Commit Graph

8075 Commits

Author SHA1 Message Date
James Zern
31509af247 vp9_encodeframe: delete auto_partition_range()
unused since:
1f00a9b Fix choose_partitioning threshold setup for speed -5

Change-Id: I51952c0a1be3e6e0aa36ff2ffcfbbea60a505960
2015-06-12 17:57:37 -07:00
James Zern
7ea431df98 vp9_mcomp: don't mark setup_center_error() inline
this function is a bit too involved for the hint; avoids a -Winline
warning

Change-Id: Ib82e424764aa78b37ddb94116e2b009a6de31d35
2015-06-12 17:56:33 -07:00
James Zern
471302a07b vp9_encoder: hide adjust_image_stat()
this function is only needed with CONFIG_INTERNAL_STATS

Change-Id: I9c8d16cb9069dd8370f8b30329933c0d97f6d0aa
2015-06-12 17:55:08 -07:00
James Zern
a4bb5f2a29 Merge "decode_tiles_mt: remove incorrect TODO" 2015-06-12 22:13:38 +00:00
James Zern
dc0f0f1cf2 Merge changes I342075eb,I222eaa4e
* changes:
  vp9_decodeframe: simplify init_read_bit_buffer signature
  vp9_decode_frame: remove explicit read_bit_buffer init
2015-06-12 22:13:28 +00:00
James Zern
e21c1eab9d Merge changes Iedb5b6a3,Iaea98508,I36580cea,Ia0574320
* changes:
  vp9_decodeframe.h: remove unused prototype
  vp9_decodeframe: move public funcs to end of file
  vp9_decodeframe: reorder some functions
  vp9_decodeframe: hide vp9_dec_build_inter_predictors_sb
2015-06-12 22:12:30 +00:00
Jingning Han
d8985f5360 Merge "Fix potential overflow issue in hadamard_16x16()" 2015-06-12 19:03:21 +00:00
Jingning Han
bea691b5c9 Merge "Add dynamic range comment to hadamard_8x8()" 2015-06-12 18:43:41 +00:00
Parag Salasakar
ecbbef6b67 Merge "mips msa vp9 filter by weight optimization" 2015-06-12 18:30:11 +00:00
Jingning Han
176c291d9c Fix potential overflow issue in hadamard_16x16()
This commit fixes a potential integer overflow issue in function
hadamard_16x16. It adds corresponding dynamic range comment.

Change-Id: Iec22f3be345fb920ec79178e016378e2f65b20be
2015-06-12 10:56:18 -07:00
Jingning Han
4f52d49f1e Add dynamic range comment to hadamard_8x8()
Add comment to assist SIMD optimization.

Change-Id: I300d5a848e6e9947e451de2a871a88940703fc9f
2015-06-12 10:39:49 -07:00
Yunqing Wang
254a4c033c Merge "Allocate tile data adaptively to accommodate the frame size increase" 2015-06-12 15:49:40 +00:00
Parag Salasakar
fbac961b47 mips msa vp9 filter by weight optimization
filter by weight - average improvement ~2x-3x

Change-Id: I4832033335d339cdafdce697f07ce3e643920057
2015-06-12 12:06:42 +05:30
James Zern
e2b52f6f01 vp9_filter: make all filter tables static
these are returned via vp9_get_interp_kernel()

Change-Id: I45ed75e5b1515c4f5be9212759dcb50a456b5548
2015-06-11 15:15:52 -07:00
James Zern
33b3953c54 vp9_filter: restore vp9_bilinear_filters alignment
the declaration containing the alignment in vp9_filter.h was removed in:
eb88b17 Make vp9 subpixel match vp8

fixes a crash in 32-bit builds

Change-Id: I9a97e6b4e8e94698e43ff79d0d8bb85043b73c61
2015-06-11 15:15:25 -07:00
Scott LaVarnway
0fbc277746 Merge "inline vp9_get_segdata()" 2015-06-11 19:48:19 +00:00
Yunqing Wang
2c838ede68 Allocate tile data adaptively to accommodate the frame size increase
If the frame size increases, the tile data buffer needs to be
re-allocated according to the number of tiles existing in current
frame. This patch makes the multi-tile encoding work in spatial
SVC usage case, and partially solved WebM issue 1018.

Change-Id: I1ad6f33058cf5ce6f60ed5024455a709ca80c5ad
2015-06-11 11:30:18 -07:00
Scott LaVarnway
cca866f578 inline vp9_get_segdata()
and change name.

Change-Id: I706645cf9d9dc04f1b3b6ac80df80edb7f101854
2015-06-11 09:52:00 -07:00
Marco
2aa67ce20f Move adjustment of some CR parameters to existing function.
Refactor/no change in behavior.

Change-Id: Idb3c55b1304feaf689b90403f79bc96dba26f060
2015-06-11 08:31:03 -07:00
Scott LaVarnway
a49c701529 Merge "inline vp9_segfeature_active()" 2015-06-11 12:29:45 +00:00
Scott LaVarnway
42c0b1b1f1 inline vp9_segfeature_active()
and changed name.

Change-Id: Ie023ca66cc2c823032f58d4faeb53fd1863c94f3
2015-06-11 04:20:55 -07:00
Parag Salasakar
c7489f4815 Merge "mips msa vp9 intra-pred optimization" 2015-06-11 03:31:49 +00:00
James Zern
15c839f563 decode_tiles_mt: remove incorrect TODO
all allocated workers are used, the final one in the main thread.

Change-Id: I04647d4fb2e01d6d404790e7899515289047f553
2015-06-10 15:56:51 -07:00
James Zern
b105414118 vp9_decodeframe: simplify init_read_bit_buffer signature
Change-Id: I342075eb5a4ba2a85d36d47ae52d1f3476039e31
2015-06-10 15:53:30 -07:00
James Zern
92146eb8ac vp9_decode_frame: remove explicit read_bit_buffer init
this is done by init_read_bit_buffer()

Change-Id: I222eaa4e9758ff9f7e1e4122106c5c4652ffa99c
2015-06-10 15:50:54 -07:00
James Zern
dca319040c vp9_decodeframe.h: remove unused prototype
vp9_init_dequantizer() was deleted in:
bdd249b Optimize the dequantization process on decoder side.

