The restructure moves the decision into the rd pick
modes loop and makes a decision based at the 16x16
block level instead of only the 64x64 level.
This gives finer granularity and better visual results
on the clips I have tested. Metrics results are worse
than the old AQ2 especially for PSNR and this mode
now falls between AQ0 and AQ1 in terms of visual
impact and metrics results.
Further tuning of this to follow.
It should be noted that if there are multiple iterations
of the recode loop the segment for a MB could change
in each loop if the previous loop causes a change in the
complexity / variance bin of the block. Also where a block
gets a delta Q this will alter the rd multiplier for this block
in subsequent recode iterations and frames where the
segmentation is applied.
Change-Id: I20256c125daa14734c16f7cc9aefab656ab808f7
The function vp9_filter_block1d16_h8_ssse3 uses the PSHUFB instruction which has a 3 cycle latency and slows execution when done in blocks of 5 or more on Atom processors.
By replacing the PSHUFB instructions with other more efficient single cycle instructions (PUNPCKLBW + PUNPCHBW + PALIGNR) performance can be improved.
In the original code, the PSHUBF uses every byte and is consecutively copied.
This is done more efficiently by PUNPCKLBW and PUNPCHBW, using PALIGNR to concatenate the intermediate result and then shift right the next consecutive 16 bytes for the final result.
For example:
filter = 0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,8
Reg = 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15
REG1 = PUNPCKLBW Reg, Reg = 0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7
REG2 = PUNPCHBW Reg, Reg = 8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15
PALIGNR REG2, REG1, 1 = 0,1,1,2,2,3,3,4,4,5,5,6,6,7,7,8
This optimization improved the function performance by 23% and produced a 3% user level gain on 1080p content on Atom processors.
There was no observed performance impact on Core processors (expected).
Change-Id: I3cec701158993d95ed23ff04516942b5a4a461c0
the flag in the header wasn't being set based on the encoder
configuration in non-intra only mode
broken since:
fbc2fbf Adding oxcf temp variable.
Change-Id: Ib4cff9901889824bc4e68d7f0f6deb1e41df2f53
The initial reset of this_rdc in vp9_pick_inter_mode is not needed,
since it will be re-assign when used.
Change-Id: Ic0e12d741cbab292fc214c1eabb48b129af7839b
Compare the current best mode rate-distortion cost with the skip
threshold to decide if performing motion search.
Change-Id: Ia071824f8dd3b7db485f424692a485a2da6a1a9f
These speed-up features for key frame coding are only turned on
in the settings of hybrid non-RD and RD mode decision. It provides
about 20% speed-up to the hybrid key frame coding at the expense
of certain compression performance loss. For vidyo1, the key frame
coding statistics are changed
9838F, 35.020 dB, 61677 us -> 9920F, 34.834 dB, 47556 us
Overall rtc set compression performance is down by -0.257%.
Change-Id: I0025447fda26bb7855e982955642b5f55d71b51f
When block size is below 16x16, the encoder swap from non-RD to
RD mode for key frame coding. This largely brough back the key
frame compression performance. For vidyo1 at 1000 kbps, the key
frame coding statistics are changed
9978F, 34.183 dB, 36807 us -> 9838F, 35.020 dB, 61677 us
As compared to the full RD case
7187F, 34.930 dB, 214470 us
The overall rtc set coding performance (single key frame setting)
is improved by 1.5%.
Change-Id: I78a4ecf025d7b24ec911e85be94e01da05e77878
Change 72193 made the encoder behave differently
when configured with and without high bitdepth.
This change means the same algorithm is used for both.
Change-Id: I707a44a94afca773a9e0c2f7ebeeea83030257c5
No more checking of corrupted reference frame as we skip
decoding any non-intra frame in case of frame corrupted.
Change-Id: I77d41bbb02fc5f61972740e2d411441eb6a17073
Currently, VP9 supports column-tile encoding, which allows a frame
to be encoded in multiple column tiles independently. The number of
column tiles are set by encoder option "--tile-columns". This
provides a way to encode a frame in parallel.
Based on previous set of patches, this patch implemented the tile-
based multi-threaded encoder. Each thread processes one or more
tiles.
Usage:
For HD clips:
--tile-columns=2 --threads=1/2/3/4
While using 4 threads, tests showed that the encoder achieved
2.3X - 2.5X speedup at good-quality speed 3, and 2X speedup at
realtime speed 5.
Change-Id: Ied987f8f2618b1283a8643ad255e88341733c9d4
Change 71789 renamed CONFIG_VP9_HIGH to CONFIG_VP9_HIGHBITDEPTH.
However, one use of CONFIG_VP9_HIGH was missed.
Change-Id: I0ebb9c71380c6d810a25708d15471abf9533e695
For key frame at speed 6: enable the non-rd mode selection in speed setting
and use the (non-rd) variance_based partition.
Adjust some logic/thresholds in variance partition selection for key frame only (no change to delta frames),
mainly to bias to selecting smaller prediction blocks, and also set max tx size of 16x16.
Loss in key frame quality (~0.6-0.7dB) compared to rd coding,
but speeds up key frame encoding by at least 6x.
Average PNSR/SSIM metrics over RTC clips go down by ~1-2% for speed 6.
Change-Id: Ie4845e0127e876337b9c105aa37e93b286193405
This commit reworks the ONE_LOOP_REDUCED coefficient probability
model update process. It allows model update for every coefficient
across the spectrum at a coarser resolution, instead of performing
precise update only for certain subset of probability models.
The overall runtime remains nearly same (<1% change) for speed -6.
The compression performance is improved by 7.5% in PSNR for speed
-5 and 4.57% for speed -6, respectively.
Change-Id: Ifb17136382ee7e39a9f34ff4a4f09a753125c8d1
Synchronize all threads immediately as a subsequent decode call may
cause a resize invalidating some allocations.
fixes one aspect of crbug.com/437655
Change-Id: Ie993b62c2756478543206ddbe43ec6268d90a470
Change 72056 unfolded some macro definitions,
but lost some alternative behaviour required for
high bitdepth encodes.
This causes the encoder to crash, see issue 884.
Change-Id: I8ce4d73c9fe0a3c10ccb86fba210fabc8b2f0ccc
Also removes some spurious changes in common/vp9_blockd.h which
was introduced by a rebase issue between nextgen and master branches.
