Call the idct/add after the tokenize. This is WIP with
the goal of creating a common idct/add for the encoder and
decoder. This move is necessary because the decoder's version
of the idct clobbers qcoeff, which is used by the tokenize.
Change-Id: I6b08d8e8397cd873647fa4fb9469884e3c876756
Added ARM optimized intra 4x4 prediction
- 2x faster on Profiler compared to C-code compiled with -O3
- Function interface changed a little to improve BLOCKD structure
access
Change-Id: I9bc2b723155943fe0cf03dd9ca5f1760f7a81f54
vp8cx_mb_init_quantizer() needs to be called at least once to get
all values calculated. This change added one check to decide if
we could skip initialization or not.
Change-Id: I3f65eb548be57580a61444328336bc18c25c085b
Added code to clip the buffer level to the maximum buffer
size. Without this the buffer level would increase
unchecked.
This bug was found when encoding an essentially static
scene at 2Mb/s. The encoder is unable to generate frames
consistent with the high data-rate because Q bottoms out
at Qmin.
As frames generated are consistently undersized the buffer
level increases and does not get checked against the
maximum size specified by the user (or default).
Change-Id: Id8a3c6323d3246da50f7cb53ddbf78b5528032c6
Fix compiler warning for passing a non const array
to a function expecting a const array by using an
intermediary pointer and casting.
Change-Id: I9bdd358ebdc926223993fb8fb2098ffedd2f3fc7
Changed 'int eob' to 'char *eob' in BLOCKD so that both encoder and
decoder will use eobs[25] array from MACROBLOCKD structure. In future,
this will enable use of the decoder side IDCT in the encoder.
Change-Id: I6e1c011628cb8864fd4a0b80f0279ce16a5ca978
In some situations (f.g. error-resilient is turned on), vp8cx_mb
_init_quantizer() was called once per macroblock. Added checks
to avoid calculations when there is no change.
Change-Id: Ie4f0a5ade2202041254990a4e9d5b03bd1ac5aea
It is discovered that in rare situations the 2nd order block may
produce a few small magnitude coefficients that has no effect on
reconstruction. The situations are a combination of low quantizer
values (high quality) and low energy in residual signals (content
dependent). This commit added code to detect such cases and reset
the 2nd order block to all 0.
Patch 1 to 4 used code to do all-zero-check on idct result buffer,
and tests on derf set showed a consistent gain of .12%-.14% on all
metrics.But due to a recent change Ie31d90b, the idct result buffer
is not longer populated. So patch 5&6 use an alternative method to
detect the situations. Tests on derf set now shows a consistent
quality gain of .16%-.20%.
As suggested by Jim, Patch 7&8 removed the condition of all first
order block not having any coefficient, instead we reset 2nd order
coefficients to all 0 if sum of absolute value of the coefficients
is small. So it does slightly more than just detecting the oddity
as discussed above, but tests on derf set now show a consistent
gain of .20%-.23% on all metrics.
It is worth noting here that this change does not have any effect
on mid/high quantizer range, it only affects the quantizer value
18 or blow. Within this range, the change helps compression by up
to 2.5% on clips in the derf set.
Change-Id: I718e19cf59a4fc2462cb7070832759beb9f7e7dd
The partial frame copy function used to copy an extra 8 lines above
and below. The partial frame filtering can only modify 3 pixel rows
above the partial frame. Reduce copy to bare minimum needed, which is
4 lines, so that partial filtering on copied frame is possible.
Define the "magic" fraction number for partial filtering in
loopfilter.h .
Change-Id: I4791ffc541b6884b12759a0d0714a8faf16147ec
check to make sure that cx_data buffer has enough room before
writting to it, prior behavior did not which could result in a crash.
Change-Id: I3fab6f2bc4a96d7c675ea81acd39ece121738b28
During the _pick only the Y plane is examined. In addition, data beyond
the borders of the frame is not read.