Change-Id: Iedb5b6a3a03964dd6901c1e3b2325194d94bc708
2015-06-10 15:48:59 -07:00
James Zern
b0bafd0439 vp9_decodeframe: move public funcs to end of file
Change-Id: Iaea9850890b726c7b5552c5f02b3a309086edc85
2015-06-10 15:47:57 -07:00
James Zern
38dd0448cd vp9_decodeframe: reorder some functions
removes the need for a prototype for the static function
dec_build_inter_predictors_sb

Change-Id: I36580ceae061d27f341ab0a16ece479f92e98004
2015-06-10 15:44:21 -07:00
James Zern
587bd3669e vp9_decodeframe: hide vp9_dec_build_inter_predictors_sb
+ strip 'vp9_' prefix

Change-Id: Ia057432095e5741473275d4da03ab665c37e924e
2015-06-10 15:31:55 -07:00
James Zern
44afbbb72d Merge "vp9_reconintra/d45_predictor: remove temp storage" 2015-06-10 19:23:57 +00:00
Paul Wilkins
59114915bc Merge "Changes to active maxq calculation in two pass." 2015-06-10 13:33:53 +00:00
Scott LaVarnway
97880c3324 Merge "Reducing size of MODE_INFO struct" 2015-06-10 13:15:19 +00:00
Marco
997ac14c6a Adjust some parameters for cyclic refresh for low bitrates.
Reduce motion threshold and boost factor for second segment,
for low bitrates, at low resolutions for now.
This is to reduce the rate fluctuation/frame dropping that occurs
at these low bitrates.

Change-Id: Ia66c3be41831882fca8c1e4fe104f5ea8fbe7142
2015-06-09 15:10:03 -07:00
Paul Wilkins
faf8c63b0f Changes to active maxq calculation in two pass.
Some initial experiments into discounting dead zone
formating bars and intra skip blocks (common in some
types of animation and graphics) in the calculation of
the active max Q for each ARF/GF group.

TODO: check for vertical formating bars and validate the
horizontal bar at the bottom edge of the image.

As expected, this change as it stands, does not make much
difference for the natural videos in the std-hd and derf sets.
However, for the yt and yt hd set there is a significant rise
in the average PSNR with overall PSNR and SSIM remaining
neutral.

The mean rise for the YT-HD test set was > 6%. This is mainly
because the change allows Q to drop further on titles and
other graphics sections where spending a small number of
extra bits gives a sharp rise in PSNR.

Change-Id: I3f878ae91fc1854312d7ecf9fa792c17bc1aa6b7
2015-06-09 15:31:24 +01:00
Paul Wilkins
4a28da5843 Enable more split modes for animated content.
For content that is identified as likely to contain some
animation or graphics content, increase the availability
of split modes for good quality speeds 1-3.

On a problem test animation clip this improves metrics
results by about 0.25 db and makes a noticeable difference
visually. It also causes a small drop in file size (~0.5%) but
a rise in encode time of about 5-6% at speed  2.

For more normal content it should have no effect.

Change-Id: Ic4cd9a8de065af9f9402f4477a17442aebf0e439
2015-06-09 14:50:44 +01:00
Paul Wilkins
b19b16cfa1 Merge "Animation and dead zone detection." 2015-06-08 14:26:07 +00:00
Scott LaVarnway
c9976b32b4 Update use_prev_frame_mvs flag in decoder.
Added check to see if last frame was all intra.  This will
eliminate two checks in find_mv_refs_idx().  Also, do not
update the frame mvs if the current frame is all intra.

This improved performance on material with frequent
intra-only frames.

Change-Id: I44a4042c3670ab0d38439d565062a0e2a1ba9d1e
2015-06-08 03:38:13 -07:00
Parag Salasakar
a2288d274c mips msa vp9 intra-pred optimization
intra pred - average improvement ~2x-3x

Change-Id: Ie3f7d6eded5ecb7ed7ee506ba8e4d98f93803b09
2015-06-06 22:29:32 +05:30
hkuang
87c21a95ae Optimize the decode_partition. About 0.7% gain on N10.
Change-Id: Ia689c254bd2d4f274abcc451a9b758f62e3a2b1f
2015-06-05 15:47:09 -07:00
James Zern
9c6eea35b6 Merge "vp9_reconintra: simplify d63_predictor" 2015-06-05 21:49:13 +00:00
Scott LaVarnway
5831a7127e Merge "BUG FIX: Remove counts param" 2015-06-05 20:04:26 +00:00
Scott LaVarnway
63819c033c BUG FIX: Remove counts param
member access within null pointer of type 'FRAME_COUNTS'

Change-Id: Id3bf75e0a6f2a1abf8522cf9fbb98b3a4443de38
2015-06-05 11:54:04 -07:00
Frank Galligan
8c854769fc Merge "Add control to skip loop filter in VP9 decoder." 2015-06-05 18:03:51 +00:00
Frank Galligan
bfb6d48812 Add control to skip loop filter in VP9 decoder.
This control allows the application to skip the loop filter in the
decoder. This is an advanced control that should only be used in
extreme circumstances as it may introduce and accumulate decode
artifacts.

Change-Id: I278c65c60826f84c9141ebe06c6eeed3c2335fa8
2015-06-05 10:07:09 -07:00
Johann
a4dad3e961 Merge "Duplicate reference variance code" 2015-06-05 16:54:33 +00:00
Marco
8710cceb45 Fix to spatial svc: set reference_frame masking.
For real-time mode: keep reference_frame masking off
for spatial svc.