Change-Id: If359f0e9a71bca9c2ba685a87a355873536bb282
(cherry picked from commit 005d80cd05)
(cherry picked from commit 08d2f54800)
(cherry picked from commit 4230c2306c)
AQ2 modified to use mb_av_energy in defining variance
thresholds used alongside complexity when defining the
segment to be used for an SB64.
Slight improvements in metrics (ssim and PSNR).
Change-Id: Idb9cb73f7d9c4f7118cd7e84ac77b0f25cacbf81
Incorporate segment delta-q into estimated bits.
This generally improves the rate control under cyclic refresh (aq=3) mode.
Change-Id: I1dc60fb230e7d08357fae18909d8ed27bf58e037
This patch greatly increase the strength of AQ1.
Visual tests show strong gains on many clips but their is a big
hit on psnr.
SSIM is more mixed with some winners and losers.
Change-Id: Idaa5d3b41d8576096bfa000b62bc531c3d8bf6a1
Each tile's tok starting address is calculated before the encoding
process. These addresses are stored so that the same calculation
won't be done again in packing bit stream.
Change-Id: I0a3be0301f002260c19a850303f2f73ebc47aa50
When the golden frame is boosted, the rate correction factor is not
correlated well with other inter frames even in CBR mode. This commit
changes to use GF specific rate_correction_factor when gf_cbr_boost
is greater than 20%.
Change-Id: I6312c1564387bcacc11f4c5e8a9cfdc781b5c3ab
This commit allows the encoder to increase the mode test kick-off
thresholds if the previous best mode renders all zero quantized
coefficients, thereby saving motion search runs when possible.
The compression performance of speed -5 and -6 is down by -0.446%
and 0.591%, respectively. The runtime of speed -6 is improved by
10% for many test clips.
vidyo1, 1000 kbps
16578 b/f, 40.316 dB, 7873 ms -> 16575 b/f, 40.262 dB, 7126 ms
nik720p, 1000 kbps
33311 b/f, 38.651 dB, 7263 ms -> 33304 b/f, 38.629 dB, 6865 ms
dark720p, 1000 kbps
33331 b/f, 39.718 dB, 13596 ms -> 33324 b/f, 39.651 dB, 12000 ms
mmoving, 1000 kbps
33263 b/f, 40.983 dB, 7566 ms -> 33259 b/f, 40.978 dB, 7531 ms
Change-Id: I7591617ff113e91125ec32c9b853e257fbc41d90
This patch modified struct VP9_COMP. Created a struct ThreadData
to include data that need to be copied for each thread. In
multiple thread case, one thread processes one tile. all threads
share one copy of VP9_COMP,
(refer to VP9_COMP *cpi in the code)
but each thread has its own copy of ThreadData,
(refer to ThreadData *td in the code).
Therefore, within the scope of encode_tiles(), both cpi and td
need to be passed as function parameters.
In single thread case, the FRAME_COUNTS pointer in ThreadData
points to "counts" in VP9_COMMON.
Change-Id: Ib37908b2d8e2c0f4f9c18f38017df5ce60e8b13e
The intra mode penalty is covered by intra_cost_penalty. This
commit removes the other intra cost threshold, provided that the
constant 50 is negligible in normal rate-distortion cost.
Change-Id: I9b8b7483c43b9a41741622e7057def1f7d51bb72
This change is made in preparation for a
subsequent patch which adds acceleration
for the highbitdepth transform functions.
The highbitdepth transform functions attempt
to use 16/32bit sse instructions where possible,
but fallback to using the C implementations if
potential overflow is detected. For this reason
the dct routines are made global so they can be
called from the acceleration functions in the
subsequent patch.
Change-Id: Ia921f191bf6936ccba4f13e8461624b120c1f665
(cherry picked from commit 454342d4e7)
This commit makes a non-RD coding mode decision process for key
frame coding. It can be optionally turned on in speed -6 and above.
Change-Id: I0847258b392877a0210b4768bef88ebc9ad009b5
This commit allows more aggressive decision to skip forward
transform and quantization for luma component in RTC coding mode.
The chroma components remains going through the normal coding
routine, since they are not included in the non-RD mode search
process.
It reduces the runtime cost by 2% - 10%. In speed -6,
vidyo1 1000 kbps
16576 b/f, 40.281 dB, 8402 ms -> 16576 b/f, 40.323 dB, 7764 ms
nik720p 1000 kbps
33337 b/f, 38.622 dB, 7473 ms -> 33299 b/f, 38.660 dB, 7314 ms
dark720p 1000 kbps
33330 b/f, 39.785 dB, 13505 ms -> 33325 b/f, 39.714 dB, 13105 ms
The compression performance of speed -6 is improved by 0.44% in
PSNR and 1.31% in SSIM.
Change-Id: Iae9e3738de6255babea734e5897f29118bebc6d7
In AQ1 a rate adjustment was applied for blocks coded with a
deltaq. This tends to skew the partition selection and cause
rate overshoot.
For example, consider a 64x64 super block where some but not all
sub blocks are in a low q segment and some are in a high q segment.
The choice of Q when considering large partition and transform sizes
is defined by the lowest sub block segment id (currently this implies the
lowest Q). If some parts of the larger partition are very hard this will
cause a high rate component.
The correct behavior here is for the rd code to discard the large partition
choice and break down to sub blocks where some have low and some
have high Q. However the rate correction factor above mask the high
cost of coding at a larger partition size.
Change-Id: Ie077edd0b1b43c094898f481df772ea280b35960
Make the midpoint variance used in AQ mode 1 segmentation
depend on the overall complexity of the frame in two pass.
Change-Id: I452814ec57f7a32352e41bb250e78066abe952dd
Add an additional restriction to bit/complexity based
segmentation based on spatial variance.
Only lower Q when both the number of bits spent
in the initial encoding pass and the spatial complexity are
below a threshold. This will prevent the low Q segments
being used just because there is a surfeit of bits.
Small metrics gains especially opsnr.
derf ~0.2% std-hd ~0.3%
Change-Id: I6a8496d466d673f9b0e2b2ca6304ea7b6d8e1cce
This is the first of a series of patches to restructure and
improve AQ mode 1 (variance based AQ).
Change-Id: Idcf693131a3ea2459dcfd957a54a65b971fa4a2a
Similar to mask_filter, the filter_cache in RD_OPT struct can be
moved out, and declared as a local variable since it is only
used in pick_inter_mode functions.