Change-Id: Ic549adfca70fc6e0b55f8aab0efe81f0afac89f9
Instead of using the predict buffer, the decoder now writes
the predictor into the recon buffer. For blocks with eob=0,
unnecessary idcts can be eliminated. This gave a performance
boost of ~1.8% for the HD clips used.
Tero: Added needed changes to ARM side and scheduled some
assembly code to prevent interlocks.
Patch Set 6: Merged (I1bcdca7a95aacc3a181b9faa6b10e3a71ee24df3)
into this commit because of similarities in the idct
functions.
Patch Set 7: EC bug fix.
Change-Id: Ie31d90b5d3522e1108163f2ac491e455e3f955e6
It was crashing when number of partitions was bigger than the number
of MB rows (ex. 128x96 with 8 partitions).
Start point was not checked against mb_rows, plus extra
"empty" partitions were not written out.
Change-Id: I9c2f013b9ec022354b658fab4ef799ff8b1de93d
Added the ability to create rate-targeted, temporally
scalable, VP8 compatible bitstreams.
The application vp8_scalable_patterns.c demonstrates how
to use this capability. Users can create output bitstreams
containing upto 5 temporally separable streams encoded
as a single VP8 bitstream.
(previously abandoned as:
I92d1483e887adb274d07ce9e567e4d0314881b0a)
Change-Id: I156250a3fe930be57c069d508c41b6a7a4ea8d6a
buffer_level in VP8_COMP and starting_buffer_level, optimal_buffer_level
and maximum_buffer_size in VP8_CONFIG changed from int to int64_t
to avoid potential crash issues for larger target bit rates.
Change-Id: I0d5ab6c8a44c2fef51f30cd8df4bb4b739c5df26
Previous entropy probs need to be saved (and restored) only when
current updates are not propagated.
Change-Id: Ie6ee0543066e30874e56258be0a6b7d2dd2fdb2b
Uninitialized data could be written to the first pass file when no
motion vectors are present in the frame.
Also fix a number of compiler warnings.
Change-Id: Icc9f53b6d33da9de4563d86d9fd591910473ea90
vp8_find_near_mvs() is being called on all possible reference frames
but the data computed may be used if the loop exits early, which can
be due to x->skip beign set to 1.
Optimize this by call vp8_find_near_mvs() laziy only if it is going
to be used and not computed yet.
Change-Id: Iccdbd4c962a670c9f2c99b8aca8096042ca5dc98
Changes to the selection of Q limits for two pass
and two pass CQ mode.
Allowance made for Mode and motion vector costs.
Some refactoring of common code.
For Derf and YT sets CQ mode average improvement
circa 1% (SSIM and Global PSNR).
Some increased tendency to undershoot even when
user CQ not reached.
Patch2: Removed some test code accidentally merged.
Change-Id: Icf74d13af77437c08602571dc7a97e747cce5066
Sync with loopfilter thread just at the beginning of next frame encoding.
This returns control to application faster and allows a better multicore scaling.
When PSNR packets are generated the final filtered frame is needed imediatly
so we cannot delay the sync.
Change-Id: I288d97b5e331d41d6f5bb49d97986fa12ac6f066
- Removed fast_fdct4x4_neon and fast_fdct8x4_neon
- Uses now short_fdct4x4 and short_fdct8x4
- Gives ~1-2% speed-up on Cortex-A8/A9
Change-Id: Ib62f2cb2080ae719f8fa1d518a3a5e71278a41ec
Rd and Rm registers should be different in 'mul'. This register
combination results in unpredictable behaviour. GCC will give
a warning and RVCT an error in this case.
Restriction applies only to armv5 targets and not for armv6 and above.
Change-Id: I378d17c51e1f16a6820814fbed43e115aaabb03e
- Updated walsh transform to match C
(based on Change Id24f3392)
- Changed fast_fdct4x4 and 8x4 to short_fdct4x4 and 8x4
correspondingly
Change-Id: I704e862f40e315b0a79997633c7bd9c347166a8e
Modified original patch If2f07220885c4c3a0cae0dace34ea0e36124f001
according to comments. Scheduled code a little bit to prevent some
interlocks.