Change-Id: I15e123c06f67ea040172b8d4042a672f3525b9d8
2015-06-05 08:25:33 -07:00
Parag Salasakar
78b434e8b1 Merge "mips msa vp9 loopfilter 4, 8 optimization" 2015-06-05 05:32:18 +00:00
Parag Salasakar
d43fd99822 mips msa vp9 loopfilter 4, 8 optimization
average improvement ~3x-4x

Change-Id: I59279293ce4b2a1e99bd10579ac97740e943643f
2015-06-05 09:56:08 +05:30
Marco
0d0db3e8c3 Merge "Bugfx in setting layer framerate." 2015-06-05 04:26:06 +00:00
James Zern
60d0b3364c vp9_reconintra/d45_predictor: remove temp storage
dst row 0 can be reused in the same way

Change-Id: Id977da62545dcc4a89cebbcbad90ba84f8ff5d6b
2015-06-04 20:11:53 -07:00
James Zern
7012ba6395 vp9_reconintra: simplify d63_predictor
calculate the averages needed for even and odd rows once; this removes a
conditional from the inner loop
the final average calculated currently relies on above[] being extended,
it could be reduced to use
above[block_size - 2] + 3 * above[block_size - 1]

Change-Id: I70f5eac8d8a2a959c7114844a95826f445c3dd4d
2015-06-04 19:21:05 -07:00
Parag Salasakar
dc07cc6fed Merge "mips msa vp9 loopfilter 16 optimization" 2015-06-05 02:15:26 +00:00
James Zern
c2cf347fe2 Merge "vp9_reconintra: use AVG[23] consistently" 2015-06-05 02:15:22 +00:00
James Zern
2b6d62140e Merge "vp9_reconintra_neon_asm/tm4x4: simplify left load" 2015-06-05 01:46:39 +00:00
James Zern
6c3b691c49 Merge "vp9_reconintra: fix d45/d63 discrepancies" 2015-06-04 22:56:43 +00:00
James Zern
faea038f4f vp9_reconintra: fix d45/d63 discrepancies
the final index in rows 2, 3 differ from vp8

Change-Id: I0fcea907b4ab44e266c0f1fd77b290d2236b280a
2015-06-04 14:49:56 -07:00
Marco
8f7e7663ad Bugfx in setting layer framerate.
Index for ts_rate_decimator should be temporal layer (tl) index.

Change-Id: I0320b7f7ae987ef64fdfe7c45099e7978a8fef17
2015-06-04 13:12:09 -07:00
Scott LaVarnway
baaaa57533 Reducing size of MODE_INFO struct
Reduced size from 124 bytes to 104 bytes.  For decode only builds,
it is reduced to 68 bytes.

Change-Id: If9e6b92285459425fa086ab5a743d0a598a69de3
2015-06-04 07:32:16 -07:00
Scott LaVarnway
8bb37dd069 Remove cm parameter from vp9_decode_block_tokens() part 2
Change-Id: Iee24b6bb095f748333223e6036fc5c9d9e7e5f1c
2015-06-04 07:13:19 -07:00
Scott LaVarnway
877fac122b Merge "Remove counts param" 2015-06-04 13:46:42 +00:00
Parag Salasakar
914f8f9ee0 mips msa vp9 loopfilter 16 optimization
average improvement ~3x-4x

Change-Id: I8ef263da6ebcf8f20aabaefeccf25a84640ba048
2015-06-04 11:50:41 +05:30
Johann Koenig
c005792951 Merge "Make vp9 subpixel match vp8" 2015-06-04 06:16:13 +00:00
Parag Salasakar
fd891a9655 Merge "mips msa vp9 convolve8 avg hv optimization" 2015-06-04 05:44:24 +00:00
Johann
eb88b172fe Make vp9 subpixel match vp8
The only difference between the two was that the vp9 function allowed
for every step in the bilinear filter (16 steps) while vp8 only allowed
for half of those. Since all the call sites in vp9 (<< 1) the input, it
only ever used the same steps as vp8.

This will allow moving the subpel variance to vpx_dsp with the rest of
the variance functions.

Change-Id: I6fa2509350a2dc610c46b3e15bde98a15a084b75
2015-06-03 22:10:51 -07:00
hkuang
ce5e17072d Merge "Optimize the idct assembly code." 2015-06-04 04:32:11 +00:00
James Zern
4fcabf5169 vp9_reconintra: use AVG[23] consistently
Change-Id: Iab7215f82be0c0c831cd81b6f8091afc3710dd54
2015-06-03 19:52:46 -07:00
Parag Salasakar
bdfbc3e876 mips msa vp9 convolve8 avg hv optimization
average improvement ~4x-6x

Change-Id: I7c8b4f2334491be8a859592606e568bc95d019aa
2015-06-04 08:11:01 +05:30
James Zern
2da8d24e8f Merge "vp9_reconintra: simplify d45_predictor" 2015-06-04 01:59:10 +00:00
James Zern
a9f55e8324 Merge changes from topic 'vp9-intra-pred'
* changes:
  vp9_reconintra: specialize d135 4x4
  vp9_reconintra: specialize d117 4x4
  vp9_reconintra: specialize d207 4x4
  vp9_reconintra: specialize d153 4x4
  vp9_reconintra: specialize d63 4x4
  vp9_reconintra: specialize d45 4x4
2015-06-04 01:58:28 +00:00
Marco
2561b4dd4c Merge "Remove ABI check for 1 pass CBR SVC." 2015-06-04 01:54:19 +00:00
James Zern
65d9599807 vp9_reconintra_neon_asm/tm4x4: simplify left load
use vld1.8 {d0[]}, [r0] rather than ldrb+vdup; mildly faster

Change-Id: Ia5ffc736bcb0f5497b7d9e55a93bf5a5f5f6928c
2015-06-03 18:51:13 -07:00
Marco
a8c5ab2ca6 Remove ABI check for 1 pass CBR SVC.
Remove the ABI check for the controls needed for SVC 1 pass CBR mode.
Bump up the ABI version.

Change-Id: I35b79ee010e14af83c6d1e801d574deaaa2fc7eb
2015-06-03 17:43:22 -07:00
hkuang
98e88e6ad8 Optimize the idct assembly code.
Change-Id: Ia0ff859ff1c813dbe100e2f27b1ef78167483f4e
2015-06-03 17:20:35 -07:00
Paul Wilkins
668e804504 Animation and dead zone detection.
Adds code to detect dead zone bars at the top and bottom
of reformatted letterbox video (note that the code only
looks at the top of the image and assumes any dead zone
is symmetrical).  Use of this to adapt rate control etc.
will follow in a subsequent patch.