Change-Id: I412b99cca82bade07ac912064ec03dd1de6b2c17
Correct calculation of number of mbs in two pass code when
frame resizing is enabled. Always use initial number of mbs if
scaling is enabled, as this is what was used in the first pass.
Change-Id: I49a4280ab5a8b1000efcc157a449a081cbb6d410
The mask_filter in RD_OPT struct is used to record rd result in
filter decision. It is only used in pick_inter_mode functions,
and is removed from the struct and declared as a local variable.
Change-Id: I3c95c8632ba7241591ce00ef2ef5677b5e297d7b
The max_partition_size and max_partition_size are set at the
beginning while setting speed features, and then adjusted at
SB level. Moving them to mb struct ensures there is a local
copy for each thread.
Change-Id: I7dd08dc918d9f772fcd718bbd6533e0787720ad4
VP9/DatarateTestVP9Large.ChangingDropFrameThresh/[34] fails post the
merge of commit#ffa06b37. This commit adds reset of rc tracking info
when frame is dropped, and fixes the causes of the bad interaction
between the tests and the previous commit.
Change-Id: I848acfd9fcb336359662274325190f94aac76eae
This commit reworks the forward transform and quantization process
for 8x8 block coding. It combines the two operations in a single
function to save a store/load stage of the original transform
coefficients. Overall the speed -6 is slightly faster (around 1%
range). The compression performance of speed -6 is improved by
3.4%.
Change-Id: Id6628daef123f3e4649248735ec2ad7423629387
In rare cases, the interaction between rate correction factor and Q
choices may cause severe oscillating frame sizes that are way off
target bandwidth. This commit adds tracking of rate control results
for last two frames, and use the information to prevent oscillating
Q choices.
Change-Id: I9a6d125a15652b9bcac0e1fec6d7a1aedc4ed97e
vp9_quantize_fp is the quantization process used by rtc coding
mode. This commit adds a sse2 implementation of it. The
implementation is modified based on vp9_quantize_b_sse2. No speed
difference from ssse3 version.
Change-Id: I24949c5b27df160b4f35117d28858d269454e64a
Current setting had active_worst_quality set too high (close to worst_quality)
for first frame(s) following first key frame. This changes that to be somewhat
more aggressive in allowing active_worst_quality to be lower following key frame.
Also remove the 4/5 reduction in active_worst for key frame as
this should be set by the user qp_max setting.
Change-Id: I0530b3ddcc85c00e3eb7568de1b14a31206c4a4c
The function pointer in compressor instance does not change, so this
commit changes to call the function directly.
Change-Id: I9c9c460e3475711c384b74c9842f0b4f3d037cc5
This commit adds a check condition to the prediction buffering
operation used in the rtc coding mode. This resolves a unit test
warning in example/vpx_tsvc_encoder_vp9_mode_7.
Change-Id: I9fd50d5956948b73b53bd8fc5a16ee66aff61995
These 2 members in RD_OPT were moved to TileDataEnc struct
already, and therefore were removed here.
Change-Id: I22fee3b67f96e473a58e194a7edc76dbd48bfa04
Several frame counters in encoder are updated at SB level. Combine
those counters and put them in a separate struct, which allows us
to allocate one copy for each thread.
Change-Id: I00366296a13c0ada4d8fa12f5e07728388b6cab7
Modified VP9_COMP struct to include MACROBLOCK *mb. This change
makes it feasible in multi-thread case to allocate a mb for each
thread.
Change-Id: I624d6d1aa9c132362200753e5d90b581b1738d6e
A flush bug is discovered during putting frame parallel decoder
into Android. This test will expose that bug.
Change-Id: Ia047f27972f4da0471649f79f1f91e7695297473
Two members in struct CYCLIC_REFRESH
int64_t projected_rate_sb;
int64_t projected_dist_sb;
are updated at the superblock level, which makes them shared data
in the multi-thread situation, and requires extra work to handle
them. However, those values are updated and used immediately, and
therefore can be removed. This patch cleaned up the code and
removed the two members.
Change-Id: I2c6ee4552bf49fb63ce590cdb47f9723974fffb1
Prepare for the introduction of frame-size change
logic into the recode loop.
Separated the speed dependent features into
separate static and dynamic parts, the latter being
those features that are dependent on the frame size.
Change-Id: Ia693e28c5cf069a1a7bf12e49ecf83e440e1d313
Add extra vp9_clear_system_state() calls to fix
double / mmx issue introduced into first pass
code for 32 bit builds.
Change-Id: I84cd2986b80d83650a091ab25c43755efeb82e03
Rate correction factor is used to correct the estimated rate for any
given quantizer, and feeds into rate control for quantizer selection.
We make use of the actual bits used to calculate this rate correction
factor with an adjustment limit to prevent over-adjustment.
This commit adapts the adjustment limit to the difference between the
estimated bits and the actual bits, allows the adjustment limit to vary
between 0.125 (when estimate is close to actual) and 0.625 (when there
is >10X factor off between estimated and actual bits). By doing this,
the commit appears to have largely corrected two observed issues:
1. Adjustment is too slow when the actual bits used is way off from
estimate due to the small adjustment limit.
2. Extreme oscillating quantizer choices due to the feedback loop.
Change-Id: I4ee148d2c9d26d173b6c48011313ddb07ce2d7d6
This commit makes the speed -6 and above use the reconstructed
boundary pixels for precise intra prediction. This allows more
intra prediction modes to be tested in the non-RD coding process.
Enabling horizontal and vertical intra prediction modes can
improve the speed -6 compression performance for rtc set
by 0.331%.
Change-Id: I3a99f9d12c6af54de2bdbf28c76eab8e0905f744
I0c5f010 changed to allow update golden reference buffer in CBR mode,
this commit changes the use of rate_correction_factor for those frames
to be aligned with the new usage. This commit attempts to solve two
issues:
a. Initialization of rate correction factor for Golden Frame
Prior to this patch, even the regular inter frame has been update
the rate correction factor based on content and encoding results,
the first golden frame would still use the ininitialized value
that can be way off.
b. Allowing rate correction factor update to be slightly faster
Prior to this patch, when the rate correction factor is off, the
update to the factor is too slow, the factor could not get close
to a semi-correct value even after many frames.
The commit helps all clips in psnr/ssim metric, but especially to
a few clip in RTC set that rate correction was way off. For example
thaloundeskmtgvga gained about .5dB for both overall/average psnr.