Change-Id: I338f02b881098782f82af63d97f042b85e63e902
vp8_update_zbin_extra() is called all the time even though the fast
quantizer doesn't use it. Skip this call if fast quantizer is used.
Change-Id: Ia711c38431930cc2486cf59b8466060ef0e9d9db
This change makes sure that no key frame recoding in real-time mode
even if CONFIG_REALTIME_ONLY is not configured.
Change-Id: Ifc34141f3217a6bb63cc087d78b111fadb35eec2
for SPLITMV and B_PRED modes. Modified code to use the bmi
found in mode_info_context instead of BLOCKD. On the decode
side, the uvmvs are calculated only when required, instead of
every macroblock. This is WIP. (bmi should eventually be
removed from BLOCKD)
Small performance gains noticed for RT encodes and decodes.(VGA)
Change-Id: I2ed7f0fd5ca733655df684aa82da575c77a973e7
Prepend . to local labels in assembly code. This
allows non unique labels within a file. Also
makes profiling information more informative
by keeping the function name with the loop name.
Change-Id: I7a983cb3a5ba2413d5dafd0a37936b268fb9e37f
Calculations were incorrectly classified as either
SSE3 or SSSE3. Only using SSE2 instructions.
Cleanup function names and make non-RTCD code work
as well.
Change-Id: I48ad0218af0cc51c5078070a08511dee43ecfe09
Calculations were incorrectly classified as either
SSE3 or SSSE3. Only using SSE2 instructions.
Cleanup function names and make non-RTCD code work
as well.
Change-Id: I29f5c2ead342b2086a468029c15e2c1d948b5d97
When active map is specified and the current frame is not a key frame,
golden frame nor a altref frame then copy only those active regions.
This significantly reduces encoding time by as much as 19% on the test
system where realtime encoding is used. This is particularly useful
when the frame size is large (e.g. 2560x1600) and there's only a few
action macroblocks.
Change-Id: If394a813ec2df5a0201745d1348dbde4278f7ad4
Instead of a single mid GF boost apply a few extra bits to
every other frame. This gives a very small average metrics
improvement on both derf and YT sets.
Also use min GF interval as min KF interval.
Change-Id: Iee238b8cae0ffaed850a5a944ac825cee18da485
This reverts commit b5ea2fbc2c. Further
testing showed noticable keyframe popping in some cases, reverting this
for now to give time for a proper fix.
Conflicts:
vp8/encoder/onyx_if.c
vp8/encoder/ratectrl.c
Change-Id: I159f53d1bf0e24c035754ab3ded8ccfd58fd04af
This patch fixes a bug in the interaction between the recode loop and
spatial resampling. If the codec was in a spatial resampling state,
and a subsequent iteration of the recode loop disables resampling,
then the source buffer must be reset to the unscaled source.
Change-Id: I4e4cd47b943f6cd26a47449dc7f4255b38e27c77
Changed motion search in vp8_find_best_half_pixel_step() to be the
same as in vp8_find_best_sub_pixel_step(), which checks 5 points
instead of 8 points. This only affects real-time mode with
cpu-used >=9. Tests showed it gives 2% encoding speedup with
a quality loss(psnr) of up to 0.5%.
Change-Id: I16049cad1535002346d46cfdfad345bfc3dc5146
This change implemented same idea in change "Preload reference area
to an intermediate buffer in sub-pixel motion search." The changes
were made to vp8_find_best_sub_pixel_step() and vp8_find_best_half
_pixel_step() functions which are called when speed >= 5. Test
result (using tulip clip):
1. On Core2 Quad machine(Linux)
rt mode, speed (-5 ~ -8), encoding speed gain: 2% ~ 3%
rt mode, speed (-9 ~ -11), encoding speed gain: 1% ~ 2%
rt mode, speed (-12 ~ -14), no noticeable encoding speed gain
2. On Xeon machine(Linux)
Test on speed (-5 ~ -14) didn't show noticeable speed change.
