Commit Graph

9345 Commits

Author SHA1 Message Date
Johann Koenig
cc43012674 Merge changes I267050a5,Iebade0ef,Id96a8df3
* changes:
  quantize_fp_32x32 highbd ssse3: enable existing function
  quantize_fp highbd ssse3: use tran_low_t for coeff
  quantize_fp highbd sse2: use tran_low_t for coeff
2017-02-16 20:34:48 +00:00
Yunqing Wang
0bf6b51572 Merge "Structured the mode ordering code to avoid redundant memcpy" 2017-02-16 16:22:54 +00:00
Johann
ff37a911ce quantize_fp_32x32 highbd ssse3: enable existing function
This was created as part of the quantize_fp_ssse3 change. Both
functions use the same source file with different macro parameters.

Change-Id: I267050a559426a85955d215aa0aaca270439c5ab
2017-02-16 07:40:56 -08:00
Johann
4682130b60 quantize_fp highbd ssse3: use tran_low_t for coeff
Change-Id: Iebade0efc0efbb0a80a0f3adbef4962e3a2f25e8
2017-02-16 07:40:56 -08:00
Johann
ac3996a6d1 quantize_fp highbd sse2: use tran_low_t for coeff
Change-Id: Id96a8df33354a7987ce890a3d6798c7375ffa4aa
2017-02-16 07:40:55 -08:00
Johann
44600442dc bitdepth conversion: really use num elements
The previous implementation confused bit/bytes/elements. It was using
'32' as the multiplier but that was mistakenly adopted because a 32x32
transform embedded the stride.

Change-Id: Ieeb867a332416b9a40580b5e7c9b20088e9e691a
2017-02-16 15:02:48 +00:00
Ranjit Kumar Tulabandu
5127e58dab Structured the mode ordering code to avoid redundant memcpy
Change-Id: I4f5d6b54018bd1928cd9e5e42619e6f55b334803
2017-02-16 14:12:33 +00:00
Paul Wilkins
60a10116d1 Merge "Disconnect ARF breakout from frame boost." 2017-02-16 10:02:09 +00:00
Paul Wilkins
543ebc900f Merge "Remove unnecessary factor." 2017-02-16 10:01:58 +00:00
Paul Wilkins
9216ba58d8 Merge "Bug in scale_sse_threshold()" 2017-02-16 10:01:46 +00:00
Paul Wilkins
e6c1993f1b Merge "Additional first pass stats." 2017-02-16 09:39:29 +00:00
Marco
158b300952 vp9: Some code cleanup for aq-mode = 3.
The weight segment needs to only be computed once per frame,
so remove it from the funciton vp9_cyclic_refresh_rc_bits_per_mb(),
which is called within a loop inside vp9_rc_regulate_q.

Change-Id: Ia0e18b89abb97e42c466d4dbc47700d7f76555db
2017-02-15 14:07:04 -08:00
Marco
f82280820a vp9. Use same source_sad threshold for all speeds.
Only affects real-time mode.

Change-Id: Iba836f110c4da936f5173cc0f54424d5b6121bff
2017-02-15 11:28:26 -08:00
Marco
716c1d5ff5 Vp9: Speed 8 aq-mode=3: Reduce computation in estimating bits per mb.
vp9_compute_qdelta_by_rate has almost 2% overhead in profiling on Nexus 6.
Reduce the calling of that function in speed 8 by estimating the delta-q.
Both rtc and rtc_derf show little/no change in avg psnr/ssim.
Encoding speed is 2~3% faster on Nexus 6.

Change-Id: If25933715783f31104a18a5092ea347b1221b5f5
2017-02-15 09:28:16 -08:00
Linfeng Zhang
ccada0636b Merge "Add vpx_highbd_idct16x16_38_add_c()" 2017-02-15 17:06:17 +00:00
paulwilkins
cfc79a357a Disconnect ARF breakout from frame boost.
This small change replaces the frame boost check in the arf group
length break out clause with a test against a prediction decay value.

The boost value is in fact partly dependent on the decay value but
this change means that the per frame boost calculation can be adjusted
without influencing the group length calculation.

The value chosen gives a close match on all the test sets with the previous
code (on average) but it was noted that a lower threshold was slightly better
for 1080P and up and a slightly higher value for small image sizes.

Change-Id: I4d5b9f67d5b17b0d99ea3f796d3d6202fd61ee0c
2017-02-15 10:46:14 +00:00
paulwilkins
b89ba05ab4 Remove unnecessary factor.
Removed unnecessary scaling factor to simplify.