Change-Id: I0be5c41691be57891d824505348b64be87fa3545
Adds support for one-pass rc-enabled SVC encoder with callbacks for
getting per-layer packets.
- the callback function registration is implemented as an encoder
control function.
- if the callback function is not registered, the old way of
aggregating packets with superframe will take effect.
- one more control function “VP9E_GET_SVC_LAYER_ID” has been
implemented to get the temporal/spatial id from the encoder
within the callback. This can be used to get the ids to put on RTP
packet.
Change-Id: I1a90e00135dde65da128b758e6c00b57299a111a
This commit rename a reserved color space entry to BT_2020, it intends
to provide support for VP9 bitstream to pass along the color space
type defined in BT.2020(Rec.2020)
please note this entry does not have any effect on encoding/decoding
behavior, but allow applications to the pass the information along
from encoding end to decoding end.
Change-Id: I4678520e89141ea5e8900f7bd1c0e95b710b7091
This commit integrates the non-RD mode decision process and the
encoding process into a single recursion scheme.
Change-Id: I6a7e72a0b84d567554801ebbe01ec75d54c1f77d
This patch was to fix the vpxdec fuzzing3 test failure. When an
error occurs, setjmp() is invoked, which calls the decoder
removing routine. In multiple thread situation, other threads
could try to access the frame context memory that is already
deallocated, thus causing a segfault.
An invalid unit test was added for this issue.
Change-Id: Ida7442154f3d89759483f0f4fe0324041fffb952
The aim of this patch is to apply a positive weighting to
frames that have a significant number of blocks that are
of low spatial complexity and are dark. The rationale behind
this is that artifacts tend to be more visible in such frames.
In this patch the weight is only applied in regard to the distribution
of bits between frames. Hence if all the frames share similar
characteristics (as is the case for most of our short test clips) there
will be little or no net effect.
However, the effect can be seen on some longer form test content.
For example Tears of steel baseline test:
2323.09 Kbit/s opsnr 39.915 ssim 74.729
With this patch:-
2213.34 Kbit/s opsnr 39.963 ssim 74.808
(Sligtly better metrics and about 5% smaller)
The weighting may well need some further tuning along side changes
to the aq modes.
Change-Id: Ieced379bca03938166ab87b2b97f55d94948904c
This commit removes the cyclic aq mode dependency on
in_static_area and reworks the corresponding cut-off thresholds.
It improves the compression performance of speed -5 by 1.47% in
PSNR and 2.07% in SSIM, and the compression performance of speed
-6 by 3.10% in PSNR and 5.25% in SSIM. Speed wise, about 1% faster
in both settings at high bit-rates.
Change-Id: I1ffc775afdc047964448d9dff5751491ba4ff4a9
This will save the memory and improve the decode speed due to
removing unnecessary memset of big prev_mi array for
all the key frames.
Decoding a all key frames 1080p video shows speed improve around 2%.
Change-Id: I6284a445c1291056e3c15135c3c20d502f791c10
This commit makes the RTC coding mode to conditionally skip the
reference frame mode search, when the predicted motion vector of
the current reference frame gives more than two times sum of
absolute difference compared to that of other reference frames.
It reduces the runtim by 1% - 4% for speed -5 and -6. The average
compression performance is improved by about 0.1% in both settings.
It is of particular benefit to light change scenarios. The
compression performance of test clip mmmovingvga.y4m is improved by
6.39% and 15.69% at high bit rates for speed -5 and -6, respectively.
Speed -5
vidyo1 16555 b/f, 40.818 dB, 12422 ms ->
16552 b/f, 40.804 dB, 12100 ms
nik 33211 b/f, 39.138 dB, 11341 ms ->
33228 b/f, 39.139 dB, 11023 ms
mmmoving 33263 b/f, 40.935 dB, 13508 ms ->
33256 b/f, 41.068 dB, 12861 ms
Speed -6
vidyo1 16541 b/f, 40.227 dB, 8437 ms ->
16540 b/f, 40.220 dB, 8216 ms
nik 33272 b/f, 38.399 dB, 7610 ms ->
33267 b/f, 38.414 dB, 7490 ms
mmmoving 33255 b/f, 40.555 dB, 7523 ms ->
33257 b/f, 40.975 dB, 7493 ms
Change-Id: Id2aef76ef74a3cba5e9a82a83b792144948c6a91
This commit unfolds the legacy macro definitions used in the
sub-pixel motion search and refactors the operational flow for
later optimizations.
Change-Id: I3e3f770cad961d03d1a6eb0b2a0186cc77eaf2b8
The current logic was allowing for disabling golden refresh only
for two pass svc encoding. This change disables it as long as
more than 1 layer encoding is used (for example temporal layers under 1pass CBR).
Change-Id: I4dc5204a7ad365c821ec7963e93b59da82e1826b
In the function mb_lpf_horizontal_edge_w_avx2_16 the usage of the intrinsic
_mm256_cvtepu8_epi16 cause a compiler bug in gcc 4.9.1.
until it will be fixed I created a workaround that create the up convert by
using broadcast128+shuffle.
The bug was reported here:
https://code.google.com/p/webm/issues/detail?id=867
Change-Id: I73452e6806f42e0fadcde96b804ea3afa7eeb351
A recent change has introduced big quality drops for speed 7 and 12
for --rt mode. The change reverted the big drop and improved quality
by 9.5% for speed 7 and 13.4% for speed 12.
Change-Id: I07b82e3bb6002a73af486a083458c88877bdad01
This will save a lot of memory for decoder due to removing of prev_mi,
but prev_mi is still needed in encoder. So this will increase a little bit
memory for encoder.
Change-Id: I24b2f1a423ebffa55a9bd2fcee1077dac995b2ed
This commit makes the inter prediction buffer system to support
hybrid partition search. It reduces the runtime of speed -5 by
about 3%. No compression performance change.
vidyo1 720p 1000 kbps
11831 ms -> 11497 ms
nik 720p 1000 kbps
10919 ms -> 10645 ms
Change-Id: I5b2da747c6395c253cd074d3907f5402e1840c36
Combined vp9_denoiser_8xM_sse2 and vp9_denoiser_4xM_sse2 into one
function vp9_denoiser_NxM_sse2_small and passed the bitexact testing.
Changed the name of the function vp9_denoiser_64_32_16xM_sse2 to
vp9_denoiser_NxM_sse2_big.
