Commit Graph

224 Commits

Author SHA1 Message Date
Dmitry Kovalev
aa464eca5e Adding search_site_config struct.
Change-Id: I2ad333553e673dbabcdc0f0366aea311e90849bf
2014-04-29 10:34:53 -07:00
Dmitry Kovalev
b87130093e Cleaning up vp9_full_search_sadx8() function.
Change-Id: I61f977a7d0b177881bd5c1a7179fa5f8f089e9a6
2014-04-25 14:55:02 -07:00
Dmitry Kovalev
fae988dc3b Merge "Cleaning up vp9_full_search_sadx3()." 2014-04-25 13:52:20 -07:00
Dmitry Kovalev
ef003078e8 Renaming "onyx" to "encoder".
Actual renames:
  vp9_onyx_if.c -> vp9_encoder.c
  vp9_onyx_int.h -> vp9_encoder.h

Change-Id: I80532a80b118d0060518e6c6a0d640e3f411783c
2014-04-22 14:57:05 -07:00
Dmitry Kovalev
6733ad19e1 Cleaning up vp9_full_search_sadx3().
Change-Id: Ia63fee65c827fd33080119184e6cf3167421807c
2014-04-14 11:13:33 -07:00
Dmitry Kovalev
58c40986e0 Removing redundant mvsadcost local variables.
Change-Id: I19a197b9525172579f253702c77fa76af88f6565
2014-04-11 14:10:37 -07:00
Dmitry Kovalev
145b24719a Merge "Removing offset argument of mvcomp macros." 2014-04-11 14:09:17 -07:00
Dmitry Kovalev
63fa722179 Removing unused cost arguments from mcomp functions.
Change-Id: Id81a76d18be6b2de69f81bb563d74c3bb356d434
2014-04-11 10:24:36 -07:00
Dmitry Kovalev
d3127382fd Removing offset argument of mvcomp macros.
Change-Id: I8bb72b790a0783720833f27b7cf41055bee6becd
2014-04-10 16:07:41 -07:00
Dmitry Kovalev
c8167f5ba7 Fixing invalid ref buffer in vp9_diamond_search_sad_c().
Problem has been introduced recently with the cleanup patch
I0816ec12ec0a6f21d0f25f10c214b5fd327afc6c

Change-Id: Iaacb956a6039eb5826b82618dc03be32053fb892
2014-04-10 15:16:48 -07:00
Dmitry Kovalev
9243edc737 Cleaning up vp9_diamond_search_sad_c() function.
Change-Id: I0816ec12ec0a6f21d0f25f10c214b5fd327afc6c
2014-04-07 16:28:35 -07:00
Dmitry Kovalev
1ea18dc258 Cleaning up vp9_refining_search_sadx4() function.
Change-Id: I47a7148bf746f61d0c4947f9b22624a8c071359f
2014-04-02 18:59:21 -07:00
Dmitry Kovalev
5c185a0587 Merge "Cleaning up vp9_pattern_search() function." 2014-04-02 18:28:47 -07:00
Dmitry Kovalev
224d986858 Cleaning up vp9_get_mvpred_{av_,}var() functions.
Change-Id: I0df8c2a6d9863f92ee406010f2daeb5e40627649
2014-03-26 15:40:41 -07:00
Dmitry Kovalev
0038bbec32 Cleaning up vp9_pattern_search() function.
Change-Id: I8da5e13b23a3d6b73c82bbe443c82dd87fd689f3
2014-03-25 17:41:19 -07:00
Dmitry Kovalev
ac49c8ca13 Merge "Adding get_buf_from_mv() function." 2014-03-24 15:52:09 -07:00
Jim Bankoski
1a8a37ceda vp9_variance.h header static version converted to global
Change-Id: I8d02cdb5e7fef02b8c95413dfd239fa51a8c7777
2014-03-24 09:55:58 -07:00
Dmitry Kovalev
6e9ca1ecef Adding get_buf_from_mv() function.
Change-Id: I21aff45546778b8393e2edf2d810448dec1f4cdb
2014-03-21 15:57:31 -07:00
Dmitry Kovalev
cac2cf886c Cleaning up vp9_mcomp.c.
Change-Id: I24a3f0f215d04fb8888c303642dd5663cc593317
2014-03-17 11:41:20 -07:00
Dmitry Kovalev
65dc919103 Merge "Cleaning up vp9_diamond_search_sad_c & vp9_diamond_search_sadx4." 2014-03-17 11:23:49 -07:00
Deb Mukherjee
e39ecfaa98 Preliminary code for variance based paritioning
Brings back most of Jim's previous patch for choosing
partitioning based on variance while making it compatible
with the current state of the code. Also adds a
nonrd_use_partition() function to recursively encode  for any
arbitrary sb_type decisions within a 64x64 block; and
includes some refactoring.

