Frank Galligan [Sun, 22 Mar 2015 20:41:13 +0000 (13:41 -0700)]
Refactor dec_build_inter_predictors
Refactor the loops in dec_build_inter_predictors to try and decrease
the number of instructions. Limited testing saw about 1% perf
increase on x86 and about 0.67 % perf increase on Arm.
Modifies a special handling that improves rate control accuracy in
the constrained quality mode, when the undershoot and overshoot
limits are set tighter.
Jingning Han [Tue, 7 Apr 2015 19:39:54 +0000 (12:39 -0700)]
Compute prediction filter type cost only when needed
Skip redundant prediction filter type cost in filter search loop,
if the rate value will be reset in Hadamard transform based rate
distortion estimate.
webmdec: Fix for reaching eof in webm_guess_framerate
Reset the reached_eos flag in webm_guess_framerate in case it ends
up consuming the entire file. Also adding a vpxdec shell test to
verify this behavior.
* changes:
vp8_regular_quantize_b_sse2: remove dead init
vp8cx_pick_filter_level*: remove dead inits
vp8_decode_frame: remove dead increment
rdopt: remove dead stores
find_next_key_frame: remove dead init & store
multiframe_quality_enhance_block: remove dead stores
vp8_print_modes_and_motion_vectors: remove dead stores
Jingning Han [Fri, 3 Apr 2015 18:33:24 +0000 (11:33 -0700)]
Enable Hadamard transform based cost estimate for all block sizes
This commit turns on the Hadamard transform based rate distortion
estimate for all block sizes in RTC coding mode. It conditionally
skips the rate distortion estimation if all zero block flag is set
on. No significant encoding speed change is observed. The
compression performance of speed -6 is improved by 1.7% over using
it only for block sizes of 32x32 and below.
Yunqing Wang [Fri, 3 Apr 2015 22:35:31 +0000 (15:35 -0700)]
Fix the scaling factor in UV skipping test
The threshold scaling factor was calculated wrong using partition
size "bsize". Thank Yaowu for pointing it out. It was fixed and no
speed change was seen.
Ed Baker [Thu, 2 Apr 2015 20:57:13 +0000 (13:57 -0700)]
Test loopfilters with count=2
The following functions use the count parameter to either loop or select
dedicated paths:
vp9_lpf_horizontal_16_c
vp9_lpf_horizontal_16_sse2
vp9_lpf_horizontal_16_avx2
vp9_lpf_horizontal_16_neon
vp9_highbd_lpf_horizontal_16_c
vp9_highbd_lpf_horizontal_16_sse2
Jingning Han [Fri, 3 Apr 2015 16:20:25 +0000 (09:20 -0700)]
Account for eob cost in the RTC mode decision process
This commit accounts for the transform block end of coefficient flag
cost in the RTC mode decision process. This allows a more precise
rate estimate. It also turns on the model to block sizes up to 32x32.
The test sequences shows about 3% - 5% speed penalty for speed -6.
The average compression performance improvement for speed -6 is
1.58% in PSNR. The compression gains for hard clips like jimredvga,
mmmoving, and tacomascmv at low bit-rate range are 1.8%, 2.1%, and
3.2%, respectively.
Yunqing Wang [Thu, 2 Apr 2015 20:57:22 +0000 (13:57 -0700)]
Set vbp thresholds for aq3 boosted blocks
The vbp thresholds are set seperately for boosted/non-boosted
superblocks according to their segment_id. This way we don't
have to force the boosted blocks to split to 32x32.
Speed 6 RTC set borg test result showed some quality gains.
Overall PSNR: +0.199%; Avg PSNR: +0.245%; SSIM: +0.802%.
No speed change was observed.
Adrian Grange [Tue, 24 Mar 2015 15:55:35 +0000 (08:55 -0700)]
Fix use of scaling in joint motion search
To enable us to the scale-invariant motion estimation
code during mode selection, each of the reference
buffers is scaled to match the size of the frame
being encoded.
