count0+(total>>1)/total),
this changes the encoding result, so this patch warrants some extensive
testing.
Paul Wilkins [Wed, 5 Dec 2012 16:23:38 +0000 (16:23 +0000)]
Further changes to mv reference code.
Some further changes and refactoring of mv
reference code and selection of center point for
searches. Mainly relates to not passing so many
different local copies of things around.
Yaowu Xu [Wed, 28 Nov 2012 23:15:51 +0000 (15:15 -0800)]
experiment with CONTEXT conversion
This commit changed the ENTROPY_CONTEXT conversion between MBs that
have different transform sizes.
In additioin, this commit also did a number of cleanup/bug fix:
1. removed duplicate function vp9_fix_contexts() and changed to use
vp8_reset_mb_token_contexts() for both encoder and decoder
2. fixed a bug in stuff_mb_16x16 where wrong context was used for
the UV.
3. changed reset all context to 0 if a MB is skipped to simplify the
logic.
John Koleszar [Thu, 6 Dec 2012 21:56:25 +0000 (13:56 -0800)]
libvpx_test: ensure rtcd init functions are called
In addition to allowing tests to use the RTCD-enabled functions (perhaps transitively)
without having run a full encode/decode test yet, this fixes a linking issue with
Apple's G++ whereby the Common symbols (the function pointers themselves) wouldn't
be resolved. Fixing this linking issue is the primary impetus for this patch, as none
of the tests exercise the RTCD functionality except through the main API.
Don't use vp9_decode_coefs_4x4() for 2nd order DC or luma blocks. The
code introduces some overhead which is unnecessary for these cases.
Also, remove variable declarations that are only used once, remove
magic offsets into the coefficient buffer (use xd->block[i].qcoeff
instead of xd->qcoeff + magic_offset), and fix a few Google Style
Guide violations.
Use these, instead of the 4/5-dimensional arrays, to hold statistics,
counts, accumulations and probabilities for coefficient tokens. This
commit also re-allows ENTROPY_STATS to compile.
This adds Debargha's DCT/DWT hybrid and a regular 32x32 DCT, and adds
code all over the place to wrap that in the bitstream/encoder/decoder/RD.
Some implementation notes (these probably need careful review):
- token range is extended by 1 bit, since the value range out of this
transform is [-16384,16383].
- the coefficients coming out of the FDCT are manually scaled back by
1 bit, or else they won't fit in int16_t (they are 17 bits). Because
of this, the RD error scoring does not right-shift the MSE score by
two (unlike for 4x4/8x8/16x16).
- to compensate for this loss in precision, the quantizer is halved
also. This is currently a little hacky.
- FDCT and IDCT is double-only right now. Needs a fixed-point impl.
- There are no default probabilities for the 32x32 transform yet; I'm
simply using the 16x16 luma ones. A future commit will add newly
generated probabilities for all transforms.
- No ADST version. I don't think we'll add one for this level; if an
ADST is desired, transform-size selection can scale back to 16x16
or lower, and use an ADST at that level.
Additional notes specific to Debargha's DWT/DCT hybrid:
- coefficient scale is different for the top/left 16x16 (DCT-over-DWT)
block than for the rest (DWT pixel differences) of the block. Therefore,
RD error scoring isn't easily scalable between coefficient and pixel
domain. Thus, unfortunately, we need to compute the RD distortion in
the pixel domain until we figure out how to scale these appropriately.
John Koleszar [Thu, 6 Dec 2012 21:56:25 +0000 (13:56 -0800)]
libvpx_test: ensure rtcd init functions are called
In addition to allowing tests to use the RTCD-enabled functions (perhaps transitively)
without having run a full encode/decode test yet, this fixes a linking issue with
Apple's G++ whereby the Common symbols (the function pointers themselves) wouldn't
be resolved. Fixing this linking issue is the primary impetus for this patch, as none
of the tests exercise the RTCD functionality except through the main API.
Johann [Mon, 3 Dec 2012 20:26:51 +0000 (12:26 -0800)]
Begin to refactor vpx_scale usage in VP9
Only declare the functions in vpx_scale RTCD and include the relevant
header.
Remove unused files and functions in vpx_scale to avoid wasting time
renaming. vpx_scale/win32/scaleopt.c contains functions which have not
been called in a long time but are potentially optimized.
The 'vp8' functions have not been renamed yet. That is for after the
cleanup.
Paul Wilkins [Tue, 4 Dec 2012 17:21:05 +0000 (17:21 +0000)]
Change to MV reference search.
This patch reduces the cpu cost of the MV ref
search by only allowing insert for candidates
that would be in the current top 4.
This could alter the outcome and slightly favors
near candidates which are tested first but also
limits the worst case loop count to 4 and means in
many cases it will drop out and not happen.
Yaowu Xu [Tue, 4 Dec 2012 16:35:37 +0000 (08:35 -0800)]
Fix the build with MSVC
1. remove the dependency on non existing "vp9_temporal_filter_x86.h"
2. prefix filenames with vp9_ in obj_int_extract.bat to reflect the
change of the actual filenames.
Yaowu Xu [Mon, 3 Dec 2012 22:53:45 +0000 (14:53 -0800)]
merged optimiz_b_16x16() into optmize_b()
The commit changed the trellis quantization function optimize_b() to
work for MBs using all transform sizes, and eliminated the function
for MB using 16x16 transform only, optimize_b_16x16.
Johann [Mon, 3 Dec 2012 20:26:51 +0000 (12:26 -0800)]
Begin to refactor vpx_scale usage in VP9
Only declare the functions in vpx_scale RTCD and include the relevant
header.
Remove unused files and functions in vpx_scale to avoid wasting time
renaming. vpx_scale/win32/scaleopt.c contains functions which have not
been called in a long time but are potentially optimized.
The 'vp8' functions have not been renamed yet. That is for after the
cleanup.
projects / libvpx / log