Commit Graph

8659 Commits

Author SHA1 Message Date
Yaowu Xu
eace551c87 Merge changes Icf9b57c3,I9e12da84,Idf5ee179
* changes:
  Fixed interval, fixed Q 1 pass test patch.
  1 pass VBR mode bug fix.
  Fixed interval, fixed Q 1 pass test patch.
2015-12-15 17:51:33 +00:00
Marco Paniconi
12084f6d57 Merge "Revert "Add "unknown" status for noise estimation."" 2015-12-15 16:46:06 +00:00
Marco Paniconi
f3e7539c67 Revert "Add "unknown" status for noise estimation."
This reverts commit e15fedb925.

Change-Id: Ibf2bce008c727a9754f88814b7630095fa7b8253
2015-12-15 16:44:40 +00:00
Marco Paniconi
93c0b879d4 Merge "SVC 1 pass mode: Constrain inter mode search within superframe." 2015-12-15 16:25:20 +00:00
Paul Wilkins
a5af49331d Merge "1 pass VBR mode bug fix." 2015-12-15 15:50:05 +00:00
paulwilkins
fc50d95b2e Fixed interval, fixed Q 1 pass test patch.
For testing implemented a fixed pattern and delta, 1 pass,
fixed Q, low delay mode.

This has not in any way been tuned or optimized.

Change-Id: Idf5ee179b277fa15d07a97f14f2ce5bbaae80a04
2015-12-15 15:00:38 +00:00
paulwilkins
cea5e1c1e3 1 pass VBR mode bug fix.
The one pass VBR mode selects a Q range based on a
moving average of recent Q values. This calculation
should have been excluding arf overlay frames as these
are usually coded at the highest allowed value. Their
inclusion skews the average and can cause it to drift
upwards even when the clip as a whole is undershooting.

As such it can undermine correct adaptation of the allowed
Q range especially for easy content.

Change-Id: I7d10fe4227262376aa2dc2a7aec0f1fd82bf11f9
2015-12-15 10:27:51 +00:00
James Zern
b81f04a0cc Merge "move vp9_avg to vpx_dsp" 2015-12-15 03:41:22 +00:00
Jacky Chen
b7654afb6b Merge "Add "unknown" status for noise estimation." 2015-12-15 00:41:23 +00:00
jackychen
e15fedb925 Add "unknown" status for noise estimation.
Change-Id: I0fe95332ccfa2e1ad2a01a8e7ddd631289e0f8eb
2015-12-14 15:38:20 -08:00
Marco
c760c33b99 SVC 1 pass mode: Constrain inter mode search within superframe.
Keep track of frame indexes for the references, and
constrain inter mode search for reference with same
temporal alignment.

Improves speed by about ~15%, no noticeable loss in
compression performance.

Change-Id: I5c407a8acca921234060c4fcef4afd7d734201c8
2015-12-14 15:19:29 -08:00
James Zern
d36659cec7 move vp9_avg to vpx_dsp
Change-Id: I7bc991abea383db1f86c1bb0f2e849837b54d90f
2015-12-14 14:42:12 -08:00
Marco
6f17954f85 Non-rd variance partition: Adjust logic for 32->16 force split.
Lower the threshold for splitting 32x32->16x16 based on average variance,
and add lower bound condition for this split to occur. This prevents
unneccassry splitting for areas with very low variance.

Change-Id: Ibeb33b3d993632c2019f296eb87ef3b7e3568189
2015-12-14 12:54:10 -08:00
Marco Paniconi
e19b7df8d3 Merge "Non-rd variance partition: Adjustments to reduce dragging artifact." 2015-12-12 02:59:33 +00:00
Marco
d4440614ae Non-rd variance partition: Adjustments to reduce dragging artifact.
For non-rd variannce partition, speed >= 5:
Adjustments to reduce dragging artifcat of background area near
slow moving boundary.

-Decrease base threshold under low source noise conditions.
-Add condition to split 64x64/32x32 based on average variances
of lower level blocks.

PSNR/SSIM metrics go down ~0.7/0.9% on average on RTC set.
Visually helps to reduce dragging artifact on some rtc clips.

Change-Id: If1f0a1aef1ddacd67464520ca070e167abf82fac
2015-12-11 16:16:02 -08:00
Jingning Han
27bbfd652d Fix sub8x8 motion search on scaled reference frame
This commit makes the sub8x8 block rate-distortion optimization
scheme use precise motion compensated prediction to compute the rd
cost. It fixes a potential buffer overflow issue related to sub8x8
motion search on scaled reference frame.

Change-Id: I4274992ef4f54eaacfde60db045e269c13aaa2de
2015-12-11 10:08:51 -08:00
Jingning Han
72760976a0 Merge "Sync high bit-depth temporal filter" 2015-12-10 22:54:59 +00:00
Paul Wilkins
449e46958c Merge "Backport temporal filter approach to VP9" 2015-12-10 09:47:25 +00:00
Jingning Han
d3c972403a Sync high bit-depth temporal filter
Change-Id: Ifdcfb91416be8189569f703bee9be253d7b3d9b6
2015-12-09 15:06:36 -08:00
Jingning Han
ece4fd5d22 Backport temporal filter approach to VP9
This commit enables the new temporal filter system for VP9. For
speed 1, it improves the compression performance:
derf  0.54%
stdhd 1.62%

Change-Id: I041760044def943e464345223790d4efad70b91e
2015-12-09 13:39:06 -08:00
Yaowu Xu
74c67e3da3 Merge "Changes to exhaustive motion search." 2015-12-09 15:57:10 +00:00
Jacky Chen
d9bba21306 Merge "Add vp9_avg_4x4_neon and the unit test." 2015-12-09 06:09:33 +00:00
jackychen
303f144eef Add vp9_avg_4x4_neon and the unit test.
Change-Id: I3ef9a9648841374ed3cc865a02053c14ad821a20
2015-12-08 17:23:36 -08:00
Marco Paniconi
835f16ea36 Merge "vp9 denoiser: Re-evaluate mode selection for golden reference." 2015-12-09 00:34:09 +00:00
paulwilkins
4e692bbee2 Changes to exhaustive motion search.
This change has been imported from VP9 and
alters the nature and use of exhaustive motion search.

Firstly any exhaustive search is preceded by a normal step search.
The exhaustive search is only carried out if the distortion resulting
from the step search is above a threshold value.

Secondly the simple +/- 64 exhaustive search is replaced by a
multi stage mesh based search where each stage has a range
and step/interval size. Subsequent stages use the best position from
the previous stage as the center of the search but use a reduced range
and interval size.

For example:
  stage 1: Range +/- 64 interval 4
  stage 2: Range +/- 32 interval 2
  stage 3: Range +/- 15 interval 1

This process, especially when it follows on from a normal step
search, has shown itself to be almost as effective as a full range
exhaustive search with step 1 but greatly lowers the computational
complexity such that it can be used in some cases for speeds 0-2.

