Rebased.
Remove the old matrix multiplication transform computation. The 16x16
ADST/DCT can be switched on/off and evaluated by setting ACTIVE_HT16
300/0 in vp9/common/vp9_blockd.h.
Change-Id: Icab2dbd18538987e1dc4e88c45abfc4cfc6e133f
rebased.
This patch includes 16x16 butterfly inverse ADST/DCT hybrid
transform. It uses the variant ADST of kernel
sin((2k+1)*(2n+1)/4N),
which allows a butterfly implementation.
The coding gains as compared to DCT 16x16 are about 0.1% for
both derf and std-hd. It is noteworthy that for std-hd sets
many sequences gains about 0.5%, some 0.2%. There are also few
points that provides -1% to -3% performance. Hence the average
goes to about 0.1%.
Change-Id: Ie80ac84cf403390f6e5d282caa58723739e5ec17
Removal of the NEWCOEFCONTEXT experiment to
reduce code clutter and make it easier to experiment with
some other changes to the coefficient coding context.
Change-Id: Icd17b421384c354df6117cc714747647c5eb7e98
fixed format issues.
Implement the inverse 4x4 ADST using 9 multiplications. For this
particular dimension, the original ADST transform can be
factorized into simpler operations, hence is retained.
Change-Id: Ie5d9749942468df299ab74e90d92cd899569e960
Refactor the 8x8 inverse hybrid transform. It is now consistent
with the new inverse DCT. Overall performance loss (due to the
use of this variant ADST, and the rounding errors in the butterfly
implementation) for std-hd is -0.02.
Fixed BUILD warning.
Devise a variant of the original ADST, which allows butterfly
computation structure. This new transform has kernel of the
form: sin((2k+1)*(2n+1) / (4N)). One of its butterfly structures
using floating-point multiplications was reported in Z. Wang,
"Fast algorithms for the discrete W transform and for the discrete
Fourier transform", IEEE Trans. on ASSP, 1984.
This patch includes the butterfly implementation of the inverse
ADST/DCT hybrid transform of dimension 8x8.
Change-Id: I3533cb715f749343a80b9087ce34b3e776d1581d
Linking when we don't use it but it is available is probably harmless.
Gtest requires pthreads. Don't automatically enable unit tests if we
don't have it.
Change-Id: I5e6c3b609f840c7b6dbb36fc65809f0ef84685f8
This experiment gives little gains and adds relatively much code
complexity (and it hinders other experiments), so let's get rid of
it.
Change-Id: Id25e79a137a1b8a01138aa27a1fa0ba4a2df274a
This patch removes the old pred-filter experiment and replaces it
with one that is implemented using the switchable filter framework.
If the pred-filter experiment is enabled, three interopolation
filters are tested during mode selection; the standard 8-tap
interpolation filter, a sharp 8-tap filter and a (new) 8-tap
smoothing filter.
The 6-tap filter code has been preserved for now and if the
enable-6tap experiment is enabled (in addition to the pred-filter
experiment) the original 6-tap filter replaces the new 8-tap smooth
filter in the switchable mode.
The new experiment applies the prediction filter in cases of a
fractional-pel motion vector. Future patches will apply the filter
where the mv is pel-aligned and also to intra predicted blocks.
Change-Id: I08e8cba978f2bbf3019f8413f376b8e2cd85eba4
This is to add to the 64x64 transform experiment as an alternative to
a 64x64 DCT.
Two levels of wavelet decomposition is used on a 64x64 block, followed
by 16x16 DCT on the four lowest subbands. The highest three subbands
are left untransformed after the first level DWT.
Change-Id: I3d48d5800468d655191933894df6b46e15adca56
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.
Change-Id: Ie708b54d8e1898af742846ce2d1e2b0d89fd4ad5
First attempt at avoiding all the compile-time environment detection for
cases where you can generate the environments statically, as when the
real build is being performed by another build system.
Change-Id: Ie3cf95d71d6c5169900f31e263b84bc123cdf73f
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.