Currently, when the VAR_BASED_PARTITIONING mode is turned on
for speed 7, there is a 10+% speed-up observed.

Experiments/improvements with this new partitioning method
will be conducted subsequently.

Change-Id: Ie6f43bfbde30583e941f450bf07c3b48828c9571
2014-03-12 15:06:20 -07:00
Deb Mukherjee
44a203f5cd Merge "Support for a fast diamond search" 2014-03-10 09:53:21 -07:00
Deb Mukherjee
bead2e8fb4 Support for a fast diamond search
Adds a fast diamond search which is about 5% faster than FAST_HEX
with only a 0.1% drop in psnr when turned on for both speeds 5 and 7.
This search is turned on for speed 7.

Change-Id: I497630aa88a5148926086bb3038e7975e5f4eb98
2014-03-07 17:15:09 -08:00
Dmitry Kovalev
ee7d26d50c Cleaning up vp9_get_mvpred_var().
Change-Id: Ic535f0a1c2501c1af143237af3b2b51b4b4980f4
2014-03-06 16:27:15 -08:00
Dmitry Kovalev
a2d846cac4 Cleaning up vp9_diamond_search_sad_c & vp9_diamond_search_sadx4.
Change-Id: I4639cdc29d17ef13cabb774b455295454f1d02ae
2014-03-06 15:00:01 -08:00
Deb Mukherjee
26fa8cad01 Fixes static analysis issues
Removes some unused variables and assignments

Change-Id: I228f43d658ce50237d65ce8dc8ab2ccf0a0b21ca
2014-03-06 11:07:06 -08:00
Dmitry Kovalev
0b6440ce02 Cleaning up vp9_refining_search_sadx4().
Change-Id: I3ed0a95645a66be069ce92a1fad8083a87d01001
2014-03-05 10:46:46 -08:00
Dmitry Kovalev
91e5e5a680 Merge "Cleaning up vp9_init_dsmotion_compensation()." 2014-03-03 17:27:51 -08:00
Deb Mukherjee
b80020d4db Refactoring motion search libs
The core motion estimation fucntions all return sad now consistently.
The only exception is vp9_full_pixel_diamond(), however the core diamond
and refining search routines called from vp9_full_pixel_diamond() also
return SAD. If variance of pred error + mv cost is desired it must be
calculated explicitly outside these functions. For very fast encoding,
hopefully this will eliminate some redundant computations.

Also suggests reimplementing FAST_HEX with the vp9_pattern_search
framework. It is not exactly the same as the existing FAST_HEX, but
performance is slightly better and speed is very similar. Enables
removing a lot of duplicate code.

Change-Id: I152736393438c25bdf7e96b37cbb8ce330f4f94a
2014-03-03 15:13:59 -08:00
Dmitry Kovalev
f1a6cdbf8d Cleaning up vp9_init_dsmotion_compensation().
Change-Id: Id5b56c7bc5671859f54ae69693064804d3fadd98
2014-02-28 10:39:31 -08:00
Dmitry Kovalev
7bca32a6a3 Merge "Changing vp9_full_search_sad{, x3, x8} signatures." 2014-02-25 10:51:17 -08:00
Dmitry Kovalev
55214924ac Merge "Cleaning up vp9_refining_search_sad() function." 2014-02-25 10:44:19 -08:00
Yunqing Wang
360c80cfb9 Use fast HEX search in real time mode
In good quality mode motion search, the best matches are normally
found after searching in a large area. In real time mode, to make
encoding fast, a center-biased fast HEX search is used, which
converges quickly most of the time. A 4-point diamond search is
also carried out as the following refining search, which gives more
precise results, and maintains good motion search quality.

At speed 5, the borg test on rtc set showed an overall PSNR loss of
0.936%. The encoding speed gain is 4% - 5%.

Change-Id: I42cd68bb56a09ca1b86293c99d5f7312225ca7ae
2014-02-18 14:35:16 -08:00
Dmitry Kovalev
35092585d8 Cleaning up vp9_refining_search_sad() function.
Change-Id: Ife6131d31882177953a22a69d25adc35940429ed
2014-02-17 14:51:58 +01:00
Dmitry Kovalev
36420009ea Changing vp9_full_search_sad{, x3, x8} signatures.
Passing block MV pointer instead of block index into
vp9_full_search_sad{, x3, x8} functions.