This patch also removes a double exhaustive search for sub 8x8 blocks
which also contained  a bug (the two searches used different distortion
metrics).

For best quality in my test animation sequence this patch has almost
no impact on quality but improves encode speed by more than 5X.

Restricted use in good quality speeds 0-2 yields significant quality gains
on the animation test of 0.2 - 0.5 db with only a small impact on encode
speed. On most natural video clips, however, where the step search
is performing well, the quality gain and speed impact are small.

Change-Id: Iac24152ae239f42a246f39ee5f00fe62d193cb98
2015-12-08 16:54:42 +00:00
Scott LaVarnway
c7e557b82c Merge "VP9: Add ssse3 version of vpx_idct32x32_135_add()" 2015-12-07 21:13:35 +00:00
Marco
d5b3f29f3c Adjust variance threshold based on source noise level.
For non-rd variance partition: Adjust variance threhsold based
on noise level estimate. This change allows the adjustment to be
updated more frequently.

Change-Id: Ie2abf63bf3f1ee54d0bc4ff497298801fdb92b0d
2015-12-04 14:43:39 -08:00
Marco Paniconi
64e46a033f Merge "Non-rd partition: Use force split on 16x16 for low resolutions." 2015-12-04 19:21:26 +00:00
Marco
6490fc71a7 Non-rd partition: Use force split on 16x16 for low resolutions.
For low resolutions, whem 4x4downsample is used for variance,
use the same force split (that is used for 8x8downsample) for 16x16 blocks.

No change in metrics. Small improvement visually.

Change-Id: I915b9895902d0b9a41e75d37fee1bf3714d2366d
2015-12-04 09:24:28 -08:00
Paul Wilkins
2b5baea8fd Merge "Fix bug when overlaying middle arfs in multi-arf groups." 2015-12-04 10:33:55 +00:00
Marco Paniconi
6202ce5ada Merge "vp9-noise estimate: Move level setting to a function." 2015-12-04 00:24:49 +00:00
Marco
dd998adc7a vp9-denoiser: Increase threshold for mode re-evaluation.
Change-Id: I57a15aec1cb2d6638f5211d30c2c9f15fb62494f
2015-12-03 13:48:35 -08:00
Marco
b12e353424 vp9-noise estimate: Move level setting to a function.
This is so we may update level at any time (e.g., to be used
for setting thresholds in variance-based partition).

Change-Id: I32caad2271b8e03017a531f9ea456a6dbb9d49c7
2015-12-03 13:11:49 -08:00
paulwilkins
4a79503b3e Fix bug when overlaying middle arfs in multi-arf groups.
Do not reset the alt ref active flag when overlaying the middle
arf(s) of a multi arf group.

Change-Id: Ia55a55a376973f3fd17161429fd2afb07b4df31f
2015-12-03 15:19:02 +00:00
Scott LaVarnway
f0b0b1fe62 VP9: Add ssse3 version of vpx_idct32x32_135_add()
Change-Id: I9a780131efaad28cf1ad233ae64c5c319a329727
2015-12-02 04:50:46 -08:00
Debargha Mukherjee
7ceba7c26b Fix a spatial svc assert failure
Fixes spatial svc rc assert failure introdcued in:
https://chromium-review.googlesource.com/#/c/312959/1

Change-Id: I6096bfbc484859d71a5fb55e6a3248a31885af61
2015-12-01 14:24:50 -08:00
Debargha Mukherjee
01a2b40e95 Merge "Spatial SVC crash fix" 2015-12-01 21:24:46 +00:00
Debargha Mukherjee
d3409bad9a Fix a spatial svc bug related to scaling
Fixes bug introduced in
https://chromium-review.googlesource.com/#/c/299482/5

Change-Id: If542c1a917380465dd9bc4ce5e32b0adbb20e340
2015-12-01 10:40:59 -08:00
Marco
1abf575f32 vp9 denoiser: Re-evaluate mode selection for golden reference.
Under certain denoising conditons, check for re-evaluation of
zero_last mode if best mode was golden reference.

Change-Id: Ic6cdfd175eef2f7d68606300c7173ab6654b3f6e
2015-12-01 09:39:01 -08:00
Marco
f78b7daec4 Condition use of minmax in variance partition on speed setting.
For non-rd variance partition: only allow minmax computation
(which currently has no arm-neon optimization) for speeds < 8.

Performance loss is small: On RTC set with speed 8, few clips lose ~2/3%,
average loss is < 1%.

Change-Id: Ia9414f4d0b77dc83c3e73ca8de5d903f64b425ce
2015-11-30 17:23:32 -08:00
Marco Paniconi
23831545a0 Merge "vp9 denoiser: Fix to re-evaluate mode selection." 2015-11-30 19:00:39 +00:00
Marco
f1f74a4e6c vp9: Update to noise estimation for denoising.
Change initial state of noise level, and only update
denoiser with noise level when estimate is done.

Change-Id: If44090d29949d3e4927e855d88241634cdb395dc
2015-11-30 10:03:20 -08:00
Marco
ad7e765319 vp9 denoiser: Fix to re-evaluate mode selection.
This fix allows to enable reuse_inter_pred.

Change-Id: I53f2bf1163bb0036ffb6df92117a86debdca11d1
2015-11-30 08:59:10 -08:00
Marco Paniconi
610b413d7b Merge "vp9 denoiser: Re-evaluate ZEROMV after denoiser filtering." 2015-11-25 04:24:00 +00:00
James Zern
fd51d90159 Merge changes Iaf8cbe95,I6748183d,I2a49811d
* changes:
  add vp9_satd_neon
  fix vp9_satd_sse2
  vp9_satd: return an int
2015-11-25 01:48:53 +00:00
Marco
5b0ddb931d vp9 denoiser: Re-evaluate ZEROMV after denoiser filtering.
For denoising, and for noise level above threshold, re-evaluate
ZEROMV for mode selection after denoising.
Current change only does this check if selected best mode (before denoising)
was intra.

Change-Id: I4b1435b68d26c78f7597b995ee7bff0ddd5f9511
2015-11-24 17:30:32 -08:00
Debargha Mukherjee
e807517a93 Spatial SVC crash fix
Fixes a spatial_svc breakage introduced in
https://chromium-review.googlesource.com/#/c/305228/3.

Change-Id: I7f2cecbdca980addb85d5e58b58b5454f4730ada
2015-11-24 16:40:27 -08:00
James Zern
eb1d0f8d60 add vp9_satd_neon
~60-65% faster at the function level across block sizes

Change-Id: Iaf8cbe95731c43fdcbf68256e44284ba51a93893
2015-11-24 16:09:10 -08:00
Marco
fbd245c598 vp9-denoiser: Fix to reset frame_stats.
zeromv_lastref_sse was not reset.

