the final macroblock rows are scheduled in the main thread. prior to
this change one additional macroblock row would be scheduled in the
worker forcing the main thread to wait before finishing.
Change-Id: I05f3168e5c629b898fcebb0d77eb6d6a90d6105e
Currently, the best quality mode in VP9 is not very well developed,
and unnecessarily makes the encode too slow. Hence the command line
default is changed to "good" quality. Also, the number of passes
default is changed to 2 passes as well, since 1-pass encoding is
not very efficient in VP9.
Besides, a number of VP9 defaults are set to the currently
recommended settings. With these changes, vpxenc
run with --codec=vp9 --kf-max-dist=9999 --cpu-used=0 should
work about the same as our borg results.
Note when the --cpu-used=0 option is dropped there will be a slight
difference in the output, because of a difference in the cpu-used
value for the first pass. Specifically, the default when unspecified
is to use cpu_used=1 for the first pass and cpu_used=0 for the
second pass. But when specified, both passes will use the cpu-used
value specified.
Note that this also changes the default for VP8 as being "good"
but other options stay unchanged.
Change-Id: Ib23c1a05ae2f36ee076c0e34403efbda518c5066
Adding set_contexts contexts function and call it instead of
set_contexts_on_border. Calling txfrm_block_to_raster_xy to get aoff and
loff.
Change-Id: I41897e344afd2cae1f923f4fdbe63daccf6fe80e
Moving foreach_predicted_block_in_plane function to vp9_reconinter.c
because there is only one usage.
Change-Id: I9852feae43fc3cf809b817fc541d043bc5496209
Updating implementation of vp9_get_pred_context_single_ref_p2 using
has_second_ref function to make code easier to read.
Change-Id: I5ba642712f59861a48aab974e73aa01640d086fe
vp9_short_idct10_8x8_add is used to handle the block that only have valid data
at top left 4x4 block. All the other datas are 0. So we could cut several
unnecessary calculations in order to save instructions.
Change-Id: I34fda95e29082b789aded97c2df193991c2d9195
Check the minimum rate-distortion cost of regular quantization and
all zero coeffs cases in the sub8x8 inter prediction rd loop for
luma components. Use this as the cumulative rdcost sent to UV rd
estimation.
Change-Id: Ia4bc7700437d5e13d7cdad4cf9ae57ab036d3e97
Cleans up the switchable filter search logic. Also adds a
speed feature - a variance threshold - to disable filter search
if source variance is lower than this value.
Results: derfraw300
threshold = 16, psnr -0.238%, 4-5% speedup (tested on football)
threshold = 32, psnr -0.381%, 8-9% speedup (tested on football)
threshold = 64, psnr -0.611%, 12-13% speedup (tested on football)
threshold = 96, psnr -0.804%, 16-17% speedup (tested on football)
Based on these results, the threshold is chosen as 16 for speed 1,
32 for speed 2, 64 for speed 3 and 96 for speed 4.
Change-Id: Ib630d39192773b1983d3d349b97973768e170c04
Changes to code to auto select a partition size range
based on data from spatial neighbors.
Now looks at the sb_type in each 8x8 block of above
and left SB64.
The effect on speed 1 is now weaker giving better
quality but less speed gain. Now also used in speed 2.
Change-Id: Iace33a97d5c3498dd2a9a8a4067351941abcbabc
Updating implementation of vp9_get_pred_context_single_ref_p1 using
has_second_ref function to make code easier to read.
Change-Id: Ie8f60403a7195117ceb2c6c43176ca9a9e70b909
As the pixel values beyond image border are duplicates of pixels
on edge, the change limits the mv search range, any mv beyond
the limits no longer produce new/different prediction values
as entire block with pixels used for subpel interpolation are
outside image border.
Change-Id: I4c6fdf06e33c1cef1489f5470ce0fb4e5e01fb79
For certain partition size, the function poniter may not be intialized
at all. The patch prevent the call if the pointer is not set.
Change-Id: I78b8c3992b639e8799a16b3c74f0973d07b8b9ac
This commit enables early termination in the rate-distortion
optimization search loop for chroma components. When the cumulative
rd cost is above the current best value, skip the rest per-block
transform/quantization/coeff_cost and continue to the next
prediction mode.