Change-Id: Ib22478df585994dd347ebae04202c0b701e7f451
This commit changes to allow the usage of golden reference frame in
VP9 CBR mode to improve quality. VP9 supports potentially up to 8
reference buffers, it has reference buffers available for this
purpose. This was not possible in VP8 as golden and alt-ref buffers
were used for temporal scalability purpose in CBR mode in WebRTC.
For frames that update golden frame, there can be a quality boost.
The amount of allowed bitrate boost can be controlled via parameter
rc_max_inter_bitrate_pct. The inital value of the boost ratior is
currently based on over_shoot_pct. Further experiments will work
out the adaption of this boost value.
Change-Id: I0c5f010c8fd8b7b598f69779c1b30e5b2ac30a4d
Added code to relax the active maximum Q in response
to extreme local overshoot to reduce bandwidth peaks.
The impact is small in metrics terms, but it this helps reduce
bandwidth spikes and overall overshoot in a number of
clips in our tests sets (especially the YT test set).
In particular this should help prevent very big spikes where a clip
is mainly easy but has a short hard section. In such a case a choice
of maximum Q for the clip as a whole may allow us to hit the overall
target rate but give some extreme spikes. The chunked encoding in YT
mitigates this problem but it can show up where a longer clip is
coded as a single chunk.
Change-Id: I213d09950ccb8489d10adf00fda1e53235b39203
The zero motion vector was effectively used in the subsampled pixel
based variance calculation. This commit makes it directly use zero
mv to generate prediction.
Change-Id: Ica83dc843e9f8da2f89c3ef451e50f16214c0def
0 means that golden boost is off, and uses average frame target rate,
a non-zero number means the percentage of boost over average frame
bitrate is given initially to golden frames in CBR mode.
Change-Id: If4334fe2cc424b65ae0cce27f71b5561bf1e577d
The point at which frames are scaled to their
coded dimensions is moved into the re-code loop.
This is in preparation for a further patch that
will add logic into the re-code loop to reduce
the coded frame size if the encoder is struggling
to hit the target data rate at the native frame
size.
Change-Id: Ie4131f5ec6fb93148879f6ce96123296442bf2d1
Add second level arf Q adjustment when using dual arfs
in constant Q mode.
Previously in constant Q mode enabling dual arf hurt by ~5%
but with this change the average benefit is ~1-1.5% with some
mid range data points up ~10%.
Note however that it still hurts on some clips including
some very low motion show content.
Change-Id: I5b7789a2f42a6127d9e801cc010c20a7113bdd9b
This patch allocated frame contexts outside VP9_COMMON. This allows
multiple threads to share the same copy of frame contexts, and
reduces the overhead. It also guarantees the correct update of
these contexts during bitstream packing. This patch doesn't change
encoding result.
Change-Id: Ic181a2460b891d1d587278a6d02d8057b9dbd353
The initialization of this_mode_pred does not work when the ref_frame
loop ever goes beyond LAST_FRAME. This commit fixes the subtle issue
and allows potentially expanding the loop to test GOLDEN_FRAME.
Change-Id: Ibbd427a22160d1d9eacb8ed0c87f88d6cef9c0f3
Using 4 threads, frame parallel decode is ~3x faster than single thread
decode and around 30% faster than tile parallel decode for frame parallel
encoded video on both Android and desktop with 4 threads. Decode speed is
scalable to threads too which means decode could be even faster with more threads.
Change-Id: Ia0a549aaa3e83b5a17b31d8299aa496ea4f21e3e
This commit refactors the rate distortion structure used in the
non-RD coding mode and saves a few RDCOST calculations.
Change-Id: I62c3416c300d2c5372f21b96d93a6b633a34ab3a
The existing speed features produce horrible encoding results, almost
30% worse than cpu-used=4, this commit adjust the speed features to
produce relatively resonable results to be within 3%-5% of cpu-used=4.
Change-Id: I0ca6ebafb33024d4a0cbcf04c78a4a00b8dd1ecf
Its functionality has been replaced with choose_partitioning and
threshold based control on split mode check.
Change-Id: Ic9bb321df06b524f5c38ea5874dc6f6a8f93c5e3
This speed feature has been deprecated in both yt and rtc coding
modes. This commit removes the related operations.
Change-Id: I079c79c6adafe45581af2ebf8b98faebcface1ce
This commit re-designs the recursive partition search scheme in
rtc speed -5. It first checks if the current block is under cyclic
refresh mode. If so, apply recursive partition search. Otherwise,
perform sub-sampled pixel based partition selection. When the
pre-selection finds the partition size should be 32x32 or above,
use the partition size directly. Otherwise, apply partition search
at nearby levels around the preset partition size.
It is enabled in speed -5. The compression performance of rtc
speed -5 is improved by 9.4%. Speed wise, the run-time goes slower
from 1% to 10%.
nik_720p, 1000 kbps
33220 b/f, 38.977 dB, 10109 ms -> 33200 b/f, 39.119 dB, 10210 ms
vidyo1_720p, 1000 kbps
16536 b/f, 40.495 dB, 10119 ms -> 16536 b/f, 40.827 dB, 11287 ms
Change-Id: I65adba352e3adc03bae50854ddaea1b421653c6c
Extend --auto-alt-ref from parameter so we can use it to
turn multi-arf on and off from the command line.
For now the range is 0-off, 1-on, 2-multi-arf on.
Rename play_alternate to enable_auto_arf
Change-Id: Id7b64407cfbe76ba0090a83b588a03e22a240386
All sad function that process above 32 consecutive elements are optimized
for AVX2:
vp9_sad64x64
vp9_sad64x32
vp9_sad32x64
vp9_sad32x32
vp9_sad32x16
vp9_sad64x64_avg
vp9_sad64x32_avg
vp9_sad32x64_avg
vp9_sad32x32_avg
vp9_sad32x16_avg
The functions that appeared as a hotspot is vp9_sad32x32 and vp9_sad64x64
vp9_sad32x32 was optimized by 68% and vp9_sad64x64 was optimized by 90%
both of them gave and overall ~2.3% user level gain
Change-Id: Iccf86b375a2b54c5fbbe685902ead0c9a561b9fd
Currently, the tokens for a tile are stored immediately after its
preceding tile, which causes a dependency. This is unnecessary
since we always allocate enough memory for tokens. Removing
the dependency allows token writing done in parallel. This patch
doesn't change encoding result.