Change-Id: Ica07356633471c2c8f81b583a7aeba85a436bafb
2014-02-17 14:24:57 +01:00
Jim Bankoski
e727bcf7ce Merge "vp9_mcomp.c visual studio warnings resolved" 2014-02-14 09:08:34 -08:00
Jim Bankoski
6025ec7901 vp9_mcomp.c visual studio warnings resolved
Change-Id: I7546cc8db4decbcd094ad76c64385b97666331b0
2014-02-14 06:35:36 -08:00
Dmitry Kovalev
fde1b6d7f2 Using MV instead of int_mv inside vp9_full_pixel_diamond().
Change-Id: If33a5a12c4025d9b5ec863dfccea7ee70f800665
2014-02-13 11:23:05 -08:00
Jim Bankoski
d2cd088cb7 Merge "vp9_mcomp.c static analysis issues" 2014-02-13 06:20:16 -08:00
Dmitry Kovalev
50712fcaa9 Adding consts to mv search function arguments.
Change-Id: Ie79114bba4f0cea55d9f701e20d2be2017630f3b
2014-02-12 14:28:23 -08:00
Jim Bankoski
f324f0c54e vp9_mcomp.c static analysis issues
Change-Id: I8c0023e3f03297993431e21158ee3821af75f82d
2014-02-12 11:54:47 -08:00
Dmitry Kovalev
dcb48761cb Cleaning up vp9_full_search_sad_c() function.
Change-Id: Ief941ff75c5e25f3288326a162b1bd0c0f17a779
2014-02-11 17:24:22 -08:00
Dmitry Kovalev
c1749e14f6 Adding is_mv_in() function.
Change-Id: I9bdb18c5cde386d3caaa2c2846649cfbda609927
2014-02-04 16:58:33 -08:00
Dmitry Kovalev
ae1c298733 Inlinging get_next_chkpts macro in vp9_mcomp.c.
Change-Id: I38ca35dccfb9051d08b026d75ba2f3a35adb0259
2014-02-03 14:26:36 -08:00
Dmitry Kovalev
3b3f1d8303 Cleaning up vp9_mcomp.{c, h}.
Change-Id: Ib71d9ed3f98e9468ad951bdc24c9ab565216eb38
2014-01-31 18:11:41 -08:00
Dmitry Kovalev
c2ca97caaf Merge "Cleaning up motion compensation code." 2014-01-31 17:33:40 -08:00
Dmitry Kovalev
abf37f1d6c Inlining IFMVCV macro.
Change-Id: I5173f996612e410d9cd495df9414d194b1ab18f3
2014-01-24 11:13:39 -08:00
Dmitry Kovalev
6c98df29e4 Cleaning up motion compensation code.
Change-Id: I74cf028e8c732cd0dbc070326152d3085b824a80
2014-01-23 17:15:30 -08:00
Dmitry Kovalev
5f75fda9e9 Merge "Cleaning up vp9_refining_search_sad() function." 2014-01-22 17:15:22 -08:00
James Zern
eb4cb43397 Merge "Enforce effective motion vector search range" 2014-01-22 16:50:30 -08:00
Jingning Han
20cf22a128 Enforce effective motion vector search range
This commit explicitly enforces the effective motion vector range
in the motion search stage. The range needs to be the intersection
of UMV border, effective absolute motion vector value range, and
the target search area.

Change-Id: I1cf7c563e02b1086040dad6c1f4f6be1538635a6
2014-01-21 12:03:18 -08:00
Dmitry Kovalev
575a083bcc Merge "Replacing macros with inline functions." 2014-01-21 10:17:29 -08:00
Dmitry Kovalev
d8bfe9e24c Cleaning up vp9_refining_search_sad() function.
Change-Id: I660b53da8ebf3049832ce8a10721051c4e0ebb00
2014-01-17 15:20:28 -08:00
Dmitry Kovalev
67246764a2 Replacing macros with inline functions.
Change-Id: Ibde8fb45594cd259fc3281d7874de8fc877fd4f2
2014-01-16 18:00:44 -08:00
Dmitry Kovalev
55d0230221 Cleaning up vp9_refining_search_8p_c() function.
Change-Id: Icf3b3dd96d7e133a4ad7260cd95288f6217998a6
2014-01-16 16:05:36 -08:00
Dmitry Kovalev
80aae67f44 Making motion search code more compact.
Change-Id: I04d6a006eaf2676f2d8d1d96b87f69e081969382
2014-01-13 15:24:06 -08:00
Dmitry Kovalev
2344e3a2e1 Merge "Converting CHECK_POINT macro to inline function." 2014-01-03 18:02:10 -08:00
Dmitry Kovalev
6e8c103620 Merge "Replacing int_mv with MV." 2014-01-03 17:38:28 -08:00
Dmitry Kovalev
513666e78c Converting CHECK_POINT macro to inline function.
Change-Id: I08533d59a78346dac30a7dcbc12146f23ef88bbc
2014-01-03 16:16:28 -08:00
Dmitry Kovalev
ed2fba2a79 Merge "Replacing CHECK_BOUNDS macro with inline check_bounds function." 2014-01-03 15:57:36 -08:00
Dmitry Kovalev
672c355a26 Replacing int_mv with MV.
Change-Id: Ifd432fa3741ba47102d298e0b348eb00f5a9ce53
2014-01-03 11:48:07 -08:00
Yunqing Wang
a7248a04b7 Merge "Remove a unused sub-pixel search" 2013-12-27 14:05:38 -08:00
Dmitry Kovalev
987810ad95 Removing vp9_findnearmv.{h, c} files.
Moving all code from that files to vp9_mvref_common.{h, c}.