For bus_cif at 2000 kbps, the average run-time goes down from
168546ms -> 164678ms, (2% speed-up) at speed 0
36197ms -> 34465ms, (4% speed-up) at speed 1
Change-Id: I9d3043864126e62bd0166250d66b3170d520b3c0
Updating all foreach_transformed_block_visitor functions to work with
plane block size instead of general block. Removing a lot of duplicated
code.
Change-Id: I6a9069e27528c611f5a648e1da0c5a5fd17f1bb4
This change set is intermediate. The next one will remove all repetitive
plane_bsize calculations, because it will be passed as argument to
foreach_transformed_block_visitor.
Change-Id: Ifc12e0b330e017c6851a28746b3a5460b9bf7f0b
The intent was to initialize the deltas for the
segment to the computed value, irrespective of mode
and reference frame if (mode_ref_delta_enabled == 0).
(In response to bug posted by Manjit Hota to codec-devel
and webm-discuss lists)
Change-Id: I10435cb63d0f88359bb4c14f22181878a1988e72
Return the distortion value in vp9_rd_pick_intra_mode_sb as sum of
dist_y and dist_uv. Remove the right shift operation on dist_uv,
and make it consistent with that of vp9_rd_pick_inter_mode_sb.
Change-Id: I9d564e242d9add38e32595d33b0e0dddb1d55e5b
When the frame size changes the last frame segment map must
be resized to match and initialized to 0.
Change-Id: Idc10de109f55dbe9af3a6caae355a2974712243d
This commit makes the rate-distortion optimization search of chroma
components consistent across all block sizes. It removes redundant
codes.
Change-Id: I7e76f54d045e8efdd41d84a164c71f55b484471b
VP9_COMMON is the right place to segmentatation struct because it has
global segmentation parameters, not something specific to macroblock
processing.
Change-Id: Ib9ada0c06c253996eb3b5f6cccf6a323fbbba708
Adds a speed feature to disable split partition search based on a
given threshold on the source variance. A tighter threshold derived
from the threshold provided is used to also disable horizontal and
vertical partitions.
Results on derfraw300:
threshold = 16, psnr = -0.057%, speedup ~1% (football)
threshold = 32, psnr = -0.150%, speedup ~4-5% (football)
threshold = 64, psnr = -0.570%, speedup ~10-12% (football)
Results on stdhdraw250:
threshold = 32, psnr = -0.18%, speedup is somewhat more than derf
because of a larger number of smoother blocks at higher resolution.
Based on these results, a threshold of 32 is chosen for speed 1,
and a threshold of 64 is chosen for speeds 2 and above.
Change-Id: If08912fb6c67fd4242d12a0d094783a99f52f6c6
This commit unifies the rate-distortion cost calculation process of
luma and chroma components. It allows early termination to be enabled
later in the rd search loop of chroma components, in consistent with
luma pixels.
Change-Id: I2e52a7c6496176bf2a5e3ef338d34ceb8aad9b3d
Making foreach_transformed_block_in_plane more clear (it's not finished
yet). Using explicit tx_size variable consistently instead of
(ss_txfrm_size / 2) or (ss_txfrm_size >> 1) expression.
Change-Id: I1b9bba2c0a9f817fca72c88324bbe6004766fb7d
The macro block mode info context originally contained an
entry for each 16x16 macroblock. In VP9 each entry refers
to an 8x8 region not a macro block, so the naming is misleading.
This first stage clean up changes the names of 3 entries in the
structure to remove the mb_ prefix.
TODO clean up the nomenclature more widely in respect of
mbmi and bmi.
Change-Id: Ia7305c6d0cb805dfe8cdc98dad21338f502e49c6
Don't do vertical or horizontal splits if subsize < min_partition_size,
except for edge blocks where it makes sense.
Change-Id: I479aa66ba1838d227b5de8312d46be184a8d6401
Enable SSE2 implementation of high precision 32x32 forward DCT. The
intermediate stacks are of 32-bits. The run-time goes down from
32126 cycles to 13442 cycles.