Change-Id: I7bc57a5fb6dbf1f85eac1543daaeb3a61633275c
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.
Change-Id: I00386f20f35d7fabb19aba94c8162f8aee64ef2b
A patch on compound inter-intra prediction.
In compound inter-intra prediction, a new predictor for
16x16 inter coded MBs are obtained by combining a single
inter predictor with a 16x16 intra predictor, in a manner
that the weight varies with distance from the top/left
boundary. The current search strategy is to combine the best
inter mode with the best intra mode obtained independently.
Results so far:
derf +0.31%
yt +0.32%
std-hd +0.35%
hd +0.42%
It is conceivable that the results would improve somewhat
with a more thorough search strategy where all intra modes
are searched given the best mv, or even a joint search for
the best mv and the best intra mode.
Change-Id: I7951f1ed0d6eb31ca32ac24d120f1585bcd8d79b
Preliminary patch on a new 4x4 intra mode B_CONTEXT_PRED where the
dominant direction from the context is used to encode. Various decoder
changes are needed to support decoding of B_CONTEXT_PRED in conjunction
with hybrid transforms since the scan order and tokenization depends on
the actual direction of prediction obtained from the context. Currently
the traditional directional modes are used in conjunction with the
B_CONTEXT_PRED, which also seems to provide the best results.
The gains are small - in the 0.1% range.
Change-Id: I5a7ea80b5218f42a9c0dfb42d3f79a68c7f0cdc2
Creates a merge between the master and experimental branches. Fixes a
number of conflicts in the build system to allow *either* VP8 or VP9
to be built. Specifically either:
$ configure --disable-vp9 $ configure --disable-vp8
--disable-unit-tests
VP9 still exports its symbols and files as VP8, so that will be
resolved in the next commit.
Unit tests are broken in VP9, but this isn't a new issue. They are
fixed upstream on origin/experimental as of this writing, but rebasing
this merge proved difficult, so will tackle that in a second merge
commit.
Change-Id: I2b7d852c18efd58d1ebc621b8041fe0260442c21
Documentation is typically auto-detected by checking for php and
doxygen. Add an option to explicitly disable it.
Remove toggle keywords from libraries, examples, documentation and
unit tests. They were not consistent with the default status.
Change-Id: I21049675ccfd8e58ac612cd058641b197db5c0eb
As suggested by Paul, this commit separate the subpel refmv selection
into a separate experiment. It also changed a couple variable names
to better reflect the nature of the variables.
Change-Id: Id951c3cadc61a982dd15afe641000f60213b8995
Documentation is typically auto-detected by checking for php and
doxygen. Add an option to explicitly disable it.
Remove toggle keywords from libraries, examples, documentation and
unit tests. They were not consistent with the default status.
Change-Id: I21049675ccfd8e58ac612cd058641b197db5c0eb
Results: derf (vanilla or +hybridtx) +0.2% and (+hybrid16x16
or +tx16x16) +0.7%-0.8%; HD (vanilla or +hybridtx) +0.1-0.2%
and (+hybrid16x16 or +tx16x16) +1.4%, STD/HD (vanilla or +hybridtx)
about even, and (+hybrid16x16 or +tx16x16) +0.8-1.0%.
Change-Id: I03899e2f7a64e725a863f32e55366035ba77aa62
Adds a new experiment with redesigned/refactored motion vector entropy
coding. The patch also takes a first step towards separating the
integer and fractional pel components of a MV. However the fractional
pel encoding still depends on the integer pel part and so they are
not fully independent. Further experiments are in progress to see
how much they can be decoupled without affecting performance.
All components including entropy coding/decoding, costing for MV
search, forward updates and backward updates to probability tables,
have been implemented.