Change-Id: I23c12e804d63dc7dc18514f6efe71de1d1acbd6a
2015-11-24 15:58:28 -08:00
Marco Paniconi
e99e4a64e0 Merge "vp9 non-rd pickmode: Fix logic in reference masking." 2015-11-24 19:14:35 +00:00
Alex Converse
4b038ad2ef Merge "Deduplicate some high bit depth tables" 2015-11-24 18:24:32 +00:00
Marco
eb43c8ebfc vp9 non-rd pickmode: Fix logic in reference masking.
This change makes sure last reference with zero mv
is always checked for mode selection.

No change in metrics.

Change-Id: Iaf01877bf34272b966c78bfe18daad882a0a419e
2015-11-24 10:10:03 -08:00
Scott LaVarnway
26eb806342 Merge "VP9: Only zero counts when !frame_parallel_decoding_mode (2)" 2015-11-23 23:36:46 +00:00
Scott LaVarnway
2c3b737af6 VP9: Only zero counts when !frame_parallel_decoding_mode (2)
The counts are never used when frame_parallel_decoding_mode
is set.

Change-Id: I293aa68abadcdd30973adacb9f5f5a3aecf8daa2
2015-11-23 14:42:15 -08:00
Marco
b0027b96ae vp9-svc: Fix to allow setting qp-max/min per spatial and temporal layer.
Change-Id: Ic0ec32c1d7f7c08c9f956592dccbfd9060b1f624
2015-11-23 10:46:34 -08:00
Marco
131c1600a9 vp9 denoiser: Bias to last reference for temporal filter.
Change-Id: I6a360a12e8da8cdcb8a779647512591612d64f31
2015-11-20 15:38:32 -08:00
James Zern
60760f710f fix vp9_satd_sse2
accumulate satd in 32-bits
+ add unit test

Change-Id: I6748183df3662ddb9d635f9641f9586f2fd38ad5
2015-11-20 14:35:46 -08:00
James Zern
3e0138edb7 vp9_satd: return an int
the final sum may use up to 26 bits

+ add a unit test
+ disable the sse2 as the result will rollover; this will be fixed in a
future commit

Change-Id: I2a49811dfaa06abfd9fa1e1e65ed7cd68e4c97ce
2015-11-20 14:35:38 -08:00
Marco Paniconi
64a60ce3ba Merge "vp9-svc: Fix the setting of is_key_frame." 2015-11-20 18:29:15 +00:00
Marco
80a3e2615a vp9-svc: Fix the setting of is_key_frame.
Change on affects 1 pass CBR.
On key frame, temporal layer_id is reset to 0 for 1 pass CBR,
but since "layer" is reset, the svc.layer_context[layer].is_key_frame
was not correspondingly set properly.

Change-Id: I08f6da0a55ac7429ccfbaddfb7be14479e43543b
2015-11-20 08:51:13 -08:00
Alex Converse
b1fcd1751e Fix unsigned overflow in rd_variance_adjustment.
Found with clang -fsanitize=integer

Change-Id: I2538e7483cb2d5f06bceecbd3326bdd88bfecfa1
2015-11-19 15:00:59 -08:00
Paul Wilkins
f3f6b6fe3e Merge "Changes to best quality settings." 2015-11-19 16:13:43 +00:00
Marco
eed5494fc6 vp9-svc: Fix to key frame counter for spatial layers.
Existing condition only applied to temporal layers.

Change-Id: Icef20a59d0afc61d4e14dea01aff4786fa9e41ae
2015-11-18 14:31:37 -08:00
Paul Wilkins
85aea16f17 Merge "Changes to exhaustive motion search." 2015-11-18 11:10:13 +00:00
Zoe Liu
8a782c7eac Fixed a few sanity checks.
Change-Id: Ieec4a7be5945dc6de192e2d8292ab978baf47f53
(cherry picked from commit 2096296421)
2015-11-17 22:54:03 +00:00
paulwilkins
8ba98516fd Changes to best quality settings.
Small changes to the best quality default speed trade off.
Some speedup settings are worth while even for best quality as they
have only a very small impact on quality but a significant impact on
encode time.

These changes give as much as a further 50-60% increase in encode
speed for my test animations clip with minimal impact on quality.

For this sequence these changes improve the best quality encode  speed
to about the same level as good quality speed 0 in Q3 2015 whilst
retaining the large quality gain of over 1 db

For many natural videos though the quality difference from good 0
to best is much smaller.

Change-Id: I28b3840009d77e129817a78a7c41e29cb03e1132
2015-11-17 16:20:20 +00:00
Marco
988fd77c1f Reduce sampling time for noise estimate.
Change-Id: I46abd85e2187b8f4c2846416a23fab26d9b9f67d
2015-11-13 08:11:30 -08:00
Marco Paniconi
5f5d185d01 Merge "VP9 noise estimation: add frame level motion metrics and adjust thresholds." 2015-11-13 14:09:19 +00:00
paulwilkins
0149fb3d6b Changes to exhaustive motion search.
This change alters the nature and use of exhaustive motion search.

Firstly any exhaustive search is preceded by a normal step search.
The exhaustive search is only carried out if the distortion resulting
from the step search is above a threshold value.

Secondly the simple +/- 64 exhaustive search is replaced by a
multi stage mesh based search where each stage has a range
and step/interval size. Subsequent stages use the best position from
the previous stage as the center of the search but use a reduced range
and interval size.

For example:
  stage 1: Range +/- 64 interval 4
  stage 2: Range +/- 32 interval 2
  stage 3: Range +/- 15 interval 1

This process, especially when it follows on from a normal step
search, has shown itself to be almost as effective as a full range
exhaustive search with step 1 but greatly lowers the computational
complexity such that it can be used in some cases for speeds 0-2.

This patch also removes a double exhaustive search for sub 8x8 blocks
which also contained  a bug (the two searches used different distortion
metrics).

For best quality in my test animation sequence this patch has almost
no impact on quality but improves encode speed by more than 5X.

Restricted use in good quality speeds 0-2 yields significant quality gains
on the animation test of 0.2 - 0.5 db with only a small impact on encode
speed. On most clips though the quality gain and speed impact are small.

Change-Id: Id22967a840e996e1db273f6ac4ff03f4f52d49aa
2015-11-13 10:16:31 +00:00
JackyChen
6fb3d6db99 VP9 noise estimation: add frame level motion metrics and adjust thresholds.
Change-Id: Ia1aba00603b32cee6835951d3d8f740937cf20f4
2015-11-12 23:41:42 -08:00
James Zern
34159b72d9 Merge "Add AVX vectorized vp9_diamond_search_sad" 2015-11-13 06:29:20 +00:00
Marco
419da5c734 Adjust variance threshold for 16x16 split at low resolutions.
Change-Id: I635e37f81237e9703d7d9a11ed76a043f4ec6eb0
2015-11-12 17:58:31 -08:00
Marco Paniconi
866c9357c2 Revert "Update to noise estimation."
This reverts commit 6b79a1e3e0.

Change-Id: I5a4923ca8a6de842855ce0725e92567ccbed6fb7
2015-11-13 00:13:32 +00:00
Marco
6b79a1e3e0 Update to noise estimation.
Add frame level global check and adjust some parameters.