Also counts other blocks (excluding the dead zone) that
have no intra signal. The presence of a significant
number of such blocks can be used as a identify that the frame
may be artificial (e.g.  animation, screen capture, graphics).
This patch contains plumbing only and does not use
the signal.

Change-Id: I59bc93529cd4065416cef773e405fda3ae006a20
2015-06-04 01:01:20 +01:00
Johann
ce2ca9f777 Duplicate reference variance code
Some places are using the unoptimized variance function. This was never
intended and does not fit into the optimization framework.

Change-Id: Id96238407aad03b0ffd4a46cd183555a026daedc
2015-06-03 13:28:59 -07:00
Parag Salasakar
b8c1cdcd12 mips msa vp9 convolve8 avg horiz optimization
average improvement ~5x-8x

Change-Id: I179a69ec620fbd69979bd128f05d18113618aab4
2015-06-03 11:33:42 +05:30
Parag Salasakar
c543d38ac7 mips msa vp9 convolve8 avg vert optimization
average improvement ~4x-6x

Change-Id: Ia2e6f770da46416ebec31fdcea5cc7878879a9d9
2015-06-03 09:55:25 +05:30
Parag Salasakar
622beaba67 Merge "mips msa vp9 idct4x4 and iwht4x4 optimization" 2015-06-03 04:10:14 +00:00
Scott LaVarnway
f779dba405 Remove counts param
Moved to MACROBLOCKD.

Change-Id: Icce765b334f2755f4fe2a4c39fb2ae2d7660d004
2015-06-02 09:06:00 -07:00
Marco
c139b81a13 Vidyo patch: Rate control for SVC, 1 pass CBR mode.
-Make Rate control work for SVC 1 pass CBR mode.
-Added temporal layering mode.
-Fixed bug in non-rd variance partition.
-Modified/updated the sample encoders (vp9_spatial_svc_encoder, vpx_temporal_svc_encoder).
-Added datarate unittest(s) for 1 pass CBR SVC.

Change-Id: Ie94b1b68a56ea1267b5087c625e5df04def2ee48
2015-06-02 07:54:13 -07:00
Parag Salasakar
54a6f73958 mips msa vp9 idct4x4 and iwht4x4 optimization
average improvement ~3x-4x
moved assert to respective files

Change-Id: I6c915059d456a00bdd76fab0dd2eede8b6c6ea58
2015-06-02 12:16:28 +05:30
Parag Salasakar
ebf7466cd8 mips msa vp9 updated convolve horiz, vert, hv, copy, avg module
Updated sources according to improved version of common MSA macros.
Enabled respective convolve MSA hooks and tests.
Overall, this is just upgrading the code with styling changes.

Change-Id: If5ad6ef8ea7ca47feed6d2fc9f34f0f0e8b6694d
2015-06-02 12:03:51 +05:30
Parag Salasakar
cf1c0ebc3a Merge "mips msa vp9 updated idct 8x8, 16x16 and 32x32 module" 2015-06-02 04:48:02 +00:00
James Zern
71d923232c Merge changes from topic 'vp9-intra-pred'
* changes:
  vp9_reconintra_neon/tm: improve above_left load
  vp9_reconintra_neon: cosmetics: normalize fn params
2015-06-01 20:03:47 +00:00
James Zern
b601202905 Merge "vp9_reconintra_neon_asm/tm: simplify above_left load" 2015-06-01 20:01:38 +00:00
paulwilkins
dbd3760712 Merge "Fast feedback of bits on undershoot." 2015-06-01 18:15:10 +00:00
Parag Salasakar
6af9d7f2e2 mips msa vp9 updated idct 8x8, 16x16 and 32x32 module
Updated sources according to improved version of common MSA macros.
Enabled idct MSA hooks and tests.
Overall, this is just upgrading the code with styling changes.

Change-Id: I1f488ab2c741f6c622b7a855388a202168082209
2015-06-01 09:24:23 +05:30
James Zern
acc481eaae vp9_reconintra: simplify d45_predictor
only the immediate above right pixel is needed; this removes a
conditional from the inner loop
the final average calculated currently relies on above[] being extended,
it could be reduced to use above[block_size - 2] + 3 * above_right

Change-Id: Ica4f2b8d25eec3ca1d6fa52ef0d4adc228eeea3f
2015-05-30 13:30:59 -07:00
James Zern
6e068e51b5 vp9_reconintra: specialize d135 4x4
based on webp's RD4()

Change-Id: I64c8f0a1325a8f201eaad39b396fae7a2d06efff
2015-05-30 13:29:40 -07:00
James Zern
b6782686f4 vp9_reconintra: specialize d117 4x4
based on webp's VR4()

Change-Id: Ic8c0b8ed65a63772ca0a4321592880a5e8947db5
2015-05-30 13:29:02 -07:00
James Zern
c022dbc4d3 vp9_reconintra: specialize d207 4x4
based on webp's HU4()

Change-Id: I2401ef307cd94e70cc7904f55954af04290c8af9
2015-05-30 13:28:22 -07:00
James Zern
2276eb16f3 vp9_reconintra: specialize d153 4x4
based on webp's HD4()

Change-Id: Icba1e21ec4b8f5026dc92e49741a68b059c8b9b1
2015-05-30 13:27:50 -07:00
James Zern
102123821d vp9_reconintra: specialize d63 4x4
based on webp's VL4()

Change-Id: Ibab962053843eae8752b4e74b6481a53bb034ae9
2015-05-30 13:27:03 -07:00
James Zern
6051bcc3dc vp9_reconintra: specialize d45 4x4
based on webp's LD4()

Change-Id: I74855d23ce73e1c6988fe08bf7c959b7a69b4abf
2015-05-30 13:26:21 -07:00
Parag Salasakar
71e88f903d Merge "mips msa vp9 updated macros and disable all MSA functions" 2015-05-30 02:52:27 +00:00
James Zern
7621b48a1c vp9_reconintra_neon/tm: improve above_left load
use vld1?_dup_u8 over vdup?_n_u8, reduces general register use; mildly
faster