Results so far:
derf: +0.19%
std-hd: +0.28%
yt: +0.80%
hd: +1.15%
Patch: Simplifies the fractional pel models:
derf: +0.284%
std-hd: +0.289%
yt: +0.849%
hd: +1.254%
Patch: Some changes in the models, rebased.
derf: +0.330%
std-hd: +0.306%
yt: +0.816%
hd: +1.225%
Change-Id: I646b3c48f3587f4cc909639b78c3798da6402678
Enable ADST/DCT of dimension 16x16 for I16X16 modes. This change provides
benefits mostly for hd sequences.
Set up the framework for selectable transform dimension.
Also allowing quantization parameter threshold to control the use
of hybrid transform (This is currently disabled by setting threshold
always above the quantization parameter. Adaptive thresholding can
be built upon this, which will further improve the coding performance.)
The coding performance gains (with respect to the codec that has all
other configuration settings turned on) are
derf: 0.013
yt: 0.086
hd: 0.198
std-hd: 0.501
Change-Id: Ibb4263a61fc74e0b3c345f54d73e8c73552bf926
Alternative strategy for finding a list of candidate motion
vectors to use as reference values in mv coding and as
nearest and near.
Sort by sad in vp8_find_best_ref_mvs() rather than just
pick the best. Allow 0,0 as a best ref option but not a
nearest or near unless there are no alternatives.
Encode/Decode verified on at least some clips.
Some commented out experimental and stats code still in place.
Gain over existing code averages about 1% on derf (alll metrics)
with improvement on all clips. Other test results pending.
The entropy coding of the mode (nearest/near etc) still
depends upon and requires the old "findnear" code so
this needs looking at and may provide room for further gains.
Change-Id: I871d7cba1d1c379c4bad9bcccce1fb19c46b8247
If you build with --enabled-shared on a Linux arch not explicitly
listed, the configure script will abort because it didn't detect
"linux" in the fallback generic-gnu tuple.
Since this is the fallback tuple and people are passing
--enable-shared, assume the user knows what they're in for.
Change-Id: Ia35b657e7247c8855e3a94fca424c9884d4241e3
Merges this experiment in to make it easier to run tests on
filter precision, vectorized implementation etc.
Also removes an experimental filter.
Change-Id: I1e8706bb6d4fc469815123939e9c6e0b5ae945cd
Using surrounding reconstructed pixels from left and above to select
best matching mv to use as reference motion vector for mv encoding.
Test results:
AVGPSNR GLBPSNR VPXSSIM
Derf: 1.107% 1.062% 0.992%
Std-hd:1.209% 1.176% 1.029%
Change-Id: I8f10e09ee6538c05df2fb9f069abcaf1edb3fca6
This allows building on MountainLion as the 10.6 SDK has been
removed from the latest Xcode version (4.4 4F250). Also fix
all warnings for that build.
Change-Id: Ib70bca4a25295f13595f0d10ea9f0229631de5a4
Merged in the high_precision_mv experiment to make it easier
to work on new mv encoding strategies. Also removed
coef_update_probs3().
Change-Id: I82d3b0bb642419fe05dba82528bc9ba010e90924
Fixed the code review comments.
Under the htrans8x8 experiment the 8X8 DCT in the
I8X8 mode is replaced with a combination of 8X8 ADST and
DCT.
Overall coding gains with the htrans8x8 experiment are:
derf: 0.486
std-hd: 1.040
hd: 1.063
yt: 0.506
Note that part of the gain comes from bigger transforms
(8x8 instead of 4x4) and part comes from replacing the DCT
wth the ADST.
Change-Id: I92ca6bbfce11b4165d612b81d9adfad4d010c775
Set on all 16x16 intra/inter modes
Features:
- Butterfly fDCT/iDCT
- Loop filter does not filter internal edges with 16x16
- Optimize coefficient function
- Update coefficient probability function
- RD
- Entropy stats
- 16x16 is a config option
Have not tested with experiments.
hd: 2.60%
std-hd: 2.43%
yt: 1.32%
derf: 0.60%
Change-Id: I96fb090517c30c5da84bad4fae602c3ec0c58b1c
Apply 2D-DCT transform of dimension 8x8 to encode prediction
residuals of I8X8 mode.