This fix ensures that a unit scaling factor is used in
this case rather than the one calculated assuming that
the reference frame is not scaled.
Jingning Han [Wed, 1 Apr 2015 00:46:41 +0000 (17:46 -0700)]
Refactor block_yrd function for RTC coding mode
This commit separates Hadamard transform/quantization operations
from rate and distortion computation in block_yrd. This allows one
to skip SATD computation when all transform blocks are quantized
to zero. It also uses a new block error function that skips
repeated computation of sum of squared residuals. It reduces the
CPU cycles spent on block error calculation in block_yrd by 40%.
Jingning Han [Wed, 1 Apr 2015 18:39:36 +0000 (11:39 -0700)]
Optimize quantization simd implementation
This commit allows the quantizer to compare the AC coefficients to
the quantization step size to determine if further multiplication
operations are needed. It makes the quantization process 20% faster
without coding statistics change.
Yunqing Wang [Wed, 25 Mar 2015 21:19:29 +0000 (14:19 -0700)]
Enhance the transform skipping decision-making in non-rd mode
For large partition blocks(block_size > 32x32), the variance
calculation is modified so that every 8x8 block's variance
is stored during the calculation, which is used in the
following transform skipping test. Also, the variance for
every tx block is calculated. The skipping test checks all tx
blocks in the partition, and sets the skip flag only if all tx
blocks are skippable. If the skip flag of Y plane is 1, a
quick evaluation is done on UV planes. If the current partition
block is skippable in YUV planes, the mode search checks fewer
inter modes and doesn't check intra modes.
The rtc set borg test(at speed 6) showed that:
Overall psnr: -0.527%; Avg psnr: -0.510%; ssim: -0.573%.
Average single-thread speedup on rtc set was 3.5%.
For 720p clips, more speedups were seen.
gipsrecmotion: 13%
gipsrestat: 12%
vidyo: 5 - 9%
dark: 15%
niklas: 6%
James Zern [Wed, 1 Apr 2015 03:57:25 +0000 (20:57 -0700)]
vp9: enable sse4 sad functions
sse4 isn't set by configure or used in rtcd, correct the sad entries to
use sse4_1 without changing the signatures for now.
this was done in vp8 post-vp9 branch.
Jingning Han [Tue, 31 Mar 2015 17:57:41 +0000 (10:57 -0700)]
Tuning SATD rate calculation for speed
This commit allows the encoder to check the eob per transform
block to decide how to compute the SATD rate cost. If the entire
block is quantized to zero, there is no need to add anything; if
only the DC coefficient is non-zero, add its absolute value;
otherwise, sum over the block. This reduces the CPU cycles spent
on vp9_satd_sse2 to one third.
Jingning Han [Tue, 31 Mar 2015 16:25:21 +0000 (09:25 -0700)]
Allow block skip coding option in RTC mode
When the estimated rate-distortion cost of skip coding mode is
lower than that of sending quantized coefficients, allow the
encoder to drop these coefficients. This improves the compression
performance of speed -6 by 0.268% and makes the encoding speed
slightly faster.
Jingning Han [Mon, 30 Mar 2015 19:31:46 +0000 (12:31 -0700)]
Enable 16x16 Hadamard transform in SATD based mode decision
This commit replaces the 16x16 2D-DCT transform with Hadamard
transform for RTC coding mode. It reduces the CPU cycles cost
on 16x16 transform by 5X. Overall it makes the speed -6 encoding
speed 1.5% faster without compromise on compression performance.
Jingning Han [Mon, 30 Mar 2015 18:09:29 +0000 (11:09 -0700)]
Use SATD based mode decision for block sizes below 16x16
This commit makes the encoder to select between SATD/variance as
metric for mode decision. It also allows to account chroma
component costs for mode decision as well. The overall encoding
time increase as compared to variance based mode selection is about
15% for speed -6. The compression performance is on average 2.2%
better than variance based approach, with about 5% compression
performance gains for hard clips (e.g., jimredvga, nikas720p, and
mmmoving) at lower bit-rate range.
projects / libvpx / log