Change-Id: Ie0e4e550849a207b34b378541196b553c9f12011
2015-05-29 19:18:43 -07:00
James Zern
f2d621e383 vp9_reconintra_neon: cosmetics: normalize fn params
s/y_stride/stride/

Change-Id: Ie98c3fe241dc240b653849eda356a8862bdd52f4
2015-05-29 19:01:39 -07:00
James Zern
b337c54cc4 vp9_reconintra_neon_asm/tm: simplify above_left load
use vld1.8 {d0[]}, [r0] rather than ldrb+vdup; mildly faster

Change-Id: I5c24d49a90c2855c94395184774b289da8e9d5a7
2015-05-29 18:56:16 -07:00
James Zern
7544e766e4 Merge changes from topic 'vp9-intra-pred'
* changes:
  vp9_reconintra_neon: add DC 16x16 predictors
  vp9_reconintra_neon: add DC 8x8 predictors
2015-05-30 01:51:35 +00:00
James Zern
a2a13cbe5f vp9_reconintra_neon: add DC 16x16 predictors
85-89% faster over 20M pixels

Change-Id: I9b320ed6b9e67f27df738b84c8b43b65a93c50c2
2015-05-29 15:41:44 -07:00
James Zern
e97b849219 vp9_reconintra_neon: add DC 8x8 predictors
~90% faster over 20M pixels

Change-Id: Iab791510cc57c8332c2f9a5da0ed50702e5f5763
2015-05-29 15:39:08 -07:00
Parag Salasakar
f9f078ebb6 mips msa vp9 updated macros and disable all MSA functions
Done little restructuring/styling changes to the sources like generic macro definitions, their use to reduce code lines, better code alignments etc.
Disabled all MSA hooks and tests

Change-Id: Ic6f2dce0b501f46b80c06c46c0fe2043d557b190
2015-05-29 13:34:33 +05:30
Marco
26ab314176 For non-rd pickmode: remove VAR_PARTITION condition.
Keep the logic, transform size based on cyclic refresh and bsize,
(that was conditioned on VAR_PARTITION conditions) the same
for all speeds in non-rd mode (speeds >= 5).

No change to speeds >=6.
Small improvement for speed 5, ~0.5/1.5% gain for avg psnr/ssim.

Change-Id: If9c5657f3d30efd3c7f147166bba7cb69ea55114
2015-05-28 17:29:47 -07:00
Minghai Shang
45db29784d Merge "[svc] Disable tiles for spatial svc case" 2015-05-28 22:13:54 +00:00
Johann
7c16dcc79b Merge "Check size restrictions before running test vector" 2015-05-28 22:01:53 +00:00
Scott LaVarnway
bbea7c95d8 Merge "Re-worked header files" 2015-05-28 19:56:39 +00:00
Johann
3f2a06674a Merge "Don't #define snprintf in VS 2015 or higher." 2015-05-28 19:38:57 +00:00
hkuang
5317185eb0 Merge "Add error handling when running out of free frame buffers." 2015-05-28 17:41:01 +00:00
Johann
cad0eca25c Don't #define snprintf in VS 2015 or higher.
In VS 2015 and higher snprintf is supplied and therefore vsnprintf
doesn't need to be defined. This also avoids problems caused by
_snprintf being different from snprintf.

This fixes a build break with VS 2015 and improves security.

Originally submitted via chromium by brucedawson@chromium.org
https://codereview.chromium.org/1055603003

Additionally break this MSVC-specific tweak to a new file, which will
become the home of all such MSVC-specific things.

This requires adding a dependency on msvc.h to every example which uses
args.c and tools_common.h

Change-Id: I35b5f8e7ea00f6627403aabc9ea79b0412557a99
2015-05-27 18:28:25 -07:00
Johann
1e4473b216 Check size restrictions before running test vector
Change-Id: I60ea7724e6ab06fc658f678c1b76d984a43f5a5e
2015-05-27 15:55:17 -07:00
hkuang
131cab7c27 Add error handling when running out of free frame buffers.
Change-Id: If28b59b9521204a6e3aecedcf75932d76a752567
2015-05-27 14:20:58 -07:00
Minghai Shang
cbdfdb947c Merge "[decoder] Optimize context buffer re-allocation" 2015-05-27 20:24:30 +00:00
Marco
a49fff632c Non-rd variance partition: Adjust thresholds for 1080p.
Increase the 32x32 split threshold, to allow for more 32x32
at expense of 16x16. Visually looks somewhat better.

Change-Id: Ia1439c3a0dc2d7933468b88bd59266fcd9f03505
2015-05-27 12:30:35 -07:00
Marco
109a2edf90 Merge "Refactor set_vbp_thresholds." 2015-05-27 19:10:28 +00:00
Minghai Shang
30181c46d8 Merge "[svc] Make size of empty frame to be 16x16 all the time" 2015-05-27 17:49:00 +00:00
Marco
f76d42a98a Refactor set_vbp_thresholds.
Break out the setting of the block variance split thresholds,
since they are locally modified, e.g., based on local/segment qp.

No change in performance.

Change-Id: I0a3238e6dab05140657539fc4bd27ac5ff7a554e
2015-05-27 09:25:18 -07:00
Minghai Shang
15353216c5 [svc] Make size of empty frame to be 16x16 all the time
Change-Id: Ibab09aa0e8c69cf5efea2f0ec035e5da9cc894b0
2015-05-26 16:04:36 -07:00
Johann
dee70d355f Merge "Move variance functions to vpx_dsp" 2015-05-26 23:02:11 +00:00
Johann
c3bdffb0a5 Move variance functions to vpx_dsp
subpel functions will be moved in another patch.