Brought back block type 3 probability context model for 8x8 tokens,
which is used for the coefficients of Y blocks in I8x8 modes. The
coefficient costs estimate of I8X8 mode in rate-distortion is also
changed appropriately.
Performance results:
derf: 0.246
yt: 0.114
std-hd: 0.730
hd: 0.670
Change-Id: If1d970eeb4e1827c9f0d2c5b27d33089b347ea27
Merged the enhanced_interp experiment.
Found and fixed a bug in the include files framework, whereby
certain encoder files were still using the old INTERP_EXTEND
value of 3 instead of 4. The thresholds for mv range mcomp.c
need a small adjustment to prevent crashes.
The results are more or less unchanged.
Change-Id: Iac5008390f1efc97ce1102fbb5f8989c847fb579
Allows for swtiching/setting interpolation filters at the MB
level. A frame level flag indicates whether to use a specifc
filter for the entire frame or to signal the interpolation
filter for each MB. When switchable filters are used, the
encoder chooses between 8-tap and 8-tap sharp filters. The
code currently has options to explore other variations as well,
which will be cleaned up subsequently.
One issue with the framework is that encoding is slow. I
tried to do some tricks to speed things up but it is still slow.
Decoding speed should not be affected since the number of
filter taps remain unchanged.
With the current version, we are up 0.5% on derf on average but
some videos city/mobile improve by close to 4 and 2% respectively.
If we did a full-search by turning the SEARCH_BEST_FILTER flag
on, the results are somewhat better.
The framework can be combined with filtered prediction, and I
seek feedback regarding that.
Rebased.
Change-Id: I8f632cb2c111e76284140a2bd480945d6d42b77a
The following five experiments are merged:
newentropy
newupdate
adaptive_entropy (also includes a couple of parameter changes
that improves results a little
in common/entropymode.c and encoder/modecosts.c
that were not merged from the internal branch)
newintramodes
expanded_coef_context
Change-Id: I8a142a831786ee9dc936f22be1d42a8bced7d270
Adds ADST/DCT hybrid transform coding for Intra4x4 mode.
The ADST is applied to directions in which the boundary
pixels are used for prediction, while DCT applied to
directions without corresponding boundary prediction.
Adds enum TX_TYPE in b_mode_infor to indicate the transform
type used.
Make coding style consistent with google style.
Fixed the commented issues.
Experimental results in terms of bit-rate reduction:
derf: 0.731%
yt: 0.982%
std-hd: 0.459%
hd: 0.725%
Will be looking at 8x8 transforms next.
Change-Id: I46dbd7b80dbb3e8856e9c34fbc58cb3764a12fcf
the integer version has very good precision, the float version is no
longer useful. this commit also removes the experiment option from
configure script.
Change-Id: Ibb92e63c9f5083357cdf89c559d584a7deb3353f
this commit removes a number of experiment options from configure
script. the associated experiments are already fully merged, the
options in configure script have no effect at all.
Change-Id: I8054ccaee0a04610162ed76ac9e59c4538217113
This commit adds lossless compression capability to the experimental
branch. The lossless experiment can be enabled using --enable-lossless
in configure. When the experiment is enabled, the encoder will use
lossless compression mode by command line option --lossless, and the
decoder automatically recognizes a losslessly encoded clip and decodes
accordingly.