Change-Id: Ib5ccafe3176c65bd6f2dbdef790bd47bbc880e56
Adding function build_inter_predictors_for_planes to build inter
predictors for specified planes. This function allows to remove
condition "#if CONFIG_ALPHA" and use MAX_MB_PLANE for general case.
Renaming 'which_mv' local var to 'ref', and 'weight' argument to 'ref'.
Change-Id: I1a97160c9263006929d38953f266bc68e9c56c7d
Invert loops to operate vertically in the inner loop. This allows
removing redundant loads.
Also add preloading of data.
Change-Id: I4fa85c0ab1735bcb1dd6ea58937efac949172bdc
Each iteration of the horizontal loop reuses 7 of the 11 source
values. Loading only the 4 new values saves some time.
Also add preload for source data.
Overall 4% faster on Chromebook.
Change-Id: I8f69e749f2b7f79e9734620dcee51dbfcd716b44
'skippable' can remain unset and negatively affect later decisions
address one aspect of issue #599
Change-Id: Iffdf0ac2e49ac481c27dc27c87fa546d4167bb28
Loop filter configuration doesn't belong to macroblock, so moving it from
MACROBLOCKD to VP9_COMMON. Also moving the declaration of loopfilter struct
from vp9_blockd.h to vp9_loopfilter.h.
Change-Id: I4b3e34be9623b47cda35f9b1f9951f8c5b1d5d28
The memset sets 16 bytes rather than the correct size of the
final array dimension (MAX_MODE_LF_DELTAS).
(In response to bug posted by Manjit Hota to codec-devel
and webm-discuss lists)
Change-Id: I8980f5aa71ddc9d7ef57c5b4700bc28ddf8651c7
Adds a new subpel motion estimation function that uses a 2-level
tree-structured decision tree to eliminate redundant computations.
It searches fewer points than iterative search (which can search
the same point multiple times) but has the same quality roughly.
This is made the default setting at speeds 0 and 1, while at
speed 2 and above only a 1-level search is used.
Also includes various cleanups for consistency and redundancy removal.
Results:
derf: +0.012% psnr
stdhd: +0.09% psnr
Speedup of about 2-3%
Change-Id: Iedde4866f5475586dea0f0ba4cb7428fba24eee9
Different partitionings were not being evaluated against
best_rd and there were unnecessary calls to RDCOST. This
could have resulted in a non-optimal partioning being
selected.
I simplified the variables used to track the rate,
distortion and RD values throughout the function.
Change-Id: Ifa7085ee80d824e86791432a5bc6d8fea5a3e313
Using block width and block height instead of their logarithms. Using
SUBPEL_BITS and SUBPEL_SHIFTS constants instead of magic numbers.
Change-Id: I4e10e93c907c8a5e1cb27dfe74d1fcdcc4995448
The low precision 32x32 fdct has all the intermediate steps within
16-bit depth, hence allowing faster SSE2 implementation, at the
expense of larger round-trip error. It was used in the rate-distortion
optimization search loop only.
Using the low precision version, in replace of the high precision one,
affects the compression performance by about 0.7% (derf, stdhd) at
speed 0. For speed 1, it makes derf set down by only 0.017%.
Change-Id: I4e7d18fac5bea5317b91c8e7dabae143bc6b5c8b
Removing the old one bsize_from_dim_lookup. Now we have a way to determine
block size for plane using its subsampling values (ss_size_lookup). And
then we can find the number of pixels in the block (num_pels_log2_lookup).
Change-Id: I6fc981da2ae093de81741d3d78eaefed11015db9
Removes some unused code and speed features, and organizes the
interfaces for fractional mv step functions for use in new speed
features to come.
In the process a new speed feature - number of iterations per
step during the subpel search - is exposed.
No change when this parameter is set as the original value of 3.
Results:
subpel_iters_per_step = 3: baseline
subpel_iters_per_step = 2: psnr -0.067%, 1% speedup
subpel_iters_per_step = 1: psnr -0.331%, 3-4% speedup
Change-Id: I2eba8a21f6461be8caf56af04a5337257a5693a8
Functions scale_mv_q4 and scale_mv_q3_to_q4 were almost identical except
q3->q4 conversion in scale_mv_q3_to_q4. Now q3->q4 conversion happens
directly in vp9_build_inter_predictor.