Change-Id: I42103394f2d329781195d94ce6cbb5b3383eea17
2015-11-12 09:18:35 -08:00
Marco Paniconi
1b63238b67 Merge "Non-rd partition: reduce variance threshold low resolutions." 2015-11-12 06:08:38 +00:00
Marco Paniconi
0941ff72a0 Merge "Adjust varianace threshold for high noise condition." 2015-11-12 06:06:51 +00:00
Marco
384fc5e381 Adjust motion threshold to limit cyclic refresh.
Change-Id: Icfca27a567eb8929c312c6315856ee130d982a04
2015-11-11 18:22:21 -08:00
Marco
1827764450 Adjust varianace threshold for high noise condition.
Change-Id: I91c722e480328ff95b8c57614d8176ccaceb2539
2015-11-11 18:06:21 -08:00
Marco Paniconi
4d38dbdfb5 Merge "vp9 denoiser: Add another noise level to denoising." 2015-11-11 20:40:29 +00:00
Marco
ff32369804 vp9 denoiser: Add another noise level to denoising.
Change-Id: Idc755ab54e4f78bb7d75bc97634c451804edad99
2015-11-11 11:21:26 -08:00
James Zern
8f7bc45b5b Revert "VP9: Only zero counts when !frame_parallel_decoding_mode"
This reverts commit 380a5519cc.

This causes an assertion failure in debug_check_frame_counts() which
probably isn't valid with this change; leaving the investigation for
later now.

Change-Id: Ieda5ca811ed2fa50a0cc6935919a8d10dca996e0
2015-11-11 11:11:00 -08:00
Geza Lore
5eefd3ebfd Add AVX vectorized vp9_diamond_search_sad
This function now has an AVX intrinsics version which is about 80%
faster compared to the C implementation. This provides a 2-4% total
speed-up for encode, depending on encoding parameters. The function
utilizes 3 properties of the cost function lookup table, constructed
in 'cal_nmvjointsadcost' and 'cal_nmvsadcosts'.
For the joint cost:
  - mvjointsadcost[1] == mvjointsadcost[2] == mvjointsadcost[3]
For the component costs:
  - For all i: mvsadcost[0][i] == mvsadcost[1][i]
        (equal per component cost)
  - For all i: mvsadcost[0][i] == mvsadcost[0][-i]
        (Cost function is even)
These must hold, otherwise the AVX version of the function cannot be used.

Change-Id: I6c2791d43022822a9e6ab43cd124a773946d0bdc
2015-11-11 14:03:47 +00:00
Marco
064a9eca49 Non-rd partition: reduce variance threshold low resolutions.
Change-Id: I06306905d187948a92f839357df5d21413823808
2015-11-10 15:42:58 -08:00
Marco
bd6bf25969 Add bias to zero/small motion for noisy source.
Change is only for real-time mode, speed >= 5, and non-screen content mode.
Add bias to zero/low motion for big blocks, if noise estimation
is enabled and noise level is above threshold.

Change-Id: I3a0a4608ede6aa535bda6eca528d20f8aba738e7
2015-11-10 11:23:40 -08:00
JackyChen
19272d866b VP9 noise estimate: no noise estimate if frame size change.
Change-Id: I521f7b53c143d562a88fe7de330aa3f0ef09f414
2015-11-09 19:18:29 -08:00
James Zern
e1fbc886e1 Merge "VP9: Only zero counts when !frame_parallel_decoding_mode" 2015-11-09 22:23:34 +00:00
Marco Paniconi
cdec99b243 Merge "VP9 dynamic resize: increase waiting time after key frame." 2015-11-09 21:11:51 +00:00
jackychen
3c9a424e6e VP9 dynamic resize: increase waiting time after key frame.
For 1 pass CBR mode: increase waiting time after key frame
before we start sampling rate control behavior for determining
resize. This change need to disable one internal resize(DownUp)
temporally since it requires a longer clip to do so.

Change-Id: If21beda1be23f169ee541ab4dd642f718347887a
2015-11-09 12:04:00 -08:00
Marco Paniconi
498fd551fd Merge "Use same bias (against non-zero mv for big blocks) for speed 5." 2015-11-09 19:29:35 +00:00
Alex Converse
d1a7c10325 Merge "Expand unconstrained nodes in pack_mb_tokens and loop on zeros." 2015-11-09 18:27:40 +00:00
Scott LaVarnway
380a5519cc VP9: Only zero counts when !frame_parallel_decoding_mode
The counts are never used when frame_parallel_decoding_mode
is set.

Change-Id: Ic7a566a048297f7373c9ffbb48929ea09eff674f
2015-11-09 10:14:13 -08:00
Marco
718654848a Use same bias (against non-zero mv for big blocks) for speed 5.
Use same setting for speed 5 (as it is for speed > 5).
Change is only for real-time (non-rd) mode.

Change-Id: I830250eac654328373cb318baa89d4f0e63942e1
2015-11-09 10:09:51 -08:00
Hui Su
908fbabe4e Merge "Use accurate bit cost for uv_mode in UV intra mode RD selection" 2015-11-07 00:22:50 +00:00
Alex Converse
70eb870cfe Expand unconstrained nodes in pack_mb_tokens and loop on zeros.
Reduces Linux perf estimated cycle count for pack_mb_tokens on a
lossless encode on my desktop from 61858501855 to 48154040219 or from
26% of the overall profile to 21%.

Change-Id: I9ca3426d7e3272bc7f7030abda4f0d0cec87fb4a
2015-11-06 16:00:10 -08:00
hui su
6ab6ac450b Use accurate bit cost for uv_mode in UV intra mode RD selection
On derflr, +0.1% for VP10; however, -0.03% on VP9.

Change-Id: I09c724232ede74254043d61d3cadc506256af0af
2015-11-06 14:45:43 -08:00
James Zern
eba14ddbe7 Merge "Revert "Add AVX vectorized vp9_diamond_search_sad"" 2015-11-06 22:37:20 +00:00
James Zern
30466f26b4 Revert "Add AVX vectorized vp9_diamond_search_sad"
This reverts commit f1342a7b07.

This breaks 32-bit builds:
 runtime error: load of misaligned address 0xf72fdd48 for type 'const
__m128i' (vector of 2 'long long' values), which requires 16 byte
alignment

+ _mm_set1_epi64x is incompatible with some versions of visual studio

Change-Id: I6f6fc3c11403344cef78d1c432cdc9147e5c1673
2015-11-06 13:15:01 -08:00
Marco
5f041c01ed vp9: Disable noise estimate on resize trigger frame.
Change-Id: I35767a6320943582ee11d737b5f240cea2d01b25
2015-11-06 08:42:09 -08:00
Marco
1c724d01aa vp9: Updates to noise estimation.
Add threshold/condition on spatial_variance and brightness level.
Modification to normalization of block variance.
Change resolution limit below which we disable noise estimation.