Change-Id: Ibc4afcb8cea6847166ff411130e93611ebe63b20
2013-12-19 17:39:57 -08:00
Yunqing Wang
6ff4f19269 Remove a unused sub-pixel search
The original iterative search was replaced by subpel_tree search,
and was not used anymore.

Change-Id: I998b38e1cb0ee359a08b2410d0766dbf183ab071
2013-12-19 11:20:56 -08:00
Dmitry Kovalev
095b03231c Using MV struct instead of int_mv union in encoder (2).
Change-Id: I068345f722a7116e3119927295ad23a28d3066a0
2013-12-13 17:20:40 -08:00
Dmitry Kovalev
25da21b14e Using MV struct instead of int_mv union in encoder.
Change-Id: I8b81a3e4b4fa530a654c28d9c136afa0c1d379fd
2013-12-13 15:24:48 -08:00
Dmitry Kovalev
5556090dd0 Replacing CHECK_BOUNDS macro with inline check_bounds function.
Change-Id: I8437113da256088576bf66401186e35125089c03
2013-12-11 15:00:36 -08:00
Jingning Han
9d6e7ce12d Rename clamp_mv_min_max to set_mv_search_range
This function sets the motion search range limit. Rename it to be
more informative.

Change-Id: I2e8e01073dcb99c9bea9c9acd0a61d672d615444
2013-12-11 14:00:08 -08:00
Guillaume Martres
0102f1d5ec avoid crash when using --best on cpus with SSE3 (but not SSE4) support
Change-Id: Ie100114a01b8b4da7248603c40676792cd06b32a
2013-12-10 18:17:10 +01:00
Jingning Han
44627a9677 Clean-ups in diamond_search_sad
Use the pre-fetched stride value.

Change-Id: I8d4aaedab7a83a4870be51c110da27e55ca3bc85
2013-12-06 14:25:20 -08:00
Jingning Han
b295092b8f Full range motion search for regular block sizes
Add a full range motion search for regular block sizes. This runs
exhaustive search within the given reference area. This commit further
optimizes the search process by combining 4 points test into one
pipeline, which gives 30% speed-up as compared to run each individual
point at a time.

This full range search serves as a best possible motion search reference.
When replacing the diamond search with full range search, the speed 0
runtime of bus CIF at 2000 kbps goes from 153872ms to 623051ms. The
compression performance compared to speed 0 setting gains 0.585% for
derf set.

Change-Id: Ieef1225216b0b86b4ac4872fa7fb9e18bf2eabb3
2013-12-06 12:24:53 -08:00
Jingning Han
bbe68fbd2a Constrain encoder motion search range
Explicitly constrain the upper limit of motion search range (in the
unit of full pixel) to be [-1023, +1023]. It is intended to control
the effective motion search range for 4K sequences.

Change-Id: I645539c70885eec0f155781f439d97d333336e88
2013-11-18 11:19:45 -08:00
Dmitry Kovalev
2726f383cd Adding allow_hp as an argument to mv search functions.
Making this change in order to move allow_high_precision_mv field
from MACROBLOCKD structure to VP9_COMMON (because it is a frame level
flag).

Change-Id: I1d006ba36d938e0caf4d40fa051e2e38df9c1108
2013-10-17 14:02:04 -07:00
Jim Bankoski
b79b7c354d cpplint issues resolved in vp9_mcomp.c
Change-Id: I2c2f83f4dfa2782fc6b0aa6db3ba2c4e6e423ffa
2013-10-05 16:44:40 -07:00
Dmitry Kovalev
8266da1cd1 Moving from int_mv* to MV* (3).
Change-Id: I9795d0937bc07793c13d067281995e0750f694d9
2013-09-25 16:44:19 -07:00
Dmitry Kovalev
b87696ac37 Moving from int_mv* to MV* (2).
Updating fractional_mv_step_fp and fractional_mv_step_comp_fp function
types.

Change-Id: I601c4378bc39ac3ffd4e295d9cbd8e1f74829d46
2013-09-24 12:48:12 -07:00
Dmitry Kovalev
30888742f4 Merge "Moving from int_mv to MV." 2013-09-24 12:25:56 -07:00
Yaowu Xu
838eae3961 Correct 3 step search site initialziation
39c7b01d accidently reverted the row/col initialization, which broke
mv clamps, which is dependent on the sites for valid motion vector
range. This commit fixed the issue.