To achieve the lossless coding, this commit has changed the following:
1. To encode at lossless mode, encoder forces the use of unit
quantizer, i.e, Q 0, where effective quantization is 1. Encoder also
disables the usage of 8x8 transform and allows only 4x4 transform;
2. At Q 0, the first order 4x4 DCT/IDCT have been switched over
to a pair of forward and inverse Walsh-Hadamard Transform
(http://goo.gl/EIsfy), with proper scaling applied to match the range
of the original 4x4 DCT/IDCT pair;
3. At Q 0, the second order remains to use the previous
walsh-hadamard transform pair. However, to maintain the reversibility
in second order transform at Q 0, scaling down is applied to first
order DC coefficients prior to forward transform, and scaling up is
applied to the second order output prior to quantization. Symmetric
upscaling and downscaling are added around inverse second order
transform;
4. At lossless mode, encoder also disables a number of minor
features to ensure no loss is introduced, these features includes:
a. Trellis quantization optimization
b. Loop filtering
c. Aggressive zero-binning, rounding and zero-bin boosting
d. Mode based zero-bin boosting
Lossless coding test was performed on all clips within the derf set,
to verify that the commit has achieved lossless compression for all
clips. The average compression ratio is around 2.57 to 1.
(http://goo.gl/dEShs)
Change-Id: Ia3aba7dd09df40dd590f93b9aba134defbc64e34
Added the ability to optionally filter the prediction data
when inter modes are selected (excludes SPLITMV, for now).
The mode selection loop considers both the filtered and
non-filtered prediction data when choosing mode. The filter
can be turned on/off at the frame-level, or signaled for
each MB.
Change-Id: I1b783c71d95a361ab36c761b07e8a6b06bc36822
Disable unit-tests. The logging in GTest would need to be adjusted.
Restructure ARM cpu detection. Flatten if-else logic.
Change #if defined(HAVE_*) to #if HAVE_* because we only need to check
for features that the library was actually built with. This should have
been harmless, as disabled feature sets wouldn't have any features to
call.
Change-Id: Iea21aa42ce5f049c53ca0376d25bcd0f36f38284
This patch incorporates adaptive entropy coding of coefficient tokens,
and mode/mv information based on distributions encountered in a frame.
Specifically, there is an initial forward update to the probabilities
in the bitstream as before for coding the symbols in the frame, however
at the end of decoding each frame, the forward update to the
probabilities is reverted and instead the probabilities are updated
towards the actual distributions encountered within the frame.
The amount of update is weighted by the number of hits within each
context.
Results on derf/hd/std-hd are all up by 1.6%.
On derf, the most of the gains come from coefficients, however for the
hd and std-hd sets, the most of the gains come from the mode/mv
information updates.
Change-Id: I708c0e11fdacafee04940fe7ae159ba6844005fd
The commit added an integer version of 8x8 forward DCT, based on the
orginal forward DCT from VP6. The constants, roundings, and shifts
were adjusted to improve the accuracy. The latest patch has a very
similar accuracy in term of round trip error against the floating
point version.
It should be noted here that the purpose of the patch is to help
encoding speed and facilitate all other experiments. There will be
futher review in combination with inverse DCT before finalization.
configure with "--enable--int_8x8fdct" to use the integer version
Change-Id: I5a4f80507429f0e07cf02a13768ec81cbfddc5bc
Adds differential encoding of prob updates using a subexponential
code centered around the previous probability value.
Also searches for the most cost-effective update, and breaks
up the coefficient updates into smaller groups.
Small gain on Derf: 0.2%
Change-Id: Ie0071e3dc113e3d0d7ab95b6442bb07a89970030
This is the first patch to add superblock (32x32) coding
order capabilities. It does not yet do any mode selection
at the SB level, that will follow in a further patch.
This patch encodes rows of SBs rather than
MBs, each SB contains 2x2 MBs.
Two intra prediction modes have been disabled since they
require reconstructed data for the above-right MB which
may not have been encoded yet (e.g. for the bottom right
MB in each SB).
Results on the one test clip I have tried (720p GIPS clip)
suggest that it is somewhere around 0.2dB worse than the
baseline version, so there may be bugs.
It has been tested with no experiments enabled and with
the following 3 experiments enabled:
--enable-enhanced_interp
--enable-high_precision_mv
--enable-sixteenth_subpel_uv
in each case the decode buffer matches the recon buffer
(using "cmp" to compare the dumped/decoded frames).
Note: Testing these experiments individually created
errors.