Change-Id: I3fc9c5975a2597e72f1324e09dd586dea1facfa7
2017-02-15 10:45:43 +00:00
paulwilkins
76550dfdc0 Bug in scale_sse_threshold()
The function scale_sse_threshold() returns a threshold scaled
if necessary for use with 10 and 12 bit from an 8 bit baseline.

SSE error values would be expected to rise for the 10 and 12
bit cases where there are more bits of precision.

Hence the threshold used for the test should also be scaled up.

Change-Id: I4009c98b6eecd1bf64c3c38aaa56598e0136b03d
2017-02-15 10:45:03 +00:00
paulwilkins
945ccfee59 Additional first pass stats.
Added counts that split the intra coded blocks into low and high variance.

Change-Id: Ic540144b34d5141659081bb22f7ee16fd6861f14
2017-02-15 10:44:37 +00:00
Paul Wilkins
7635ee0f37 Merge "Aggressive VBR method." 2017-02-15 10:37:02 +00:00
Johann Koenig
61927ba4ac Merge "vp9 fdct higbd neon: connect existing highbd calls" 2017-02-15 01:33:00 +00:00
Linfeng Zhang
e07e74fb0f Add vpx_highbd_idct16x16_38_add_c()
When eob is less than or equal to 38 for high-bitdepth 16x16 idct,
call this function.

BUG=webm:1301

Change-Id: I09167f89d29c401f9c36710b0fd2d02644052060
2017-02-14 17:25:52 -08:00
Yunqing Wang
f2c1aea118 Merge "Row based multi-threading of encoding stage" 2017-02-15 00:54:10 +00:00
Ranjit Kumar Tulabandu
71061e9332 Row based multi-threading of encoding stage
(Yunqing Wang)
This patch implements the row-based multi-threading within tiles in
the encoding pass, and substantially speeds up the multi-threaded
encoder in VP9.

Speed tests at speed 1 on STDHD(using 4 tiles) set show that the
average speedups of the encoding pass(second pass in the 2-pass
encoding) is 7% while using 2 threads, 16% while using 4 threads,
85% while using 8 threads, and 116% while using 16 threads.

Change-Id: I12e41dbc171951958af9e6d098efd6e2c82827de
2017-02-15 00:49:34 +00:00
Johann
3e7aa8fda9 vp9 fdct higbd neon: connect existing highbd calls
Change-Id: Ia8f822bd6e70b3911bc433a5a750bfb6f9a3a75c
2017-02-14 22:11:49 +00:00
Johann Koenig
9c2bb7f342 Merge "quantize_fp highbd neon: use tran_low_t for coeff" 2017-02-14 21:28:23 +00:00
clang-format
4b402746ca apply clang-format
Change-Id: I75e4a9e0b37bd4586f26c8d6c1fa27f3f6ff1bce
2017-02-14 12:45:52 -08:00
Johann
2b24aa87d9 quantize_fp highbd neon: use tran_low_t for coeff
Change-Id: I90fd815f15884490ad138f35df575a00d31e8c95
2017-02-14 10:26:10 -08:00
Yunqing Wang
318ca07657 The bitstream bit match test in multi-threaded encoder
While the new-mt mode is enabled(namely, allowing to use row-based
multi-threading in encoder), several speed features that adaptively
adjust encoding parameters during encoding would cause mismatch
between single-thread encoded bitstream and multi-thread encoded
bitstream. This patch provides a set_control API to disable these
features, so that the bit match bitstream is obtained in the unit
test.

Change-Id: Ie9868bafdfe196296d1dd29e0dca517f6a9a4d60
2017-02-13 13:02:26 -08:00
James Zern
3c4ea94210 cosmetics,vp9_ratectrl: apply clang-format
broken since:
c3f095c8b Merge "Fix to avoid abrupt relaxation of max qindex in recode path"
5f21aba4b Fix to avoid abrupt relaxation of max qindex in recode path

the original change pre-dated the addition of .clang-format

Change-Id: If5e399d9a805bcad9147360b13b36fbc8c560a7c
2017-02-13 11:29:39 -08:00
paulwilkins
ce7b38459a Aggressive VBR method.
VBR method that allows a wider Q range for the first normal frame
in each ARF group and then centers the min - max range for the rest of
the arf group on the chosen Q value for that first frame.

This allows for quite rapid adjustment of the active Q range even if the
initial estimate is poor.

In some cases where the ARF frames themselves are tending to
undershoot but the normal frames are overshooting this can still give
net undershoot. This can be corrected by allowing a larger Q delta for
arf frames but is usually is a sign that the allocation to the arfs was to
high.