Change-Id: Ibcce0226e0360b1ef483fe760b2e33f1af4bf494
2013-09-23 16:11:49 -07:00
Dmitry Kovalev
e51e7a0e8d Moving from int_mv to MV.
Converting vp9_mv_bit_cost, mv_err_cost, and mvsad_err_cost
functions for now.

Change-Id: I60e3cc20daef773c2adf9a18e30bc85b1c2eb211
2013-09-20 13:52:43 +04:00
Dmitry Kovalev
39c7b01d3c Cleanup in vp9_init3smotion_compensation.
Change-Id: Ie47f53e76bc9530475c8c6d24e9b7a5a0189de56
2013-09-20 12:54:14 +04:00
Yaowu Xu
014acfa2af fix integer overflow errors
Change-Id: I76f440a917832c02d7a727697b225bac66b99f56
2013-09-19 08:14:26 -07:00
Dmitry Kovalev
72fd127f8c Removing redundant code from vp9_mcomp.c.
Replacing ((1 << MV_MAX_BITS) - 1) with MV_MAX, adding const
qualifiers, reusing computed values.

Change-Id: I7b46d47f6c644b079d9c3478116a9de465a9baec
2013-09-18 13:11:38 +04:00
Scott LaVarnway
ac6093d179 New mode_info_context storage -- undo revert
mode_info_context was stored as a grid of MODE_INFO structs.
The grid now constists of pointers to MODE_INFO structs.  The
MODE_INFO structs are now stored as a stream (decoder only),
eliminating unnecessary copies and is a little more cache
friendly.

Change-Id: I031d376284c6eb98a38ad5595b797f048a6cfc0d
2013-09-11 13:45:44 -04:00
James Zern
54a03e20dd Revert "New mode_info_context storage"
This reverts commit dae17734ec

Encode crashes, leaks and increases integer overflow errors.

Change-Id: I595aa2649bb8d0b6552ff91652837a74c103fda2
2013-09-09 13:37:01 -07:00
Scott LaVarnway
dae17734ec New mode_info_context storage
mode_info_context was stored as a grid of MODE_INFO structs.
The grid now constists of a pointer to a MODE_INFO struct and
a "in the image" flag.  The MODE_INFO structs are now stored
as a stream, eliminating unnecessary copies and is a little
more cache friendly.

For the test clips used, the decoder performance improved
by ~4.3% (1080p) and ~9.7% (720p).

Patch Set 2: Re-encoded clips with latest. Now ~1.7% (1080p)
and 5.9% (720p).

Change-Id: I846f29e88610fce2523ca697a9a9ef2a182e9256
2013-09-06 12:33:34 -04:00
Yaowu Xu
f70330a906 fix a bug when null function pointer is used.
For certain partition size, the function poniter may not be intialized
at all. The patch prevent the call if the pointer is not set.

Change-Id: I78b8c3992b639e8799a16b3c74f0973d07b8b9ac
2013-08-19 17:16:12 -07:00
Dmitry Kovalev
1aedfc992a Using MV* instead of int_mv* as argument of vp9_clamp_mv_min_max.
Change-Id: I3c45916a9059f11b41e9d798e34ffee052969a44
2013-08-12 13:56:04 -07:00
Dmitry Kovalev
6a8ec3eac2 General code cleanup.
Removing redundant parenthesis and curly braces. Combining declarations
with initializations. Adding useful intermediate variables instead of
recalculating expressions every time.

Change-Id: I00106f404afd60bfc189905b0fded881684f941a
2013-08-08 21:12:34 -07:00
Deb Mukherjee
1ba91a84ad Adds a new subpel motion function
Adds a new subpel motion estimation function that uses a 2-level
tree-structured decision tree to eliminate redundant computations.
It searches fewer points than iterative search (which can search
the same point multiple times) but has the same quality roughly.

This is made the default setting at speeds 0 and 1, while at
speed 2 and above only a 1-level search is used.

Also includes various cleanups for consistency and redundancy removal.

Results:
derf: +0.012% psnr
stdhd: +0.09% psnr
Speedup of about 2-3%

Change-Id: Iedde4866f5475586dea0f0ba4cb7428fba24eee9
2013-08-08 11:41:49 -07:00
Deb Mukherjee
71b43b0ff0 Clean ups of the subpel search functions
Removes some unused code and speed features, and organizes the
interfaces for fractional mv step functions for use in new speed
features to come.

In the process a new speed feature - number of iterations per
step during the subpel search - is exposed.

No change when this parameter is set as the original value of 3.