Change-Id: Idb2e049bad0b9b32ac42cc7731cd6903de2826ce
2015-05-26 12:01:52 -07:00
Minghai Shang
9ae5fb706e Merge "[svc] Turn on frame_parallel_decoding_mode" 2015-05-26 17:50:45 +00:00
Jingning Han
55ef1ae9e7 Merge "Fix integral projection motion search for frame resize" 2015-05-26 16:08:31 +00:00
Scott LaVarnway
89ca85dacd Move inter_predictor to vp9_reconinter.h
This function was originally static.

Change-Id: I1922fa86711ace884d9f394210b6bb9ea2a0bfe3
2015-05-26 04:22:11 -07:00
James Zern
02fda6582c Merge changes Ie15e301e,Ib070c79b
* changes:
  vp9_reconintra_neon: cosmetics: reindent
  vp9_reconintra_neon: cosmetics: drop unneeded returns
2015-05-23 17:47:52 +00:00
James Zern
4e11f3ca6e vp9_reconintra_neon: cosmetics: reindent
Change-Id: Ie15e301e8f55cf928f42a03e53a8bb8b66d0e5d5
2015-05-22 21:04:30 -07:00
James Zern
ff683ab1da vp9_reconintra_neon: cosmetics: drop unneeded returns
Change-Id: Ib070c79bdbb9c1f4e25af693d7056ec9f964c789
2015-05-22 20:59:36 -07:00
James Zern
8c15ced172 vp9: move ssse3 convolve fns to intrinsics file
+ synchronize filter function signatures

this makes any intrinsics filters available for inlining and has the
side-effect of making those filters static, quieting missing-prototype
warnings.

Change-Id: I1908875caffa585bd4fc65aaf10d17a5e20cfb46
2015-05-22 20:14:16 -07:00
James Zern
2161e44025 vp9: move avx2 convolve fns to intrinsics file
+ synchronize filter function signatures

this makes any intrinsics filters available for inlining and has the
side-effect of making those filters static, quieting missing-prototype
warnings.

Change-Id: I1cd55c9d52547793ad65aa90c7620f0e426edaa2
2015-05-22 20:13:06 -07:00
James Zern
ef2b3cce50 add vp9/common/x86/convolve.h
collect the vp9_convolve function definition macros there; this will
allow some relocation of functions from vp9_asm_stubs.c

Change-Id: Idadd117fa256dd48748379856973fd985b8204e8
2015-05-22 20:12:16 -07:00
Jingning Han
96dba4902c Fix integral projection motion search for frame resize
This commit fixes the integral projection motion search crash when
frame resize is used. It fixes issue 994.

Change-Id: Ieeb52619121d7444f7d6b3d0cf09415f990d1506
2015-05-22 15:40:45 -07:00
James Zern
48d8291df4 vp9_subpixel_8t_intrin_ssse3: quiet vs9 warning
reorder includes to avoid:
warning C4985: 'ceil': attributes not present on previous declaration.

this is the same workaround used in vp9/common/vp9_systemdependent.h

Change-Id: Ia10dd63de24f96fa1507a6179220e9d6ec774db6
2015-05-22 12:05:02 -07:00
Scott LaVarnway
b962646fc5 Re-worked header files
Various header/test files had to be re-worked in order to
build "Remove cm parameter from vp9_decode_block_tokens()".

This patch reverts the "Remove cm" part and only contains
the re-worked header files.

Change-Id: I520958a88d1991fee988a3c784d0eac40e117a32
2015-05-22 11:19:51 -07:00
paulwilkins
da8c3bf218 Merge "Fix issues with mixed ARF and GF groups." 2015-05-21 12:08:15 +00:00
James Zern
a492bcef87 vp9_mvref_common.c: fix compile warning
string literal to int within an assert

Change-Id: Ifd7acc717e01ee1bb3955ef830ec0d1645942459
2015-05-20 16:45:16 -07:00
Minghai Shang
9843e7c635 [svc] Disable tiles for spatial svc case
Change-Id: I8655a6760ab61947c09f337ddd9f4c1baf803a56
2015-05-20 14:31:49 -07:00
Minghai Shang
48bfee8797 [decoder] Optimize context buffer re-allocation
1. Check existing buffer sizes when re-allocate context buffers.
2. Don't need to set mi buffers to 0 during setup_mi.

Change-Id: I6b48b0e077a4d804312b605ad0dc34aec5795a6d
2015-05-20 11:05:22 -07:00
Minghai Shang
e2c6a633fb [svc] Turn on frame_parallel_decoding_mode
Change-Id: I33b0384ee87f83950e03be6c999bc5f193055fd3
2015-05-20 10:56:48 -07:00
paulwilkins
883fdd45cf Fast feedback of bits on undershoot.
This patch provides a partial rapid feedback of bits
resulting from extreme undershoot.

Some  improvement on some problem animated material
but in its current form only a small impact on the metrics results
of our standard test sets.

Change-Id: Ie03036ea8123bc2553437cb8c8c9e7a9fc5dac5d
2015-05-20 16:47:34 +01:00
paulwilkins
ade9693a30 Fix issues with mixed ARF and GF groups.
This patch addresses two issues that can occur when the
encoder chooses to use a mixture of ARF and GF groups.

The first issue relates to a failure to reset the "ARF active" flag
correctly when transitioning from coding ARF groups to coding
GF groups. This caused some golden frames to be  encoded
with an incorrect bit rate target as if they were ARF overlay frames.

The second issue relates to the encoding of a single short GF group
just before a key frame.  Where the last group before a key frame
is an ARF group we expect the final frame before the key frame  to
be an low data rate overlay frame. However, when the last  group
is a GF group, the final frame before the key frame should be a normal
frame with a normal bit allocation. This issue had the potential to cause
a single poorly coded frame just before a key frame. If that key frame
were a forced key frame rather than a real scene cut, this might cause
pulsing.