Some problems were found with other experiments but it
is unclear what state these experiments are in:
--enable-comp_intra_pred
--enable-newentropy
--enable-uvintra
This code has not been extensively tested yet, so there
is every likelihood that further bugs remain. I also
intend to do some code cleanup & refactoring in tandem
with the next patch that adds the 32x32 modes.
Change-Id: I1eba7f740a70b3510df58db53464535ef881b4d9
When using 'make dist' after --disable-vp8[encoder|decoder] it would
fail to recognize the option. This would only occur when also specifying
--enable-install-docs and --enable-install-srcs but not
--enable-codec-srcs
Including vpx/ fixes builds with --enable-codec-srcs
vpx_timer.h is also required for vpxenc.c
Change-Id: Ie3e28b2f7ec7ee6d5961d3843f9eab869f79c35b
Pulled out super block code for the snapshot as this
is not quite ready and will need an extensive re-merge.
Change-Id: I436369b511257447a7b0ea064016cb63f5011849
Produce the token partitions on-the-fly, while processing each MB.
Context is updated at the beginning of each frame based on the
previoud frame's counters. Optimally encoder outputs partitions in
separate buffers. For frame based output, partitions are concatenated
internally.
Limitations:
- enabled just in combination with realtime-only mode
- number of encoding threads has to be equal or less than the
number of token partitions. For this reason, by default the encoder
will do 8 token partitions.
- vpxenc supports partition output (-P) just in combination with
IVF output format (--ivf)
Performance:
- Realtime encoder can be up to 13% faster (ARM) depending on the number
of threads and bitrate settings. Constant gain over the 5-16 speed
range.
- Token buffer reduced from one frame to 8 MBs
Quality:
- quality is affected by the delayed context updates. This again
dependents on input material, speed and bitrate settings. For VC
style input the loss seen is up to 0.2dB. If error-resilient=2
mode is used than the effect of this change is negligible.
Example:
./configure --enable-realtime-only --enable-onthefly-bitpacking
./vpxenc --rt --end-usage=1 --fps=30000/1000 -w 640 -h 480
--target-bitrate=1000 --token-parts=3 --static-thresh=2000
--ivf -P -t 4 -o strm.ivf tanya_640x480.yuv
Change-Id: I127295cb85b835fc287e1c0201a67e378d025d76
This is the initial patch for supporting 1/8th pel
motion. Currently if we configure with enable-high-precision-mv,
all motion vectors would default to 1/8 pel. Encode and
decode syncs fine with the current code. In the next phase
the code will be refactored so that we can choose the 1/8
pel mode adaptively at a frame/segment/mb level.
Derf results:
http://www.corp.google.com/~debargha/vp8_results/enhinterp_hpmv.html
(about 0.83% better than 8-tap interpoaltion)
Patch 3: Rebased. Also adding 1/16th pel interpolation for U and V
Patch 4: HD results.
http://www.corp.google.com/~debargha/vp8_results/enhinterp_hd_hpmv.html
Seems impressive (unless I am doing something wrong).
Patch 5: Added mmx/sse for bilateral filtering, as well as enforced
use of c-versions of subpel filters with 8-taps and 1/16th pel;
Also redesigned the 8-tap filters to reduce the cut-off in order to
introduce a denoising effect. There is a new configure option
sixteenth-subpel-uv which will use 1/16 th pel interpolation for
uv, if the motion vectors have 1/8 pel accuracy.
With the fixes the results are promising on the derf set. The enhanced
interpolation option with 8-taps alone gives 3% improvement over thei
derf set:
http://www.corp.google.com/~debargha/vp8_results/enhinterpn.html
Results on high precision mv and on the hd set are to follow.
Patch 6: Adding a missing condition for CONFIG_SIXTEENTH_SUBPEL_UV in
vp8/common/x86/x86_systemdependent.c
Patch 7: Cleaning up various debug messages.
Patch 8: Merge conflict
Change-Id: I5b1d844457aefd7414a9e4e0e06c6ed38fd8cc04