Change-Id: I21bec2d6e7fbe541fcc0f4c0366bbdf3e2076aa2
There were some situations that the start motion vectors were
out of range. This fix adjusted range checks to make sure they
are checked and clamped.
Change-Id: Ife83b7fed0882bba6d1fa559b6e63c054fd5065d
sharpness was not recalculated in vp8cx_pick_filter_level_fast
remove last_filter_type. all values are calculated, don't need to update
the lfi data when it changes.
always use cm->sharpness_level. the extra indirection was annoying.
don't track last frame_type or sharpness_level manually. frame type
only matters for motion search and sharpness_level is taken care of in
frame_init
move function declarations to their proper header
Change-Id: I7ef037bd4bf8cf5e37d2d36bd03b5e22a2ad91db
In sub-pixel motion search, the search range is small(+/- 3 pixels).
Preload whole search area from reference buffer into a 32-byte
aligned buffer. Then in search, load reference data from this buffer
instead. This keeps data in cache, and reduces the crossing cache-
line penalty. For tulip clip, tests on Intel Core2 Quad machine(linux)
showed encoder speed improvement:
3.4% at --rt --cpu-used =-4
2.8% at --rt --cpu-used =-3
2.3% at --rt --cpu-used =-2
2.2% at --rt --cpu-used =-1
Test on Atom notebook showed only 1.1% speed improvement(speed=-4).
Test on Xeon machine also showed less improvement, since unaligned
data access latency is greatly reduced in newer cores.
Next, I will apply similar idea to other 2 sub-pixel search functions
for encoding speed > 4.
Make this change exclusively for x86 platforms.
Change-Id: Ia7bb9f56169eac0f01009fe2b2f2ab5b61d2eb2f
This is done by expanding luma row to 32-byte alignment, since
there is currently a bunch of code that assumes that
uv_stride == y_stride/2 (see, for example, vp8/common/postproc.c,
common/reconinter.c, common/arm/neon/recon16x16mb_neon.asm,
encoder/temporal_filter.c, and possibly others; I haven't done a
full audit).
It also uses replaces the hardcoded border of 16 in a number of
encoder buffers with VP8BORDERINPIXELS (currently 32), as the
chroma rows start at an offset of border/2.
Together, these two changes have the nice advantage that simply
dumping the frame memory as a contiguous blob produces a valid,
if padded, image.
Change-Id: Iaf5ea722ae5c82d5daa50f6e2dade9de753f1003
allowing the compiler to inline this function. For real-time
encodes, this gave a boost of 1% to 2.5%, depending on the
speed setting.
Change-Id: I3929d176cca086b4261267b848419d5bcff21c02
This patch attempts to improve the handling of CBR streams with
respect to the short term buffering requirements. The "buffer level"
is changed to be an average over the rc buffer, rather than a long
running average. Overshoot is also tracked over the same interval
and the golden frame targets suppressed accordingly to correct for
overly aggressive boosting.
Testing shows that this is fairly consistently positive in one
metric or another -- some clips that show significant decreases
in quality have better buffering characteristics, others show
improvenents in both.
Change-Id: I924c89aa9bdb210271f2e03311e63de3f1f8f920
Optimized C-code of the following functions:
- vp8_tokenize_mb
- tokenize1st_order_b
- tokenize2nd_order_b
Gives ~1-5% speed-up for RT encoding on Cortex-A8/A9
depending on encoding parameters.
Change-Id: I6be86104a589a06dcbc9ed3318e8bf264ef4176c
Do mvp clamping in full-pixel precision instead of 1/8-pixel
precision to avoid error caused by right shifting operation.
Also, further fixed the motion vector limit calculation in change:
b748045470
Change-Id: Ied88a4f7ddfb0476eb9f7afc6ceeddbf209fffd7
Separate simple filter with reduced no. of parameters.