Change-Id: If5be08a26ceda351242d8a58d2f0bc88c0a918f0
2015-11-05 18:19:01 -08:00
Yunqing Wang
57cae22c1e Merge "Add AVX vectorized vp9_diamond_search_sad" 2015-11-05 20:17:13 +00:00
Geza Lore
f1342a7b07 Add AVX vectorized vp9_diamond_search_sad
This function now has an AVX intrinsics version which is about 80%
faster compared to the C implementation. This provides a 2-4% total
speed-up for encode, depending on encoding parameters. The function
utilizes 3 properties of the cost function lookup table, constructed
in 'cal_nmvjointsadcost' and 'cal_nmvsadcosts'.
For the joint cost:
  - mvjointsadcost[1] == mvjointsadcost[2] == mvjointsadcost[3]
For the component costs:
  - For all i: mvsadcost[0][i] == mvsadcost[1][i]
        (equal per component cost)
  - For all i: mvsadcost[0][i] == mvsadcost[0][-i]
        (Cost function is even)
These must hold, otherwise the AVX version of the function cannot be used.

Change-Id: I184055b864c5a2dc37b2d8c5c9012eb801e9daf6
2015-11-05 10:02:17 +00:00
Marco Paniconi
c6641709a7 Merge "Bias against non-zero mv for large blocks." 2015-11-04 00:01:23 +00:00
Alex Converse
246e0eaa71 Deduplicate some high bit depth tables
Change-Id: I6977f7d155cc1e81ae2393933893caac6770821f
2015-11-03 15:40:44 -08:00
Marco
04a99cb36b Bias against non-zero mv for large blocks.
Change is only for real-time mode, speed > 5, and non-screen content mode.
Bias is based on block size and motion vector level (motion above some threshold).

Helps to improves stability in background from lightning changes.
PSNR/SSIM metrics on RTC set almost no change/neutral (within +/- 0.1).

Change-Id: I7eac13c1ae10be4ab1f40acc7f9f1df5653ece9d
2015-11-03 14:51:56 -08:00
Marco Paniconi
17534d2918 Merge "Update to encoder_breakout_test, for non-rd mode." 2015-11-03 22:40:53 +00:00
Yaowu Xu
5ff1008ed9 Merge "Fix a msvc warning" 2015-11-03 21:56:25 +00:00
Hui Su
3cbe767972 Merge "Generate intra prediction reference values only when necessary" 2015-11-03 20:55:14 +00:00
Marco
9a7785b9d6 Update to encoder_breakout_test, for non-rd mode.
Only use non-zero threshold(s) for breakout if
the motion level of the current tested mode is low.

Change-Id: I22aae961cc42371b49d3f648560181cc54708502
2015-11-03 11:49:44 -08:00
Yaowu Xu
87e08f4d9f Fix a msvc warning
Change-Id: Id5b8f597fb275395232559fea7bfeb56912b88a1
2015-11-03 11:22:58 -08:00
Marco
c7da053d4b Move noise level estimate outside denoiser.
Source noise level estimate is also useful for
setting variance encoder parameters (variance thresholds,
qp-delta, mode selection, etc), so allow it to be used also
if denoising is not on.

Change-Id: I4fe23d47607b4e17a35287057f489c29114beed1
2015-11-02 12:15:26 -08:00
hui su
e085fb643f Generate intra prediction reference values only when necessary
This can help increase encoding speed substantially.

Change-Id: Id0c009146e6e74d9365add71c7b10b9a57a84676
2015-11-02 10:26:50 -08:00
Marco
c2f6a7df8d vp9 denoiser: Don't estimate noise on resized trigger frame.
Change-Id: I60461f011d1aba0b1eb6584c6940f745221915f4
2015-11-02 09:11:35 -08:00
Debargha Mukherjee
9cafc46d9e Merge "Convert motion search config from AoS to SoA" 2015-10-30 20:57:10 +00:00
James Zern
082434b274 Merge changes I3b89e7a6,I8ef772a0
* changes:
  vp9_dx_iface: move struct defs to separate header
  vp9_decodeframe.h: add missing include
2015-10-30 05:50:58 +00:00
James Zern
ca163b85bb vp9_dx_iface: move struct defs to separate header
this avoids redefining vpx_codec_vp9_dx, vpx_codec_vp9_dx_algo in
vp9_encoder_parms_get_to_decoder.cc

Change-Id: I3b89e7a62497227ee32419f1a7d30e4c10a13c05
2015-10-29 17:55:35 -07:00
Alex Converse
d2967221d2 Merge "Make the zero handling in extend_to_full_distribution more explicit." 2015-10-30 00:37:33 +00:00
James Zern
68ecfc1e62 vp9_decodeframe.h: add missing include
Change-Id: I8ef772a016a79cab88bee8e9739530aa030baaa9
2015-10-29 16:41:25 -07:00
Alex Converse
989193c797 Make the zero handling in extend_to_full_distribution more explicit.
The old workaround "p = 0 ? 0 : p -1" is misleading.

?: happens before =
assigning back to p truncates to one byte.

Therefore it is equivalent to (p - 1) & 0xFF, but the check just exists
to work around a first pass bug, so let's make the work around more
clear.

https://bugs.chromium.org/p/webm/issues/detail?id=1089

Change-Id: I587c44dd61c1f3767543c0126376f881889935af
2015-10-29 14:46:55 -07:00
Jacky Chen
039f241fc2 Merge "VP9_resizing: add limitation to the downsacling resolution." 2015-10-29 21:00:36 +00:00
Alex Converse
6f229b3e62 Merge "Shrink probability remap tables." 2015-10-29 19:58:24 +00:00
jackychen
dba2d5b3f3 VP9_resizing: add limitation to the downsacling resolution.
Width and height of downscaling resolution should not be lower
than min_width and min_height which can be set as needed, both
are 180 for now.

Change-Id: I34d06704ea51affbdd814246e22ee8d41d991f00
2015-10-29 09:42:44 -07:00
Marco Paniconi
9645cd4826 Merge "VP9-SVC: Allow frame dropping due to overshoot for spatial layers." 2015-10-28 21:59:17 +00:00
Alex Converse
663960e757 Revert "Replace the zero handling in extend_to_full_distribution."
This reverts commit 7f56cb2978.

It causes uninitialized reads in the first pass setting up later cost tables.

Change-Id: I2df498df3f5c03eff359f79edf045aed0c618dc9
2015-10-28 11:51:40 -07:00
Geza Lore
965a8dea0b Convert motion search config from AoS to SoA
This is a prerequisite for vectorizing vp9_diamond_search_sad_c.

Change-Id: I49cd9148782410ca8b16e8a468ca9e7c6d088410
2015-10-28 15:30:43 +00:00
Marco
823a47ee3b Update to vp9-denoising.
Set increase_denoising parameter for temporal filter.

Change-Id: Id98bf160db98dfa9aedf76e20b43e6f7c783fb1c
2015-10-27 15:52:56 -07:00
Marco
4fb2ba2861 VP9-SVC: Allow frame dropping due to overshoot for spatial layers.
For 1 pass CBR mode.