Change-Id: I7365a6e5e2c2833eb14377c37e1503c9d0f26543
This should be set right after decoder really start to decode frame
instead setting at the end.
Even decoder does not have a displayable frame to show and return NULL
to application, this should be set too.
Change-Id: If0313a834bc64e3b0f05a84f4459d444d9eab0d8
When early termination is triggered, properly reset the rate cost
to invalid value to avoid potential ioc issue.
Change-Id: I3444390be2e49a34bb02cf8a74c33d5dbd96d88d
Delete gfboost_qadjust() and move Q based adjustment
into calc_frame_boost(). Also remove clamping. Making
the adjustment here means that it influences not just the
boost level but also the selection of the GF/ARF interval.
This change gives a small average gain in PSNR but
larger gains in SSIM, especially for harder std-hd set (1.5%)
Change-Id: I3aa81b8feccaeff93d915e19fb9cf5cd64c86327
This commit fixes an ioc issue that will happen when the cumulative
variables are not in effective use. The fix discards these
redundant additions.
Change-Id: Idbac5bfb989c0cedc5f8a323effce938519b2457
this removes an assumption that worker->data1 would be pointing to a
TileWorkerData allocation.
additionally, within the multi-threaded loopfilter pass VP9LfSync as a
parameter to the worker hook, removing the need for a shadow pointer in
LFWorkerData.
Change-Id: Ic7b2faa34e3eb59dbcb8a7c67f333448fa047c88
move them from VP9Worker::data[12] to allow the structure to be reused a
bit more naturally by the multi-threaded loopfilter.
Change-Id: I31b49c9e93ca744fd7f6d6ed8696671188fb2c1d
This removes an unnecessary restriction that causes
a problem (noticed by AWG) when the forced key frame
interval is set to a very small value, such as 10. In this case
we were being forced to code minimal length GF groups.
Change-Id: I76ef5861a09638ff51f61fea02359554184ada53
We encode a empty invisible frame in front of the base layer frame to
avoid using prev_mi. Since there's a restriction for reference frame
scaling factor, we have to make it smaller and smaller gradually until
its size is 16x16.
Change remerged.
Change-Id: I9efab38bba7da86e056fbe8f663e711c5df38449
This reverts commit 452dc21500.
This change has introduced a significant quality regression on content
with forced key frames. (e.g. the YT and yt-hd set). It is most
noticeable in static content where the kf bits dominate. Here, despite
key frames being apparently coded at the same Q, there is a drop in all
metrics of ~20% (e.g clXR and BFa0).
Change-Id: Iba14cc61778c0846fa0a59c33c55a9fc49512cb4
Compare the estimated rate and distortion to the thresholds scaled
according to the operating block size and determine if further
split partition search will be run. The compression performance of
speed -5 is changed by -0.074%. The encoding speed is 10% - 15%
faster.
vidyo1 720p
16545 b/f, 40.492 dB, 11475 ms -> 16535 b/f, 40.486 dB, 10100 ms
nik720p
16624 b/f, 36.310 dB, 10071 ms -> 16617 b/f, 36.313 dB, 8346 ms
Change-Id: Ic9197ab5761279ae55d2fb7813b2af0e0db497b8
Reduce the intra_cost_penalty for non-rd mode,
and some updates to VAR_BASED_PARTITION.
Visual tests show some improvement at Speed 6, for RTC clips.
Change-Id: If9090daf7aed14906a32d931a538ab544bbca606
This commit replaces the use of copy_partitioning with
choose_partitioning based on the sse of subsamped pixels, which
provides significantly better coding performance and runs at
similar speed, as compared to copy_partitioning. It improves rtc
speed 5 coding performance by 3%.
Change-Id: I52d3682a12dce0147f5e52383a594fc242ca3228
We encode a empty invisible frame in front of the base layer frame to
avoid using prev_mi. Since there's a restriction for reference frame
scaling factor, we have to make it smaller and smaller gradually until
its size is 16x16.
Change-Id: I60b680314e33a60b4093cafc296465ee18169c19
Move the point at which input frames are scaled
into the recode loop. This will allow us to change
the coded frame size dynamically in response
to previous attempts to encode the frame at a
higher resolution.
A following patch will implement a scheme for
resizing the frame in the recode loop.
Change-Id: I6a59c02d6ac1626512edad6de8b60063b79433e6
This commit makes a struct that contains rate value, distortion
value, and the rate-distortion cost. The goal is to provide a
better interface for rate-distortion related operation. It is
first used in rd_pick_partition and saves a few RDCOST calculations.
Change-Id: I1a6ab7b35282d3c80195af59b6810e577544691f
Add back clamp which ensures that the Q adaptation
is turned off when the over_shoot_pct and under_shoot_pct
parameters are set to 100.
Change-Id: Id0161b114d39a3029cd3eb28020caab0c3914922
Allow min and maxQ to creep when the undershoot
or overshoot exceeds thresholds controlled by the
command line under_shoot_pct and over_shoot_pct
values.
Default is 100%,100% which ~disables adaptation.
Derf results for example undershoot% / overshoot%:-
Head:- Mean abs (%rate error) = 14.4%
This check in:-
25%/25% - Mean abs (%rate error) = 6.7%
PSNR hit -1% SSIM -0.1%
5% / 5% - Mean abs (%rate error) = 2.2%
PSNR hit -3.3% SSIM - 1.1%
Most of the remaining error and most of the quality hit is
at extreme data rates. The adaptation code still has an
exception for material that is in effect static so that we
don't over adjust and over spend on YT slide show type
content.
(Rebase of If25a2449a415449c150acff23df713e9598d64c9
to resolve a auto-merge error)
Change-Id: Iec4e1613ef0d067454751d8220edb7058dfbd816
Allow min and maxQ to creep when the undershoot
or overshoot exceeds thresholds controlled by the
command line under_shoot_pct and over_shoot_pct
values.
Default is 100%,100% which ~disables adaptation.
Derf results for example undershoot% / overshoot%:-
Head:- Mean abs (%rate error) = 14.4%
This check in:-
25%/25% - Mean abs (%rate error) = 6.7%
PSNR hit -1% SSIM -0.1%
5% / 5% - Mean abs (%rate error) = 2.2%
PSNR hit -3.3% SSIM - 1.1%
Most of the remaining error and most of the quality hit is
at extreme data rates. The adaptation code still has an
exception for material that is in effect static so that we
don't over adjust and over spend on YT slide show type
content.