MB filter level picking based on precalculated table. Level table updated for
each frame. Inside and edge limits precalculated and updated just when
sharpness changes. HEV threshhold is constant.
ARM targets use scalars and others vectors.
Change works only with --target=generic-gnu
All other targets have to be updated!
Change-Id: I6b73aca6b525075b20129a371699b2561bd4d51c
Allow the encoder to inform the application that the encoded frame will not
be used as a reference.
Change-Id: I90e41962325ef73d44da03327deb340d6f7f4860
Motion vector limits are calculated using right shifts, which
could give wrong results for negative numbers. James Berry's
test on one clip showed encoder produced some artifacts. This
change fixed that.
Change-Id: I035fc02280b10455b7f6eb388f7c2e33b796b018
In this commit I have added an experimental function
that tests prediction quality either side of a central position
to calculate a suggested boost number for an ARF frame.
The function is passed an offset from the current position and
a number of frames to search forwards and backwards.
It returns a forward, backward and compound boost number.
The new code can be deactivated using #define NEW_BOOST 0
In its current default state the code searches forwards and backwards
from the proposed position of the next alt ref.
The the old code used a boost number calculated by scanning forward
from the previous GF up to the proposed alt ref frame position.
I have also added some code to try and prevent placement of a gf/arf
where there is a brief flash.
Change-Id: I98af789a5181148659f10dd5dd2ff2d4250cd51c
There were many instances in the code of vp8_coef_tokens and
vp8_coef_tokens-1, which was a preprocessor macro despite the naming
convention. Replace these with MAX_ENTROPY_TOKENS and ENTROPY_NODES,
respectively.
Change-Id: I72c4f6c7634c94e1fa066cd511471e5592c748da
With this commit frames can be received partition-by-partition
from the encoder and passed partition-by-partition to the
decoder.
At the encoder-side this makes it easier to split encoded
frames at partition boundaries, useful when packetizing
frames. When VPX_CODEC_USE_OUTPUT_PARTITION is enabled,
several VPX_CODEC_CX_FRAME_PKT packets will be returned
from vpx_codec_get_cx_data(), containing one partition
each. The partition_id (starting at 0) specifies the decoding
order of the partitions. All partitions but the last has
the VPX_FRAME_IS_FRAGMENT flag set.
At the decoder this opens up the possibility of decoding partition
N even though partition N-1 was lost (given that independent
partitioning has been enabled in the encoder) if more info
about the missing parts of the stream is available through
external signaling.
Each partition is passed to the decoder through the
vpx_codec_decode() function, with the data pointer pointing
to the start of the partition, and with data_sz equal to the
size of the partition. Missing partitions can be signaled to
the decoder by setting data != NULL and data_sz = 0. When
all partitions have been given to the decoder "end of data"
should be signaled by calling vpx_codec_decode() with
data = NULL and data_sz = 0.
The first partition is the first partition according to the
VP8 bitstream + the uncompressed data chunk + DCT address
offsets if multiple residual partitions are used.
Change-Id: I5bc0682b9e4112e0db77904755c694c3c7ac6e74
Adding support in the encoder for generating
independent residual partitions by forcing
equal probabilities over the prev coef entropy
contexts.
Change-Id: I402f5c353255f3ca20eae2620af739f6a498cd21
I got this idea from Pascal (Thanks). Before encoding a macroblock,
copy it to a 16x16 buffer, and then read source data from there
instead. This will help keep the source data in cache, and help
with the performance.
Change-Id: Id05f4cb601299150511d59dcba0ae62c49b5b757
This reverts commit 212f618373.
Further testing shows that the overshoot accumulation/damping is too
aggressive on some clips. Allowing the accumulated overshoot to
decay and limiting to damping to golden frames shows some promise.
But some clips show significant overshoot in the buffer window, so
I think this still needs work.
Change-Id: Ic02a9ca34f55229f9cc04786f4fab54cdc1a3ef5
RDMULT/RDDIV defines a bit worth of distortion in term of sum squared
difference. This has also been used as errorperbit in subpixel motion
search, where the distortions computed as variance of the difference.