Results:
subpel_iters_per_step = 3: baseline
subpel_iters_per_step = 2: psnr -0.067%, 1% speedup
subpel_iters_per_step = 1: psnr -0.331%, 3-4% speedup

Change-Id: I2eba8a21f6461be8caf56af04a5337257a5693a8
2013-08-06 17:23:50 -07:00
Deb Mukherjee
15b5a6a2c7 Flexible support for various pattern searches
Adds a few pattern searches to achieve various tradeoffs
between motion estimation complexity and performance.
The search framework is unified across these searches so that a
common pattern search function is used for all. Besides it will
be easier to experiment with various patterns or combinations
thereof at different scales in the future.

The new pattern search is multi-scale and is capable of using
different patterns at different scales.

The new hex search uses 8 points at the smallest scale
and 6 points at other scales.
Two other pattern searches - big-diamond and square are
also added. Big diamond uses 4 points at the smallest scale and
8 points in diamond shape at the larger scales.
Square is very similar conceptually to the default n-step search
but is somewhat faster since it keeps only one survivor across
all scales.

Psnr/speed-up results on derf300:

hex: -1.6% psnr%, 6-8% speed-up
big-diamond: -0.96% psnr, 4-5% speedup
square: -0.93% psnr, 4-5% speedup

Change-Id: I02a7ef5193f762601e0994e2c99399a3535a43d2
2013-08-06 11:56:39 -07:00
Dmitry Kovalev
86053d3ae2 Cleanups around allow_high_precision_mv flag.
Change-Id: Ic07f5f8ffeaedd5b7513b464871f83afc82dcd5c
2013-08-02 11:21:16 -07:00
Dmitry Kovalev
b621e2d72e Nice looking motion vector clamping functions.
Removing assign_and_clamp_mv function, making implementation of clamp_mv
and clamp_mv2 more clear and consistent.

Change-Id: Iecd08e1c1bf0379f8314ebe01811f8253f4ade58
2013-08-01 13:40:26 -07:00
Dmitry Kovalev
500ade243a Removing unused "ishp" arguments.
Using different variable names "allow_hp" and "use_hp" instead of "usehp".

Change-Id: I0cd5996ddeb46bd754473b680a993c0aaf8eb879
2013-07-31 11:27:53 -07:00
Deb Mukherjee
a1e2d50be9 Diamond search change to accelerate movement
Optional change in diamond search to continue in the best move
direction until that move turns worse.

This is still WIP since the exact way the new method is to be used is
under investigation. One option is to make it an option in diamond
search and use it only when motion is large.

Overall slightly positive on derfraw300 +0.02%, stdhdraw +0.13%,
but works a lot better for high motion sequences (ex. football : +1%).

Change-Id: If88e01a6021daa0cda934680cdc70be1ee04f798
2013-07-22 11:19:15 -07:00
Paul Wilkins
9f3ab83486 Auto adapt step size feature.
Also tweaks to other features and experiments with
what is on and off at different speed settings.

Change-Id: I3e1d0be0d195216bf17c2ac5df67f34ce0b306b2
2013-06-26 19:48:39 +01:00
Paul Wilkins
e606cac046 Change meaning of cpi->sf.first_step and rename.
Renamed cpi->sf.first_step to cpi->sf.reduce_first_step_size
and changed its meaning such that it is a delta applied to
reduce the default first step size (>> x) in the motion search
rather than an absolute value.

The default first step size is already changed according to the image
dimensions (smaller for smaller images). cpi->sf.reduce_first_step_size
now applies a further correction from the default.

Change-Id: Ia94e08bc24c67b604831f980909af7e982fcd16d
2013-06-26 17:04:06 +01:00
Ronald S. Bultje
c24d922396 Add averaging-SAD functions for 8-point comp-inter motion search.
Makes first 50 frames of bus @ 1500kbps encode from 3min22.7 to 3min18.2,
i.e. 2.3% faster. In addition, use the sub_pixel_avg functions to calc
the variance of the averaging predictor. This is slightly suboptimal
because the function is subpixel-position-aware, but it will (at least
for the SSE2 version) not actually use a bilinear filter for a full-pixel
position, thus leading to approximately the same performance compared to
if we implemented an actual average-aware full-pixel variance function.
That gains another 0.3 seconds (i.e. encode time goes to 3min17.4), thus
leading to a total gain of 2.7%.

Change-Id: I3f059d2b04243921868cfed2568d4fa65d7b5acd
2013-06-25 12:57:28 -07:00
Ronald S. Bultje
4dc70fa7f9 Don't re-allocate comp_pred buffers for each call to comp motion search.
Instead, just allocate a few bytes on the stack, this is 4k, which isn't
all that much.

Change-Id: I82af6ee89e6ed01faaa23ff891ee7ced76df8c16
2013-06-24 14:05:13 -07:00
Dmitry Kovalev
87e1fa7627 Renaming 'nmv' to 'mv' for several functions.
Change-Id: I183a38997a9d01e4a1b869e92509f6915216fa09
2013-06-18 18:28:10 -07:00
Paul Wilkins
0b713f8c18 Merge CONFIG_COMP_INTER_JOINT_SEARCH.
Merge this experiment so that it is under a speed feature
flag not a configuration flag.