Change-Id: If25a2449a415449c150acff23df713e9598d64c9
Function will jump to error handler when ref buffer is corrupted.
So "xd->corrupted |= ref_buffer->buf->corrupted;" is useless.
Change-Id: I35353a0637ad0dbb682454e040ef69fa68280bfa
- Some fixes to surface fit.
- Returns variance function as cost rather than sad in the
pattern search and diamond search functions. Only
vp9_pattern_search_sad function used in bigdia search
uses sad as integer 1-away costs.
- Deploys SUBPEL_TREE_PRUNED_MORE for speed 4+.
Results:
derf [Speed 3]: About +0.036% in coding efficiency without any
discernible speed loss.
derf [Speed 4]: About 2-3% faster at -0.199% loss in coding efficiency.
derf [Speed 5]: About 3-4% faster at -0.149% loss in coding efficiency.
Change-Id: I8462f94f6adb46966ca964f2bd0400977357fd63
In model_rd_for_sb function, the spatial domain SSE and variance
are checked to see if transform coefficients are quantized to 0.
Besides that, this patch adds another set of thresholds that are
much more strict. These thresholds are used to conduct a partition
block level check to measure if all its TX blocks are skippable
for YUV planes. If it is true, x->skip is set for this partition
block, and thus its mode search is terminated.
This speeds up the encoding at very low prediction error case,
such as screen sharing application. This patch covers what
rd_encode_breakout_test() does, so that function is removed.
Borg test at speed 3 shows:
For stdhd set, psnr: +0.008%, ssim: +0.014%;
For derf set, psnr: +0.018%, ssim: +0.025%.
No noticeable speed change.
Change-Id: I4e5f15cf10016a282a68e35175ff854b28195944
For input source with size that is not multiple of 8, the size is
rounded to 8 and saved in width or height, the original source sizes
are saved in crop_width and crop_height. This commit corrects the
computation of bottom and right extension amounts to use the orignal
sizes, hence crop_width and crop_height.
In addition, this commit also adds the missed initialization for
uv_crop_width and uv_crop_height.
This addresses issue #834
Change-Id: I084543ca7645a4964b88f7cf8ff668f517d3a39b
The concept:
There's too much noise in source pixels for variance and at low bitrate
the reconstructed looks nothing like the source so we have problems
getting good partitionings with either. This skirts the issue by using
a box blur scaled down version for variance calculations. To compare
against source_var_ moved keyframe to be rd based like source_var.
Change-Id: Ie3babdbfadae324b7b5a76bea192893af27f0624
This commit breaks the overly broad header files into more
targeted and smaller ones, to help better structure the system
layout.
Change-Id: I7b24559d3ea6e582cf5d9bbe8f71459f9824d71b
Fixed an encoder crash. Set skip_txfm to 0 for cases that skip_txfm
isn't calculated. Put memcpy of skip_txfm at right place.
Change-Id: Ib3b6afc1b251a85b2a853c8138fb3393f48cfef6
The functions b_width_log2 and b_height_log2 only do direct
table fetch. This commit unifies such use cases by using the
table directly and removes these functions.
Change-Id: I3103fc6ba959c1182886a2799d21b8b77c8a7b6b
Add comments on the use case of these definitions. Further reduce
the scope of header file in vp9_context_tree.h.
Change-Id: Ic4a7638e838d0ac441b64abfc56e57354c059d75
The coefficient range checking is enabled when configured with
--enable-debug --enable-coefficient-range-checking
for vpxdec to detect ill-formed input stream. This addresses the
problem raised by issue #792.
Change-Id: I3f9ea541de4dc742dd64389d6c5f543fb1c4f052
This commit fixes a buffer pointer mis-use in store_coding_context.
The compression performance for stdhd set of speed 3 is improved by
0.097%. It fixes issue 869.
Change-Id: Idc59e22035eaf39f7133ca04174894374d647ff7
This SSE2 is based on VP8 denoiser's SSE2 code. In VP8, there are
only 16x16 blocks in denoiser, while in VP9, there are 13 different
block sizes.
By adding this SSE2 code, the improvement of encoder speed is around
20%(using C code vs using SSE2 code), vary for different clips.
The unit test for VP9 denoiser is to confirm that the SSE2 code is
bit-exact with the C code. The unit test covers all block size.
Change-Id: Ic8d8ac26db4ea40a5f146b5678a065af07eaaa3d
Adjustments to the GF interval choice and minimum boost.
Adjustment to the calculation of 2 pass worst q.
Compared to 09/29 head there is metrics hit on derf of
(-0.123%,-0.191%)
Compared to the September 29 head and a baseline on
September 18 baseline the accuracy of the VBR rate control
measured on the derf set is as follows:-
Mean error % / Mean abs(error %)
Sept 18 baseline (-7.0% / 14.76%)
Sept 29 head (-15.7%, 19.8%)
This check in (-1.5% / 14.4%)
The mean undershoot is reduced slightly but the
worst case overshoot on e.g. harbour/highway is
increased. This will be addressed in a later patch.
Change-Id: Iffd9b0ab7432a131c98fbaaa82d1e5b40be72b58
It is possible that the GOLDEN reference frame is not avaiable, in
which setting the predicted mv will be associated with a residual
value of INT_MAX. This commit checks this condition before
left shift and comparison with that of ALTREF frame, to avoid
overflow issue.
Change-Id: Ib98c3149dbdd016f2fe5beaafb13f67d469dd07c
This commit adds proper initialization of segment id for variance AQ
mode in non-rd coding path. It fixes the enc/dec mismatch issue of
rt=7 with --aq-mode=1, as reported in issue #816
Change-Id: I02fa41b96345bf2e66077d5ea553f85ba800f7bb
This commit enables the encoder to skip split partition search if
the bigger block size has all non-zero quantized coefficients in low
frequency area and the total rate cost is below a certain threshold.