The variance of differences is different from sum squared differences
by amount of DC squared. Typically, for inter predicted MBs, this
difference averages around 10% between the two distortion, so this patch
introduces a 110% constant in deriving errorperbit from RDMULT/RDDIV.
Test on CIF set shows small but positive gain on overall PSNR (.03%)
and SSIM (.07%), overall impact on average PSNR is 0.
Change-Id: I95425f922d037b4d96083064a10c7cdd4948ee62
The starting points are always within the limits, and bounds
checking on these points is not needed. For speed < 5, the
encoded result changes a little because different treatment
is taken while starting point equals the bounds.
Change-Id: I09a402d310f51e305a3519f1601b1d17b05c6152
Modify the second-pass code to provide a full golden-frame (GF) bit
allocation boost if the past GF group (GFG) had no alt-ref frame (ARF),
even if the current GFG does contain and ARF.
This mostly has no effect on clips, since switching ARFs on/off between
GFGs is not very common. Has a positive effect on e.g. cheer (+0.45 SSIM
at 600kbps) and football (+0.25 SSIM at 600kbps), particularly at high
bitrates. Has a negative effect (-0.04 SSIM at 300kbps) at pamphlet,
which appears only marginally related to this patch, and crew (-0.1 SSIM
at 700kbps).
Change-Id: I2e32899638b59f857e26efeac18a82e0c0b77089
The BPRED mode selection uses SSE as a distortion metric, but the early
breakout threshold being used was a variance value.
Change-Id: I42d4602fb9b548bf681a36445701fada5e73aff1
firstpass.c contains some rate adjustment code that assures that the
last few frames in a sequence abide by rate limits. If the second-to-
last group of frames contains an alt-ref frame (ARF), the last golden
frame (GF) is zero bytes, and we will thus spend a ridiculously high
number of bits on regular P-frames trying to hit the target rate. This
does slightly enhance the quality of these last few frames, but has
no perceptual value (other than hitting the target rate).
Disabling this code means we consistently (slightly) undershoot the
target rate and consequently do worse on the last few frames of a
clip, which is particularly noticeable for small clips. The quality-
per-bitrate is generally better, ~0.2% better overall on derf-set,
especially on clips such as garden, tennis, foreman at low bitrates.
Has a negative effect on hallmonitor at high bitrates.
Change-Id: I1d63452fef5fee4a0ad2fb2e9af4c9f2e0d86d23
Moved encode_intra function from firstpass.c to encodeintra.c to
prevent linking problem in real-time only build. Also changed name
of the function to vp8_encode_intra because it is not a static.
Change-Id: Ibf3c6c1de3152567347e5fbef47d1d39564620a5
Small decode performance gain (~1%) on keyframes. No
noticeable gains on encode. Also changed pick_intra4x4mby_modes()
to read the above and left block modes for keyframes only.
Change-Id: I1f4885252f5b3e9caf04d4e01e643960f910aba5
Since BPRED will be tested at most once, and SPLITMV is not enabled,
there's nothing to clobber the subblock modes, so there's no need to
save and restore them.
Change-Id: I7c3615b69190c10bd068a44df5488d6e8b85a364
In activity masking, RDO constant RDMULT is adjusted on a per MB basis
adaptive to activity with the MB. errorperbit, which is defined as
RDMULT/RDDIV, is a constant used in motion estimation. Previously, in
activity masking, errorperbit is not changed even when RDMULT is changed.
This commit changed to adjust errorperbit according to the change in
RDMULT.
Test in cif set showed a very small but consistent gain by all quality
metrics (average, overall psnr and ssim) when activity masking is on.
Change-Id: I07ded3e852919ab76757691939fe435328273823
This change is analogous to I0b67dae1f8a74902378da7bdf565e39ab832dda7,
which made the move for the non-RD path.
Change-Id: If63fc1b0cd1eb7f932e710f83ff24d91454f8ed1