Change-Id: I536f7f125a4ff5149bb3a64f791e835c324535fd
2013-05-22 11:23:31 +01:00
Paul Wilkins
6ff3eb1647 New inter mode context.
This patch creates a new inter mode contest that avoids
a dependence on the reconstructed motion vectors from
neighboring blocks. This was a change requested by
a hardware vendor to improve decode performance.

As part of this change I have also made some modifications
to stats output code (under a flag) to allow accumulation of
inter mode context flags over multiple clips

Some further changes will be required to accommodate the
deprecation of the split mv mode over the next few days.

Performance as stands is around -0.25% on derf and
std-hd but up on the YT and YT-HD sets. With further tuning
or some adjustment to the context criteria it should be
possible to make this change broadly neutral.

Change-Id: Ia15cb4470969b9e87332a59c546ae0bd40676f6c
2013-05-16 12:09:19 +01:00
Yunqing Wang
9f5811c2da Add joint motion search in comp_inter_inter mode(experiment)
In current code, motion vectors got from single prediction mode are used
in compound prediction mode directly. These motion vectors may not give
accurate prediction since they are searched independently. In this patch,
we took Pascal's suggestion, and did joint motion search in compound
prediction mode to find better motion vectors in this situation.
Test results:
Overall PSNR: 0.570%(derf), 0.918%(stdhd);
SSIM: 0.572%(derf), 1.009%(stdhd);

The encoder is a little slower. This can be improved since some c
code is used in motion search.

Change-Id: Ib30c9240f6c56c9b070867b4ca89412a76d9f3c6
2013-05-10 10:15:43 -07:00
John Koleszar
bb41ab4a0c Remove BLOCKD structure
All members can be referenced from their per-plane counterparts, and
removes assumptions about 24 blocks per macroblock.

Change-Id: I7ff2fa72d22c29163eb558981c8193765a8113d9
2013-04-26 10:35:54 -07:00
Scott LaVarnway
57f180b388 Removed bmi from blockd
This originally was "Removed update_blockd_bmi()".  Now,
this patch removed bmi from blockd and uses the bmi found
in mode_info_context.  Eliminates unnecessary bmi copies between
blockd and mode_info_context.

Change-Id: I287a4972974bb363f49e528daa9b2a2293f4bc76
2013-04-26 10:19:43 -04:00
John Koleszar
a99e1aa8ca Remove predictor pointers from BLOCKD
Access these members from MACROBLOCKD instead.

Change-Id: I2574622e577bb9feede47f6b7ccbb11f3e928ca8
2013-04-25 12:04:07 -07:00
John Koleszar
4bd0f4f646 Remove BLOCK structure
All members can be referenced from their per-plane counterparts, and
removes assumptions about 24 blocks per macroblock.

Change-Id: I593fb0715e74cd84b48facd1c9b18c3ae1185d4b
2013-04-25 11:33:17 -07:00
Dmitry Kovalev
3e01ca6a19 Fixing rounding inside vp9_mv_bit_cost function.
Change-Id: I7209a05919162a8155520bc543658ddb69ba12ce
2013-04-18 15:37:11 -07:00
Dmitry Kovalev
0db175ffed Changing argument type of vp9_get_mv_joint from MV to MV*.
Change-Id: I28c3026946fc1bde7074e6e0198da93bb0d75dfe
2013-04-17 11:21:28 -07:00
Dmitry Kovalev
ec299e2092 Encoder code cleanup.
Removing duplicated code from vp9_encodemv.c and reusing ROUND_POWER_OF_TWO
macro definitions.

Change-Id: I9caf0c17f761ada7905cb99a3e2a31f871fef0f9
2013-04-11 11:08:00 -07:00
Paul Wilkins
d8ffee4526 Changes to rd error_per_bit calculation.
Specifically changes to retain more precision
especially at low Q through to the point of use.

Change-Id: Ief5f010f2ca4daaabef49520e7edb46c35daf397
2013-03-18 23:07:51 +00:00
Jingning Han
2a5278bdbd Extend diff MV limit from +/-256 to +/-1024
Increase the motion search range by 4x. Change MV_CLASS tree of the
entropy coding to allow two additional mv classes to cover the
extended motion vector limit. The codec determines the effective
motion search range conditioned on the actual frame dimension.

It provides coding gains:

stdhd 0.39%
yt    0.56%
hd    0.47%

Major coding performance gains are packed in several sequences with
intense motion activities, e.g., ped_1080p gains 7% at high bit-rates,
and on average 3%.

TODO: Need to further tune the rate control and motion search units.

Change-Id: Ib842540a6796fbee5a797809433ef6a477c6d78d
2013-03-08 10:04:36 -08:00
Dmitry Kovalev
dcbdda8e15 Code cleanup.
Lower case variable names, converting while loops to for loops.