Change-Id: I8bceb489a850ec26f05382eecb5c0c32a1bb8883
2015-10-27 14:51:47 -07:00
Marco Paniconi
2de14eb942 Merge "Adjustments to vp9-denoising." 2015-10-27 19:10:01 +00:00
Alex Converse
a736bf6bfb Shrink probability remap tables.
Saves 2288 bytes in vp8+vp9 libvpx.a.

Change-Id: Iaa5712e59a9693ed58cea63de63781a96827e44e
2015-10-27 12:08:23 -07:00
Marco
8a2fc54508 Adjustments to vp9-denoising.
Adjust variance threshold, delta-qp, and intra penalty cost,
based on estimated noise level in source.

Replace denoising_on with a level value=L/M/H.

Change-Id: I0c017dae75a5d897367d2c42dec26f2f37e447c1
2015-10-27 10:44:19 -07:00
Alex Converse
89d10d8f3f Merge "Replace the zero handling in extend_to_full_distribution." 2015-10-27 16:54:49 +00:00
Marco Paniconi
dc9d36c0a6 Merge "Code cleanup for vp9-denoiser." 2015-10-26 20:52:16 +00:00
Paul Wilkins
cce3982c48 Merge "Incorrect frame used in KF boost loop." 2015-10-26 19:12:34 +00:00
Paul Wilkins
26abc15e04 Merge "Bug in clamping of base_frame_target." 2015-10-26 19:12:08 +00:00
Marco
f2845ed83c Code cleanup for vp9-denoiser.
Change-Id: Ibb573f50c4bf2cfb382b589803f3363db0ac1285
2015-10-26 12:04:54 -07:00
Alex Converse
7f56cb2978 Replace the zero handling in extend_to_full_distribution.
The old workaround "p = 0 ? 0 : p -1" is misleading.

?: happens before =
assigning back to p truncates to one byte.

Therefore it is equivalent to (p - 1) & 0xFF, but the check just exists
to work around a first pass bug, so let's make the work around more
clear.

https://code.google.com/p/webm/issues/detail?id=1089

Change-Id: Ia6dcc8922e1acbac0eeca23a4d564a355c489572
2015-10-26 11:29:46 -07:00
Debargha Mukherjee
65dd056e41 Merge "Optimize vpx_quantize_{b,b_32x32} assembler." 2015-10-26 18:04:49 +00:00
Debargha Mukherjee
35cae7f1b3 Merge "Optimize vp9_highbd_block_error_8bit assembly." 2015-10-26 18:03:46 +00:00
Paul Wilkins
762c0f2264 Bug in clamping of base_frame_target.
Bug relating to issue:- http://b/25090786

base_frame_target is supposed to track the idealized bit
allocation based on error score and not the actual bits
allocated to each frame.

The clamping of this value based on the VBR min and max pct values
was causing a bug where in some cases the loop that adjusts the
active max quantizer for each GF group was running out of bits at
the end of a KF group. This caused a spike in Q and some ugly artifacts.

A second change makes sure that the calculation of the active
Q range for a group DOES, however, take account of clamping.

Change-Id: I31035e97d18853530b0874b433c1da7703f607d1
2015-10-23 14:45:48 -07:00
Marco
d162934bdc VP9: Estimate noise level for denoiser.
Periodically estiamte noise level in source, and only denoise
if estimated noise level is above threshold.

Change-Id: I54f967b3003b0c14d0b1d3dc83cb82ce8cc2d381
2015-10-23 11:03:30 -07:00
Paul Wilkins
4e887f032d Incorrect frame used in KF boost loop.
Fixes a bug in the calculation of the boost for key frames.

Change-Id: I75e9c96a9e86379239fbbbecb56ccd529783dc7c
2015-10-21 22:17:53 +01:00
Geza Lore
aa8f85223b Optimize vp9_highbd_block_error_8bit assembly.
A new version of vp9_highbd_error_8bit is now available which is
optimized with AVX assembly. AVX itself does not buy us too much, but
the non-destructive 3 operand format encoding of the 128bit SSEn integer
instructions helps to eliminate move instructions. The Sandy Bridge
micro-architecture cannot eliminate move instructions in the processor
front end, so AVX will help on these machines.

Further 2 optimizations are applied:

1. The common case of computing block error on 4x4 blocks is optimized
as a special case.
2. All arithmetic is speculatively done on 32 bits only. At the end of
the loop, the code detects if overflow might have happened and if so,
the whole computation is re-executed using higher precision arithmetic.
This case however is extremely rare in real use, so we can achieve a
large net gain here.

The optimizations rely on the fact that the coefficients are in the
range [-(2^15-1), 2^15-1], and that the quantized coefficients always
have the same sign as the input coefficients (in the worst case they are
0). These are the same assumptions that the old SSE2 assembly code for
the non high bitdepth configuration relied on. The unit tests have been
updated to take this constraint into consideration when generating test
input data.

Change-Id: I57d9888a74715e7145a5d9987d67891ef68f39b7
2015-10-21 12:30:40 +01:00
Geza Lore
9cfba09ac0 Optimize vpx_quantize_{b,b_32x32} assembler.
Added optimization of the 8 bit assembly quantizer routines. This makes
these functions up to 100% faster, depending on encoding parameters.

This patch maskes the encoder faster in both the high bitdepth and 8bit
configurations. In the high bitdepth configuration, it effects profile 0
only.

Based on my profiling using 1080p input the net gain is between 1-3% for
the 8 bit config, and around 2.5-4.5% for the high bitdepth config,
depending on target bitrate. The difference between the 8 bit and high
bitdepth configurations for the same encoder run is reduced by 1% in all
cases I have profiled.

Change-Id: I86714a6b7364da20cd468cd784247009663a5140
2015-10-20 10:11:19 +01:00
James Zern
849e54cedd Merge "vp8cx: remove deprecated reference/entropy controls" 2015-10-20 02:46:36 +00:00
James Zern
a046f56491 vp8cx: remove deprecated reference/entropy controls
VP8E_UPD_ENTROPY, VP8E_UPD_REFERENCE and VP8E_USE_REFERENCE have been
deprecated since the initial public release

Change-Id: Ied16b441eec13434d85f1ab115d49ccaf5f2f7b0
2015-10-16 17:02:36 -07:00
Yaowu Xu
568429512e Add a new enum type vpx_color_range_t
to make meaning of color_range obvious.

Change-Id: I303582e448b82b3203b497e27b22601cc718dfff
2015-10-16 16:27:18 -07:00
Marco
b44c5cf639 Adjustment on limiting cyclic refresh on steady blocks.
Adjust the qp threshold and consec_zeromv threshold for
limiting cyclic refresh. Also increase the refresh period
when the limit amount is significant, and some code-cleanup.

Small gain in PSNR/SSIM metrics: ~0.25/0.3 gain on RTC set, speed 7.

Change only affects non-screen content.