Also adding useful constants: SUBPEL_BITS and SUBPEL_MASK.
Change-Id: Ia0a6ad2ac07c45fdf95a5139ece6286c035e9639
Enable use_x86inc as a commandline option. Fix Bug with sse2 when
x86inc is disabled. Adds Sad asm protection to x86inc protection
Change-Id: Iee0f9dd235ea10e8ace512eb362ba9bebe8c9df6
Adds a few pattern searches to achieve various tradeoffs
between motion estimation complexity and performance.
The search framework is unified across these searches so that a
common pattern search function is used for all. Besides it will
be easier to experiment with various patterns or combinations
thereof at different scales in the future.
The new pattern search is multi-scale and is capable of using
different patterns at different scales.
The new hex search uses 8 points at the smallest scale
and 6 points at other scales.
Two other pattern searches - big-diamond and square are
also added. Big diamond uses 4 points at the smallest scale and
8 points in diamond shape at the larger scales.
Square is very similar conceptually to the default n-step search
but is somewhat faster since it keeps only one survivor across
all scales.
Psnr/speed-up results on derf300:
hex: -1.6% psnr%, 6-8% speed-up
big-diamond: -0.96% psnr, 4-5% speedup
square: -0.93% psnr, 4-5% speedup
Change-Id: I02a7ef5193f762601e0994e2c99399a3535a43d2
There was no benefit having this function. For example, inside
read_switchable_filter_type switchable filter context was calculated twice.
Change-Id: I79cd5bf95cbc0f6d8bf91a2e32289e01b18dcff1
Converting arguments of two functions (clamp_mv_ref, lower_mv_precision)
from int_mv* to MV*. Rewriting is_inside function to make it much shorter.
Change-Id: Ie4c4cf3eccd46707c7df099ec21fb1b61c72fc7a
This is in preparation for the SSE2 version of the high-precision
32x32 forward DCT which will share a lot of code with the existing
low precision version used for rate-distortion search.
Change-Id: I7084b6bdfb480b1fabb8493fb14e3f7fcc7888c0
Support enabling it or disabling it. Moved read out to configure.sh
so that its done once instead of in make and in config.
Change-Id: I73a9190cf31de9f03e8a577f478fa522f8c01c8b
Adds a speed feature to skip all intra modes other than
DC_PRED if the source variance is small. This feature is
made part of speed 1 and up.
Results on derf300: psnr -0.07%, speedup about 1-2%
Also uses the source variance to fine-tune the early
termination criteria when FLAG_EARLY_TERMINATE is on.
This feature is made part of speed 2 and up.
Results on derf300: psnr -0.52%, speedup about 5-7%
Change-Id: I59e38aa836557cfa5405ae706fc64815cbfe4232
Currently the only threaded option for vp9 decode. Enabled when the
decoder config thread count is > 1.
Change-Id: I082959abac9e31aa4a38ed9fd68b94680e57f4df
This changeset allows to remove vp9_switchable_interp and
vp9_switchable_interp_map arrays and make code much clear. Actually we
still have to use these mapping but only inside read_interp_filter_type and
write_interp_filter_type functions.
Change-Id: I4026c6f8c4acefba6c81421b7bacbaa52cc45f50
Chromium does not support 32bit builds for Mac which use x86inc.asm.
Make the files which include it work if 64bit or not PIC enabled
starting with vp9_copy_sse2.asm
Consolidate these targets in vp9_rtcd_defs.sh
Change-Id: If18f0b957a611efd085a3ee7d245cf1eb91e8248
This is an attempt at rewriting vp9_find_mv_refs_idx. I believe that it gains
about 1-2% decode speed
Change-Id: Ia5359c94ce9bb43b32652890e605e9a385485c1b
Loading to single lanes in multiple registers is expensive since
it requires a read and write of each register which saturates
the register file access. Loading to single registers followed
by a separate transpose reduces this pressure.