It logarithmatically scales the rate threshold according to the
current block size. For speed 3, the compression performance loss:
derf -0.093%
stdhd -0.066%
Local experiments show 4% - 20% encoding speed-up for speed 3.
blue_sky_1080p, 1500 kbps
51051 b/f, 35.891 dB, 67236 ms ->
50554 b/f, 35.857 dB, 59270 ms (12% speed-up)
old_town_cross_720p, 1500 kbps
14431 b/f, 36.249 dB, 57687 ms ->
14108 b/f, 36.172 dB, 46586 ms (19% speed-up)
pedestrian_area_1080p, 1500 kbps
50812 b/f, 40.124 dB, 100439 ms ->
50755 b/f, 40.118 dB, 96549 ms (4% speed-up)
mobile_calendar_720p, 1000 kbps
10352 b/f, 35.055 dB, 51837 ms ->
10172 b/f, 35.003 dB, 44076 ms (15% speed-up)
Change-Id: I412e34db49060775b3b89ba1738522317c3239c8
Incorporates the WRAPLOW macro into the non-highbitdepth transforms
to aid hardware verification between a software C model and an
intended hardware implementation though the use of the configure
options: --enable-experimental --enable-emulate-hardware.
Note that to avoid further discrepancies between the sse/sse2
implementations of the transforms and the C implementation, when the
emulate hardware option is invoked, we also disable sse/sse2/etc.
Also incudes some minor cleanups/renaming etc.
Change-Id: Ib864d8493313927d429cce402982f1c8e45b3287
Miscellaneous bug-fixes for high bitdepth functionality.
With this patch, high bit-depth profiles become mostly functional,
except for an intermittent assert failure issue that is being
tracked.
Change-Id: I6a7fcbdcf1e5b09842e88535f8442d2e1230748c
This commit removes unused header file vp9_onyxc_int.h and repeatedly
included file vpx_ports/mem.h from vp9_block.h
Change-Id: I400b210bd1da48f1880bd50a8f4a6e2c690e15a1
Block transform skipping was implemented based on DCT's energy
conservation property. Modified the thresholds using zero bin
parameters. AC and DC coefficients were checked separately to
allow better identifying of skippable blocks.
Borg test at speed 3 showed:
stdhd set: psnr gain: 0.153%, ssim gain: 0.051%;
derf set: psnr gain: 0.023%, ssim gain: 0.036%
For most test clips, the encoding speedup is 1% - 2%.
parkrun(720p): 7.5% speedup, park_joy(1080p): 3.5% speedup.
Change-Id: If28eb81113a077414f5ca7b021c14f9069b373bb
The commit cleans up the header files in vp9_entropymv.h. This
file should only depend on vp9_mv.h and vp9_prob.h. Remove the
giant vp9_blockd.h from header file list.
Change-Id: I44cd26d2cfd10a16a9325778347dd53f888a874c
For regular inter frames, if the distance from GOLDEN_FRAME is larger
than 2 and if the predicted motion vector of LAST_FRAME gives lower
sse than that of GOLDEN_FRAME, skip the GOLDE_FRAME mode checking in
the rate-distortion optimization. It provides about 5% speed-up at
expense of -0.137% and -0.230% performance down for speed 3. Local
experiment results:
pedestrian 1080p 2000 kbps
66712 b/f, 40.908 dB, 113688 ms ->
66768 b/f, 40.911 dB, 108752 ms
blue_sky 1080p 2000 kbps
51054 b/f, 35.894 dB, 70406 ms ->
51051 b/f, 35.891 dB, 67236 ms
old_town_cross 720p 1500 kbps
14412 b/f, 36.252 dB, 60690 ms ->
14431 b/f, 36.249 dB, 57346 ms
Change-Id: Idfcafe7f63da7a4896602fc60bd7093f0f0d82ca
Moves transform type defines to vp9_common.h from vp9_idct.h
so that they can be included in vp9_rtcd_defs.pl safely.
Change-Id: Id5106227bee5934f7ce8b06f2eb9fa8a9a2e0ddb
This reverts commit eafc8c9c40.
tran_low_t/tran_high_t don't belong in a public header, they're private.
Similarly the public headers shouldn't rely on config defines,
vpx_config.h isn't installed.
Change-Id: I194ec273598da418df8dd727b6c0e78a556740ad
Some header file in vp9_idct.c has been included in vp9_idct.h.
This commit removes these redundant declarations.
Change-Id: I0238c27e4efff5c981eb437022c6bc6970c4e445
This commit fixes a compiling error in vp9_idct.h, where the codec
checks that the intermediate steps of transformation fit within
16-bit length. The issue was due to broken file dependency.
Change-Id: Ib22bba13a1e6df28489cb23d6774c561969f1fdc
When calculating delta in VP8 denoiser, since the block size is fixed to 16x16,
the divisor is 256, which is the number of the pixel.
But in VP9, the block size varies, the divisor should correspond to the block
size.
Change-Id: Ibdc1e5d23ba8c788b0d0dc6d406bcdfc34c1b142
One is a more aggressive version of the pruned subpel tree
search where only a single halfpel candidate is searched.
The search candidate is based on a surface fit result.
The other is a method to obtain the subpel position at one
shot based on the same surface fit.
The methods have not been deployed in any speed setting yet.
Change-Id: I34fef3f2e34f11396c9d1ba97f4be8c4ffca62d3
This commit enables the encoder to skip checking ALTREF inter modes
in ARF coding, if the predicted motion vectors suggest that the
GOLDEN_FRAME provides higher prediction accuracy than ALTREF_FRAME.
It improves the speed 3 encoding speed by about 5%, at the expense
of compression performance loss -0.041% and -0.225% for derf and
stdhd, respectively.
pedestrian_area 1080p 2000 kbps
66705 b/f, 40.909 dB, 118738 ms ->
66732 b/f, 40.908 dB, 113688 ms
old_town_cross 720p 1500 kbps
14427 b/f, 36.256 dB, 62746 ms ->
14412 b/f, 36.252 dB, 60690 ms
blue_sky 1080p 1500 kbps
51026 b/f, 35.897 dB, 73310 ms ->
50921 b/f, 35.893 dB, 70406 ms
bus CIF 1000 kbps
21301 b/f, 34.841 dB, 7326 ms ->
21248 b/f, 34.837 dB, 7196 ms
Change-Id: I76cf88b4d655e1ee3c0cb03c8a5745493040e8d2
This patch re-enabled the feature in Pengchong's patch
(commit 1286126073). Originally, it
was turned on while use_lastframe_partitioning > 0(not used anymore).
Now it was added as a feature, and turned on while speed >= 2.
As described in the original patch, this feature helps speed up the
slideshows in YouTube.
Change-Id: I1b0f18d65da1ee1c8d1e117dabba910c5207c471