Change-Id: I1ced87a89a132684c071e722616e445b2d18236a
2015-10-16 10:16:44 -07:00
Yaowu Xu
1832ba7509 Restore partial changes from previous commit
This portion was tested to have no effect on asan test failures.

Change-Id: I3de1dab7479148bdffc24c4568cb2e7e9963f099
2015-10-16 00:28:37 +00:00
Jacky Chen
a5d74843eb Merge "VP9_resizing: adjust the threshold and another improvement." 2015-10-15 21:35:02 +00:00
Marco Paniconi
cff15f9d3c Merge "Fix resetting of cyclic refresh on dynamic resize change." 2015-10-15 21:09:06 +00:00
JackyChen
dc002cb7b4 VP9_resizing: adjust the threshold and another improvement.
Adjust the qp threshold based on the denoising setting; not allow
to scale directly from original resolution to one half and vise versa.

Change-Id: I032a9b22f8e1c88de6bb81cf8351367223a3e40d
2015-10-15 09:27:22 -07:00
Marco
d6bbda4bc2 Fix resetting of cyclic refresh on dynamic resize change.
Put the reset at the right place, during the setup and prior
to updating the map.

Change-Id: I75e550ae9d8cc15081330b8857edc04c23947875
2015-10-15 09:03:51 -07:00
Marco
1a0a10cf3d VP9: Rate control update for re-encode screen-content.
For the re-encoding (at max-qp) on the detected high-content change:
update rate correction factor, reset rate over/under-shoot flags,
and update/reset the rate control for layered coding.

Change-Id: I5dc72bb235427344dc87b5235f2b0f31704a034a
2015-10-15 08:26:15 -07:00
Yaowu Xu
4727fa2a75 Fix two asan failures
Change-Id: I57865e9604ac162ef0d97deb16e81ca436a98428
2015-10-14 18:03:31 -07:00
Yaowu Xu
c2b8b5bfe2 Merge "Changes to partition breakout rules." 2015-10-13 22:31:56 +00:00
paulwilkins
cdc359989a Changes to partition breakout rules.
Changes to the breakout behavior for partition selection.
The biggest impact is on speed 0 where encode speed in
some cases more than doubles with typically less than 1%
impact on quality.

Speed 0 encode speed impact examples
Animation test clip: +128%
Park Joy:  +59%
Old town Cross: + 109%

Change-Id: I222720657e56cede1b2a5539096f788ffb2df3a1
2015-10-13 14:19:06 -07:00
Marco
1ce01eaaf7 VP9-SVC: Bugfix to allow skipping lower layer(s) encoding.
The setting of svc->spatial_layer_to_encode was missing
in VP9E_SET_SVC_LAYER_ID.

Change-Id: I015b1a64adb9ef2644d6477a02d9d9364c8462b9
2015-10-12 16:11:34 -07:00
James Zern
ba7ea4456f tile_worker_hook: fix -Wclobbered warning
*tile should be marked volatile like the others due to the use of
setjmp()

Change-Id: I5dbf8e6792e4c0f34a683434b4fd06e3b4c75c4b
2015-10-10 11:17:08 -07:00
James Zern
65055a5fbd Merge "vp9/decode_tiles_mt: remove unnecessary local" 2015-10-09 17:52:34 +00:00
Debargha Mukherjee
94bedd013e Merge "Optimization of 8bit block error for high bitdepth" 2015-10-09 13:36:47 +00:00
Geza Lore
0134764fa6 Optimization of 8bit block error for high bitdepth
If high bit depth configuration is enabled, but encoding in profile 0,
the code now falls back on optimized SSE2 assembler to compute the
block errors, similar to when high bit depth is not enabled.

Change-Id: I471d1494e541de61a4008f852dbc0d548856484f
2015-10-08 14:05:25 -07:00
Jacky Chen
66bf686975 Merge "VP9 denoiser: use skin map to improve denoising." 2015-10-08 21:02:46 +00:00
jackychen
bafe1a2d67 VP9 denoiser: use skin map to improve denoising.
Only denoise at small motion if it's a skin block.

Change-Id: I6235cad9dd7f76ab40e7d9cdfe6180e619c20c6e
2015-10-08 12:17:25 -07:00
jackychen
eaa101b502 vp9_skin_detection: fix some build warnings.
Change-Id: Ib779c083e9775dc9922ed6e104f6275bc453bef9
2015-10-08 09:51:34 -07:00
James Zern
50b20b90aa vp9/decode_tiles_mt: remove unnecessary local
reuse the common loop index

Change-Id: I9db45a93c219c2123917514cb8e9d4ea86454711
2015-10-07 17:46:13 -07:00
James Zern
a83e8ec008 Merge "vp9/tile_worker_hook: pass pbi directly" 2015-10-07 22:09:33 +00:00
James Zern
1f2acb7e40 Merge changes Iaee60826,I51cf1e39
* changes:
  vp9/tile_worker_hook: add multiple tile decoding
  invalid_file_test: loosen error check w/tile-threading
2015-10-07 22:09:21 +00:00
jackychen
b0a2ba2ffa VP9_denoiser: pass address in copy_frame to make it faster.
Change-Id: I65269ddb3ea5f911d5be38614b93c97be7e1ba76
2015-10-07 13:22:37 -07:00
Marco Paniconi
780ada18aa Merge "VP9 denoiser bug-fix: artifact caused by false buffer swap." 2015-10-07 19:08:07 +00:00
Alex Converse
061103dc82 Merge "vp9: simplify extrabits encoding" 2015-10-07 18:45:02 +00:00
jackychen
7231c62c9f VP9 denoiser bug-fix: artifact caused by false buffer swap.
The artifact occurs periodically when VP9 denoiser is on and
refresh_golden_frame happen. When refresh_golden_frame happen,
we should copy the frame buffer instead of swapping the pointers.

Change-Id: Ib3204c4b04db28ecf439c6d9e61f3d146f04196d
2015-10-07 11:16:15 -07:00
James Zern
0bd82af834 vp9/tile_worker_hook: pass pbi directly
reduces the size of TileWorkerData reusing the storage in the worker
itself

Change-Id: If8a62fcb35167037c3da5814ab84fb81893f9cab
2015-10-06 20:14:24 -07:00
James Zern
1f4a6c8a4e vp9/tile_worker_hook: add multiple tile decoding
this reduces the number of synchronizations in decode_tiles_mt() and
improves overall performance when the number of threads is less than the
number of tiles

Change-Id: Iaee6082673dc187ffe0e3d91a701d1e470c62924
2015-10-06 20:13:54 -07:00
Marco
bc137ff67b Move setting of refresh threshold outside loop.
Small code cleanup. consec_zeromv refresh threshold
does not need to be computed for every super-block.

No change in behavior.

Change-Id: I8c4b1b28072f42b01d917fff6d1f62722f1e1554
2015-10-06 17:51:30 -07:00
Alex Converse
2f7f482c77 vp9: simplify extrabits encoding
Change-Id: I5a2abd35cb303d8f6354b3119ab95acf90405116
2015-10-06 16:26:08 -07:00
Marco
7266bedc04 Add first_spatial_layer_to_encode to SVC.
Use the existing VP9_SET_SVC control to set the
first spatial layer to encode.