Change-Id: Icec87758925d8f7aeb2dca29aac0ff9496237469
2017-02-13 15:42:11 +00:00
Marco
22dcfa80aa vp9: Non-rd mode: use simple block_yrd for 8 bit high bitdepth builds
Temporary fix until optimization work for block_yrd is completed.
This essentially reverts back to the state before the change:
https://chromium-review.googlesource.com/c/433821/

Compression loss is about ~5-6% on RTC set.
Speed-up (from using this simple/model-based block_yrd) over the low
bitdepth builds (which uses more complex block_yrd) is ~5% on 720p.

Change-Id: Ie0af9eb0d111e5595f587870c44f08317403b8d8
2017-02-10 10:15:35 -08:00
Paul Wilkins
c3f095c8b3 Merge "Fix to avoid abrupt relaxation of max qindex in recode path" 2017-02-09 17:17:55 +00:00
Paul Wilkins
82b88a7fd0 Merge "Fix for max qindex calculation of a gf interval" 2017-02-09 17:17:44 +00:00
Linfeng Zhang
0fefc6873a Merge "Add vpx_idct16x16_38_add_c()" 2017-02-08 17:20:19 +00:00
Johann Koenig
b73f99745b Merge "block_error_fp highbd sse2: use tran_low_t for coeff" 2017-02-07 23:26:10 +00:00
Marco Paniconi
71f5314993 Merge "vp9: Denoiser speed-up: increase partition and ac skip thresholds." 2017-02-07 22:25:00 +00:00
Yunqing Wang
b106abe570 Merge "Row based multi-threading of ARNR filtering stage" 2017-02-07 19:55:41 +00:00
Marco Paniconi
259e835b1b Merge "vp9: Adjust rate_err threshold for setting active_worst factor." 2017-02-07 19:25:47 +00:00
Marco
1a5482d4d8 vp9: Denoiser speed-up: increase partition and ac skip thresholds.
Add factor to increase varianace partition and ac skip thresholds,
under certain conditions (noise level and sum_diff), to increase
denoiser speed.

Change-Id: I7671140ef3598bf5f114a72623d68792bcd7b77b
2017-02-07 10:33:13 -08:00
Linfeng Zhang
cf76ee2cb7 Add vpx_idct16x16_38_add_c()
When eob is less than or equal to 38 for 16x16 idct, call this function.

Change-Id: Ief6f3fb16a49ace3c92cebf4e220bf5bf52a6087
2017-02-07 09:40:51 -08:00
Marco
3c2f076ad0 vp9: Adjust rate_err threshold for setting active_worst factor.
Only affects 1 pass vbr.
Small improvement on ytlive set.

Change-Id: I09a7456fe658fbea82ece1035cf683bd8bd8bd14
2017-02-07 09:38:16 -08:00
Johann
537949a9df block_error_fp highbd sse2: use tran_low_t for coeff
BUG=webm:1365

Change-Id: Id2ed3ebaaaa6a4b68628c23e08b64ea5f1341761
2017-02-07 15:03:28 +00:00
Ranjit Kumar Tulabandu
91f01a2060 Row based multi-threading of ARNR filtering stage
Change-Id: Ic238d32c7e10b730342224ab56712a89a6026a8f
2017-02-07 14:03:19 +05:30
Johann Koenig
85f3a82355 Merge "highbd x86: consolidate tran_low_t conversions" 2017-02-07 02:49:58 +00:00
Jerome Jiang
aa327a1ed4 vp9: speed 8: Tune threshold of ac skip and partitioning.
Threshold for partitioning only affects VGA and lower res.
0.07% quality regression is observed in borg tests on rtc_derf
and 0.2% regression on rtc.
5.6% speed up for low res and 6.8% for VGA on Nexus 6.

Change-Id: If85a2919b48c991de66059c90f32ed06980452be
2017-02-06 16:27:53 -08:00
Johann
641fda79bb highbd x86: consolidate tran_low_t conversions
Create new helper files specifically for converting tran_low_t types.

Change-Id: I7c4c458ef910f3b3d10a3cfbf9df4de7682fd905
2017-02-06 10:43:26 -08:00
Yunqing Wang
dbc5090b5e Merge "Changes to facilitate multi-threading of encoding stage" 2017-02-04 01:02:29 +00:00
Ranjit Kumar Tulabandu
12ec948490 Changes to facilitate multi-threading of encoding stage
Modified the encoding stage to have row level entry points with relevant
initializations and to access the token information at row level

Change-Id: Ife10e55a7c1a420ee906d711caf75002688d9e39
2017-02-02 14:47:13 +05:30
Yunqing Wang
770c6663d6 Merge "Changes to facilitate row based multi-threading of ARNR filtering" 2017-02-01 22:04:15 +00:00