Change-Id: Ic3b973391eef7472a99d18d02fe79cfef5e04e62
2013-02-28 14:40:20 -08:00
Jim Bankoski
714aa9f3c0 this commit converts all sad ptrs to uint32
sse4_1 code used uint16_t for returning sad, but that
won't work for 32x32 or 64x64.   This code fixes the
assembly for those and also reenables sse4_1 on linux

Change-Id: I5ce7288d581db870a148e5f7c5092826f59edd81
2013-02-28 08:46:35 -08:00
Paul Wilkins
ef01b956d8 Entropy stats output code.
Fixes to make Entropy stats code work again

Change-Id: I62e380481a4eb4c170076ac6ab36f0c2b203e914
2013-02-20 14:33:19 +00:00
John Koleszar
7ca517f755 Replace as_mv struct with array
Replace as_mv.{first, second} with a two element array, so that they
can easily be processed with an index variable.

Change-Id: I1e429155544d2a94a5b72a5b467c53d8b8728190
2013-02-08 20:23:35 -08:00
Ronald S. Bultje
4455036cfc Merge superblocks (32x32) experiment.
Change-Id: I0df99742029834a85c4933652b0587cf5b6b2587
2013-01-08 12:54:45 -08:00
John Koleszar
879cb7d962 Merge vp9-preview changes into experimental branch
Incorportate vp9-preview changes by merging master branch into experimental.

Conflicts:
	test/test.mk
	vp9/common/vp9_filter.c
	vp9/common/vp9_idctllm.c
	vp9/common/vp9_invtrans.h
	vp9/common/vp9_mbpitch.c
	vp9/common/vp9_rtcd_defs.sh
	vp9/common/vp9_systemdependent.h
	vp9/common/vp9_type_aliases.h
	vp9/common/x86/vp9_asm_stubs.c
	vp9/common/x86/vp9_subpixel_mmx.asm
	vp9/decoder/vp9_decodframe.c
	vp9/decoder/vp9_dequantize.c
	vp9/decoder/vp9_dequantize.h
	vp9/decoder/vp9_onyxd_int.h
	vp9/encoder/vp9_bitstream.c
	vp9/encoder/vp9_encodeframe.c
	vp9/encoder/vp9_rdopt.c

Change-Id: I17f51c3666d1b59cf1a699f87607cbc5d30a87c5
2013-01-08 10:19:59 -08:00
Ronald S. Bultje
c3941665e9 64x64 blocksize support.
3.2% gains on std/hd, 1.0% gains on hd.

Change-Id: I481d5df23d8a4fc650a5bcba956554490b2bd200
2013-01-05 18:20:25 -08:00
John Koleszar
5ebe94f9f1 Build fixes to merge vp9-preview into master
Various fixups to resolve issues when building vp9-preview under the more stringent
checks placed on the experimental branch.

Change-Id: I21749de83552e1e75c799003f849e6a0f1a35b07
2012-12-26 11:21:09 -08:00
Ronald S. Bultje
4cca47b538 Use standard integer types for pixel values and coefficients.
For coefficients, use int16_t (instead of short); for pixel values in
16-bit intermediates, use uint16_t (instead of unsigned short); for all
others, use uint8_t (instead of unsigned char).

Change-Id: I3619cd9abf106c3742eccc2e2f5e89a62774f7da
2012-12-18 15:31:19 -08:00
Ronald S. Bultje
4d0ec7aacd Consistently use get_prob(), clip_prob() and newly added clip_pixel().
Add a function clip_pixel() to clip a pixel value to the [0,255] range
of allowed values, and use this where-ever appropriate (e.g. prediction,
reconstruction). Likewise, consistently use the recently added function
clip_prob(), which calculates a binary probability in the [1,255] range.
If possible, try to use get_prob() or its sister get_binary_prob() to
calculate binary probabilities, for consistency.

Since in some places, this means that binary probability calculations
are changed (we use {255,256}*count0/(total) in a range of places,
and all of these are now changed to use 256*count0+(total>>1)/total),
this changes the encoding result, so this patch warrants some extensive
testing.

Change-Id: Ibeeff8d886496839b8e0c0ace9ccc552351f7628
2012-12-12 10:01:19 -08:00
Jim Bankoski
c67873989f fixed includes to be fully specified
Change-Id: Ia1cce221f8511561b9cbd8edb7726fbc286ff243
2012-11-28 10:53:17 -08:00
John Koleszar
fcccbcbb39 Add vp9_ prefix to all vp9 files
Support for gyp which doesn't support multiple objects in the same
static library having the same basename.

Change-Id: Ib947eefbaf68f8b177a796d23f875ccdfa6bc9dc
2012-11-27 14:12:30 -08:00