Change-Id: Idf1eb5eaf63a231495a74de7899236e1ead9fb00
2015-05-20 16:46:44 +01:00
James Zern
37d03809bf Merge "rename vp9_dct_impl_sse2.c to vp9_dct_sse2_impl.h" 2015-05-19 00:01:57 +00:00
James Zern
1bcf7a35e9 Merge "rename vp9_dct32x32_sse2.c to vp9_dct32x32_sse2_impl.h" 2015-05-19 00:01:22 +00:00
James Zern
0f854e9ab5 Merge "rename vp9_dct32x32_avx2.c to vp9_dct32x32_avx2_impl.h" 2015-05-19 00:01:07 +00:00
James Zern
85076fc5ab dec_build_inter_predictors: don't return a void fn
split call of extend_and_predict() and return, fixes visual studio build
warning since:
0a80164 Move mc_buf to cut down size of MACROBLOCKD.

Change-Id: I7cdf712941ef773a07f038539cb8080dc27861cd
2015-05-16 10:33:57 -07:00
James Zern
a989c66b84 rename vp9_dct_impl_sse2.c to vp9_dct_sse2_impl.h
this file shouldn't be built directly, it is included in vp9_dct_sse2.c
to create a non-high-bitdepth and a high-bitdepth version

silences missing prototype warnings for the unused FDCT* functions

Change-Id: Ide6ff8c24ab31bdb0f833260505ae33660a1ad5b
2015-05-15 17:01:19 -07:00
James Zern
587a71f1d6 rename vp9_dct32x32_sse2.c to vp9_dct32x32_sse2_impl.h
this file shouldn't be built directly, it is included in vp9_dct_sse2.c
to create a non-high-bitdepth and a high-bitdepth version

silences missing prototype warnings for the unused FDCT32x32* functions

Change-Id: I0e38f16dae5ea1728de184ee2c89287d48675c51
2015-05-15 16:59:52 -07:00
James Zern
4ec47249bc rename vp9_dct32x32_avx2.c to vp9_dct32x32_avx2_impl.h
this file shouldn't be built directly, it is included in vp9_dct_avx2.c
to create a non-high-bitdepth and a high-bitdepth version

silences missing prototype warnings for the unused FDCT32x32* functions

Change-Id: I4c19935c0e035b393be513bde735e9a78064a494
2015-05-15 16:47:51 -07:00
James Zern
985f19bc6b Merge changes from topic 'missing-proto'
* changes:
  vp9_subexp.h: add a missing prototype
  vp9: add some missing includes
  vp9 intrinsics: add vp9_rtcd include
  vp9: correct some function signatures
  vp9_variance_sse2: sync function signatures
  vp9/encoder: make some functions static
  vp9_dct_sse2: make some functions static
  vp9_decodeframe.c: make a function static
2015-05-15 23:08:15 +00:00
Marco
beec69cfe2 Merge "Change tx_size_search_method setting for non-rd speed 5." 2015-05-15 21:29:01 +00:00
Marco
e88de49faa Change tx_size_search_method setting for non-rd speed 5.
Use the same settting as in speed >=6.
This will use same logic for tx_size selecton as in speed >=6,
which limits the transform size and reduces ringing artifact.
Also metrics go up on average with this change: ~2% for PSNR, ~10% for SSIM.

Change-Id: Ia2d50db236ae1cc72f742bfa6c9ec5ea50ff0e0a
2015-05-15 11:12:47 -07:00
James Zern
ca5a54113f vp9_subexp.h: add a missing prototype
+ include the .h in the .c
silences missing prototype warnings

Change-Id: Ia87366dccb4bf4e9f2ffa5d3ab51ac6ca5488c91
2015-05-15 10:43:48 -07:00
James Zern
97db651ce0 vp9: add some missing includes
mostly: <file>.c should include <file>.h
silences missing prototype warnings

Change-Id: Ic05ec32c6f7b2224b78825904d96d73aacad6000
2015-05-15 10:43:47 -07:00
James Zern
330fba41e2 vp9 intrinsics: add vp9_rtcd include
silences a missing declaration warning

Change-Id: I59a34e1a1377cf3529b678d7ec0122bd43ab1bf1
2015-05-15 10:43:47 -07:00
James Zern
18b60af27c vp9: correct some function signatures
silences missing prototype warnings

Change-Id: Idaf68d83d2cb03847f3ee002c4d00c2ac79da604
2015-05-15 10:43:47 -07:00
James Zern
43d5cc7fe1 vp9_variance_sse2: sync function signatures
+ include vp9_rtcd.h
silences missing prototype warnings

Change-Id: I77902f07a454029baad4fe5fe6fc37c65644e6f7
2015-05-15 10:43:47 -07:00
James Zern
700b7fd0a9 vp9/encoder: make some functions static
silences missing prototype warnings

Change-Id: I3338fcaa67b5dcdf6bf237e8b374db3befd18753
2015-05-15 10:43:47 -07:00
James Zern
8515e62e6b vp9_dct_sse2: make some functions static
silences missing prototype warnings

Change-Id: I773b6a6b5bd7c57db18c3b17c519534f80e131de
2015-05-15 10:43:47 -07:00
James Zern
f3bf5f2029 vp9_decodeframe.c: make a function static
silences a missing declaration warning

Change-Id: I2f49ebca9ba7a47f3c48f5fe919b90cd4114a9bc
2015-05-15 10:43:47 -07:00
Adrian Grange
8371c897dc Merge "Fix illegal memory access when stream starts w/ invisible frame." 2015-05-15 17:36:41 +00:00
Frank Galligan
d610ead258 Merge "Move mc_buf to cut down size of MACROBLOCKD." 2015-05-15 15:20:39 +00:00
paulwilkins
4f569e8485 Merge "Revert "Skip the last frame update for some frame repeats."" 2015-05-15 09:17:19 +00:00
Frank Galligan
0a80164c94 Move mc_buf to cut down size of MACROBLOCKD.
Change-Id: Icea64b9e5632b41aaa7cd7018c501d6add9b7a7f
2015-05-14 19:10:02 -07:00
Adrian Grange
f480c1256d Fix illegal memory access when stream starts w/ invisible frame.
Add a check to make sure we have a decoded frame available
before copying its 'corrupt' flag.

(Originally submitted to the old repository by Alexander Voronov
as: https://gerrit.chromium.org/gerrit/#/c/74305/).