Change-Id: I4cc35887ddbca80e5e635b50d2b1d158de9668ee
Removing assign_and_clamp_mv function, making implementation of clamp_mv
and clamp_mv2 more clear and consistent.
Change-Id: Iecd08e1c1bf0379f8314ebe01811f8253f4ade58
Adds a function to compute source variance for various
sb_types to be used for pruning mode and partition searches.
[The existing activity measure function is currently specialized
for only 16x16 MBs and needs to be updated].
Change-Id: I22a41e6f1430184201487326fdbebb9b47e6fc24
The inverse 32x32 transform detects all zero entries and skips the
computations accordingly per 8 rows in the first 1-D operation. The
function vp9_short_idct10_32x32_add performs differently and is not
used anywhere, hence removed.
Change-Id: Ic4fad422debbde7b6b6ffed47c69fbd4268a906c
If the partition is out of partition size range, we don't
need to process small partition information.
Change-Id: Ice9bfbbdebe1f2ef79271a3aee17de0ed4608376
use_min_partition_size and use_max_partition_size are not used
currently, and could be added back if needed later.
Change-Id: Ib22a9c06b064567a7c1d6d5445567ed77e0d3acc
The function name rd_pick_intra4x4mby_modes is confusing, so
I changed it to rd_pick_intra_sub_8x8_y_modes to better
reflect what the function does. Also added const qualifiers
to some of the input parameters and removed camel-case.
Change-Id: I23d53d4c7af5d79ed8a471acd59a09bbb47add39
This commit exploits the sparsity of quantized coefficient matrix.
It detects each 32x8 array and skip the corresponding inverse
transformation if all entries are zero.
For ped1080p at 8000 kbps, this on average reduces the runtime of
32x32 inverse 2D-DCT SSE2 function from 6256 cycles -> 5200
cycles. It makes the overall encoding process about 2% faster at
speed 0. The speed-up is more pronounceable for the decoding process.
Change-Id: If20056c3566bd117642a76f8884c83e8bc8efbcf
This commit removes redundant arguments passing in the function of
rd_pick_reference_frame. This resolves the clang warnings about
potential use of uninitialized values.
Change-Id: Ic68f949a9f8fcd0a583786b0c75321104ea44739
Refactor the frame buffer referencing in choose_partition and make
it consistent with other places. This means to prevent potential
issues when we extend reference frame buffer.
Change-Id: I5ff33ed5f671e1f4cc7049622212769a9b4578d9
The tokenize_b function is only called when output flag is on. Hence
removing the conditional branch on it therein.
Change-Id: Ib709f47f23f39ca05a695faf86fa3377f11f2dd0
This commit optimizes the tokenization and detokenization operational
flow for speed-up. It makes the coding process about 0.3% faster at
speed 0.
Change-Id: I28008df7482874e4b5f237f2d418ff82a249dd56
This commit makes the encoder skip the redundant tokenization process
in the rate-distortion optimization search loop, while updating the
entropy contexts accordingly. It makes the speed 0 encoding process
about 0.5% faster at no performance change.
Change-Id: I34a4155a0b5332afeb45c93a51c7f35a294d685c
This commit provides special handle on 16x16 inverse 2D-DCT, where
only DC coefficient is quantized to be non-zero value.
Change-Id: I7bf71be7fa13384fab453dc8742b5b50e77a277c
Fixes a warning on MSVS 2012 where the alignment of vp9_default_iscan_8x8
didn't match between its declaration and definition.
Change-Id: I1466a15635f4b22594d705d570b7e399bfb6cf21
This allows us to increment the position at the band-level only as
we go from one band to the next; more importantly, that allows us to
use an add instead of multiply instruction, and omit the instruction
altogether if the band doesn't change from one coef to the next, thus
being slightly faster (probably more noticeable on systems where a
multiply is expensive, like arm).
Change-Id: I4343fe35b9f9a47fa00b217bdcbf5f91ff96c381
This commit brought back the shortcut implementation of 8x8/16x16
inverse 2D-DCT. When the eob <= 10, it skips the inverse transform
operations on row 4:7/4:15 in the first round. For bus_cif at 1000
kbps, this provides about 2% speed-up at speed 0.