Since we loop over all spatial layers inside the encoder, the
setting of spatial_layer_id via VP9_SET_SVC has no relevance.
Use it instead to set the first_spatial_layer_to_encode,
which allows an application to skip encoding lower layer(s).

Change only affects the 1 pass CBR SVC.

Change-Id: I5d63ab713c3e250fdf42c637f38d5ec8f60cd1fb
2015-10-06 08:56:15 -07:00
jackychen
de53e6de49 Add the check of resolution in VP9 dynamic resizing.
The resolution check fixs the issue which resets resize_pending
unnecessarily and causes not-bitexact with previous one-step version.

Change-Id: I4e7660b3c8f34f59781e2e61ca30d61080c322de
2015-10-05 15:39:32 -07:00
Marco Paniconi
7777e7a8d5 Merge "Fix to denoiser with dynamic resize." 2015-10-05 14:14:35 +00:00
Marco Paniconi
3da6564f90 Merge "Stabilize the encoder buffer from going too negative." 2015-10-05 14:11:43 +00:00
JackyChen
87b2495f95 Turn on two-steps scaling in VP9 encoder dynamic resizing.
First do a 3/4 scaling and then go down to 1/2 when necessary.

Change-Id: I5689c5228ca7e1606baea7f960eb24d0dab04d4d
2015-10-02 15:27:37 -07:00
Marco
86ede50943 Fix to denoiser with dynamic resize.
Temporary fix to denoiser when dynamic resizing is on.
 -Reallocate denoiser buffers on resized frame.
 -Force golden update on resized frame.
 -Don't denoise resized frame, and copy source into denoised buffers.

Change-Id: Ife7638173b76a1c49eac7da4f2a30c9c1f4e2000
2015-10-02 11:50:57 -07:00
Marco
37293583cd Stabilize the encoder buffer from going too negative.
For screen-content mode, with frame dropper off, put a limit
on how low encoder buffer can go.

Under hard slide changes, the buffer level can go too low and then
take long time to come back up (in particular when frame-dropping
is not used), which will affect the active_worst and target frame size.

Change-Id: Ie9fca097e05cd71141f978ec687f852daf9de332
2015-10-02 11:07:59 -07:00
Debargha Mukherjee
cb5c47f20d Merge "Accelerated transform in high bit depth" 2015-10-02 06:55:55 +00:00
Marco Paniconi
194b374bb6 Merge "Two-steps scaling in VP9 encoder dynamic resizing." 2015-10-02 03:20:22 +00:00
jackychen
ba06be3844 Two-steps scaling in VP9 encoder dynamic resizing.
Dynamic resizing now support two-steps scaling: first go down to
3/4 and then 1/2. This feature is under a flag which controls the
switch between two-steps scaling and one-step scaling (1/2 only).

Change-Id: I3a6c1d3d5668cf8e016a0a02aeca737565604a0f
2015-10-01 18:18:49 -07:00
hui su
06bdc7f6db Small cleanup
Change-Id: I5aeaa94b743f84738d288f8b027fec4c164f2ec3
2015-10-01 11:19:13 -07:00
Scott LaVarnway
2f8625d824 VP9: remove plane_type from macroblockd_plane
Change-Id: Ia5072a3a92212d8565f33359f6c146469bdfbbec
2015-09-30 15:15:11 -07:00
Scott LaVarnway
13888e0eef Merge "VP9: remove plane_type checks in loopfilter functions" 2015-09-30 22:11:21 +00:00
James Zern
a18cc591a5 vp9_loopfilter: remove unnecessary masks
Change-Id: I264e75bf3ddd083ee5311c50a37fb18fe634ddc3
2015-09-30 12:12:53 -07:00
James Zern
a1914dbb31 vp9_reset_lfm: harmonize function signature
Change-Id: Ifb0f41fb43564a777be29b4c66443b366fa146a3
2015-09-29 20:46:37 -07:00
Scott LaVarnway
18373264d9 VP9: remove plane_type checks in loopfilter functions
vp9_filter_block_plane_ss11() and vp9_filter_block_plane_non420()
are only called for the uv planes.

Change-Id: Iacd3b3242c8ce581edd37c8f06d95efc8a0f88a3
2015-09-29 15:54:33 -07:00
Scott LaVarnway
66de2b710f Merge "VP9: move loopfilter build masks to decode loop" 2015-09-29 21:40:48 +00:00
Marco Paniconi
0ca0a536f5 Merge "aq-mode for SVC: Add consec_zero_mv to layer context." 2015-09-29 17:47:39 +00:00
Marco
c05c58f8ff aq-mode for SVC: Add consec_zero_mv to layer context.
Change-Id: I63fadf1c7240d4b2893384f75c519311e9659d47
2015-09-29 10:01:53 -07:00
Yaowu Xu
45948a03c0 Fix a macro definition
to be consistent with the head file name.

Change-Id: I9634332a2b3fac7e7f3b7ef58821ea7c81c5c813
2015-09-29 09:34:42 -07:00
Scott LaVarnway
7718117104 VP9: move loopfilter build masks to decode loop
The loopfilter masks are now built in the decode loop.
This is done so we can eventually reduce the number of
MODE_INFO structs required by the decoder.

The encoder builds the masks for the entire frame prior
to calling the loopfilter.

Change-Id: Ia2146b07e0acb8c50203e586dfae0c4c5b316f11
2015-09-29 05:20:49 -07:00
Julia Robson
406030d1b0 Accelerated transform in high bit depth
When configured with high bitdepth enabled, the 8bit transform
stopped using optimised code. This made 8bit content decode slowly.

Change-Id: I67d91f9b212921d5320f949fc0a0d3f32f90c0ea
2015-09-28 21:09:16 -07:00
Ronald S. Bultje
cc5dd3ec10 Merge "vp9/10: improve support for render_width/height." 2015-09-28 16:25:28 +00:00
Ronald S. Bultje
3db5721e21 Merge "Rename display_{size,width,height} to render_*." 2015-09-28 16:25:20 +00:00
Ronald S. Bultje
812945a8f1 vp9/10: improve support for render_width/height.
In the decoder, map this to the output variable vpx_image_t.r_w/h.
This is intended as an improved version of VP9D_GET_DISPLAY_SIZE,
which doesn't work with parallel frame decoding. In the encoder,
map this to a codec control func (VP9E_SET_RENDER_SIZE) that takes
a w/h pair argument in a int[2] (identical to VP9D_GET_DISPLAY_SIZE).

Also add render_size to the encoder_param_get_to_decoder unit test.

See issue 1030.

Change-Id: I12124c13602d832bf4c44090db08c1009c94c7e8
2015-09-25 22:18:22 -04:00
James Zern
db2056f341 Merge "vp9/10 encoder: prevent NULL access on failure" 2015-09-26 01:52:52 +00:00