Change-Id: I453e2d72956467d75be4ad8c04b4482ab889d572
This commit enables a special handle for the 8x8 inverse 2D-DCT,
where only DC coefficient is quantized to be non-zero. For bus_cif
at 2000 kbps, it provides about 1% speed-up at speed 0.
Change-Id: I2523222359eec26b144cf8fd4c63a4ad63b1b011
Speed feature experiment to set an upper and lower
partition size limit based on what has been seen
in spatial neighbors.
This seems to gives quite reasonable speed gains in local
(10-15%) and when used with speed 0 the losses are small
(0.25% derf, 0.35% stdhd). However, for now I am only
enabling it on speed 1 as there may be clashes with the existing
temporal partition selection in speed 2.
Using a tighter min / max around the range derived from the
neighbors increases speed further but at the cost of a
bigger quality loss. However, I think this spatial method could
be combined with data from either the last frame or a variance
method (or both) to refine the range of minimum and maximum
partition size. I.e. consider the min and max from spatial and
temporal neighbors and the variance recommendation.
Change-Id: I1b96bf8b84368d6aad0c7aa600fe141b4f07435f
Used 3 * standard_deviation in internal threshold calculation
instead of fit curve. This actually approached the algorithm
better.
For comparison, similar tests were done:
The overall psnr loss is less than before.
1. derf set:
when static-thresh = 1, psnr loss is 0.329%;
when static-thresh = 500, psnr loss is 0.970%;
2. stdhd set:
when static-thresh = 1, psnr loss is 0.922%;
when static-thresh = 500, psnr loss is 1.307%;
Similar speedup is achieved. For example,
clip bitrate static-thresh psnr time
akiyo(cif) 500 0 48.952 5.077s(50f)
akiyo 500 500 48.866 4.169s(50f)
parkjoy(1080p) 4000 0 30.388 78.20s(30f)
parkjoy 4000 500 30.367 70.85s(30f)
sunflower(1080p) 4000 0 44.402 74.55s(30f)
sunflower 4000 500 44.414 68.69s(30f)
Change-Id: Ic78833642ce1911dbbd1cb6c899a2d7e2dfcc1f3
Now read_inter_mode_info calls read_intra_block_part (renamed from
read_intra_block_modes) or read_inter_block_part (just added).
Change-Id: I541badea6b663e0ae692ec158665efb90ed20c03
This option exists in VP8, and it was rewritten in VP9 to support
skipping on different partition levels. After prediction is done,
we can check if the residuals in the partition block will be all
quantized to 0. If this is true, the skip flag is set, and only
prediction data are needed in reconstruction. Based on DCT's energy
conservation property, the skipping check can be estimated in
spatial domain.
The prediction error is calculated and compared to a threshold.
The threshold is determined by the dequant values, and also
adjusted by partition sizes. To be precise, the DC and AC parts
for Y, U, and V planes are checked to decide skipping or not.
Test showed that
1. derf set:
when static-thresh = 1, psnr loss is 0.666%;
when static-thresh = 500, psnr loss is 1.162%;
2. stdhd set:
when static-thresh = 1, psnr loss is 1.249%;
when static-thresh = 500, psnr loss is 1.668%;
For different clips, encoding speedup range is between several
percentage and 20+% when static-thresh <= 500. For example,
clip bitrate static-thresh psnr time
akiyo(cif) 500 0 48.923 5.635s(50f)
akiyo 500 500 48.863 4.402s(50f)
parkjoy(1080p) 4000 0 30.380 77.54s(30f)
parkjoy 4000 500 30.384 69.59s(30f)
sunflower(1080p) 4000 0 44.461 85.2s(30f)
sunflower 4000 500 44.418 78.1s(30f)
Higher static-thresh values give larger speedup with larger
quality loss.
Change-Id: I857031ceb466ff314ab580ac5ec5d18542203c53
Removing unused constants, macros, and function declarations. Using
ROUND_POWER_OF_TWO macro, vp9_zero, vp9_copy where possible. Moving
#include from *.h to *.c. Merging for loops for motion vectors.
Change-Id: Ic3bf841764a2bb177128bb3a6d7aa8f68229cd13