Change-Id: Iceb4686c785afb437b668015bf8818b18d60e0ce
2015-05-14 15:49:10 -07:00
paulwilkins
eb8faf1c89 Revert "Skip the last frame update for some frame repeats."
Testing on another rate control patch reveals that in some
situations, where the encoder is flipping in and out of arf
mode, we get an encoder decoder mismatch.

Whilst it is still not clear why, skipping  the last buffer
update seems to trigger the problem. Until I can establish
why, or if there is another underlying cause, I am reverting
this change.

This reverts commit e5112b3ae3.

Change-Id: I315c5200414de89458015823344b7367e9dd75ba
2015-05-14 17:21:44 +01:00
Johann
cafae5b544 Merge "Relocate memory operations for common code" 2015-05-13 19:47:24 +00:00
Johann
1d7ccd5325 Relocate memory operations for common code
With the sad functions, and hopefully the variance functions soon,
moving to the vpx_dsp location, place the defines used in the
reference C code in a common location.

Change-Id: I4c8ce7778eb38a0a3ee674d2f1c488eda01cfeca
2015-05-13 11:41:15 -07:00
Parag Salasakar
686616a989 Merge "mips msa vp9 idct 8x8 optimization" 2015-05-13 04:36:34 +00:00
Yunqing Wang
f72af26305 Merge "Remove unneeded variable declaration" 2015-05-12 23:33:31 +00:00
Yaowu Xu
a8015e217e Merge "Protect new metric computation with use_highbitdepth flag" 2015-05-12 23:20:35 +00:00
Yaowu Xu
3f42d10805 Protect new metric computation with use_highbitdepth flag
The computation of new metrics is not supported yet in highbitdepth
mode. This commit adds protection to make sure the computation is
done only when highbitdepth is not on. This protection shall be
revised when support of highbitdpeth computation is added.

This resolves the encoder crash when configured with both
--enable-internal-stats
--enable-vp9-highbitdepth

Change-Id: Id9f4bcc4fa26d9ca0e9eabade83f3f88a5b212e6
2015-05-12 15:12:05 -07:00
Yunqing Wang
8ba2d2d5a0 Remove unneeded variable declaration
This patch fixed the following warning:
src\third_party\libvpx\source\libvpx\vp9\encoder\vp9_pickmode.c(1607) :
warning C6246: Local declaration of 'this_mode' hides declaration of the
same name in outer scope.

Change-Id: I1d93c4a47a13cb13089fec5bd61e8b58e6cd8d58
2015-05-12 15:01:40 -07:00
Adrian Grange
17fc3e94c9 Merge "Recompute tile params on frame resize" 2015-05-12 17:48:16 +00:00
James Zern
465ce0e420 Merge "build_intra_predictors*: reduce above_data size" 2015-05-11 23:48:58 +00:00
Adrian Grange
5f0dc57653 Merge "Reduce border extension by 1 line" 2015-05-11 22:47:49 +00:00
Adrian Grange
65b768fdf9 Recompute tile params on frame resize
When the frame size changes we must recompute details
of the tile dimensions.

Change-Id: Ie519bd6da47b5cd43933c0bcfc0f2429bcb01986
2015-05-11 15:45:26 -07:00
Jingning Han
2dcfd16fbd Merge "Sort variables dependency in read_uncompressed_header" 2015-05-11 19:11:56 +00:00
Jingning Han
2b2b461d39 Sort variables dependency in read_uncompressed_header
Remove a few repeated data structure reads from
read_uncompressed_header.

Change-Id: I6eb741b39f9415ad0aa4631dfbf4a1ace4eba56a
2015-05-11 10:07:55 -07:00
Marco
913862be8c Fix rate control issue with layers and aq-mode=3.
When aq-mode=3 is enabled, only for base layer frames should the
qp of the frame incorporate the segment delta-qp.

This was causing more rate mismatch for the enhancement layer frames
when running temporal layers with aq-mode=3 on.

Change-Id: I1c5e69d1ef8a51188af8696753c17fd8f67699b3
2015-05-11 10:04:18 -07:00
paulwilkins
aa5c1b4c5d Merge "Skip the last frame update for some frame repeats." 2015-05-11 12:28:45 +00:00
James Zern
a5e4ca8390 build_intra_predictors*: reduce above_data size
currently this needs to be 2x (NEED_ABOVERIGHT) the size of the largest
block (32) + 1 (for above_left). reduce the buffer size from 128 + 16
(alignment) to 64 + 16.

Change-Id: Idaca1806c7e1214e9437de24e15edc2ebf18f95d
2015-05-08 20:17:20 -07:00
James Zern
6d22713722 Merge "build_intra_predictors*: reduce left_col size" 2015-05-09 00:53:55 +00:00
hkuang
d53fb0fda5 Fix clang ioc warning due to NULL mi pointer.
The warning only happens in VP9 encoder's first pass due to src_mi
is not set up yet. But it will not fail the encoder as left_mi and
above_mi are not used in the first_pass and they will be set up again
in the second pass.

Change-Id: I0713b4660d71e229e196654cb0970ba6b1574f28
2015-05-08 15:42:50 -07:00
hkuang
f5574fb44c Merge "Add more sse2 code for intra prediction." 2015-05-08 17:26:30 +00:00
paulwilkins
e5112b3ae3 Skip the last frame update for some frame repeats.
Where a frame appears to be a repeat of an earlier
frame or frame buffer,  but the first pass code
does not anticipate this (usually because it is matching
the GF or ARF buffer not the last frame buffer), do not
update the last frame buffer.

This helps ensure that the content of the last frame buffer
is kept "different" where possible, and not updated to
match the GF or ARF. This is particularly helpful in some
animated sequences where there are groups of repeating
frames. Here it has quite a big impact. However, in most
of our standard test clips it has little or no impact.

Change-Id: I77332ee1a69f9ffc0c6080bfeb811c43fd8828e6
2015-05-08 17:51:26 +01:00
Parag Salasakar
7c5f00f868 mips msa vp9 idct 8x8 optimization
average improvement ~4x-6x

Change-Id: I5edf713721b9e24c7e0ce2e69d8fc3ecab625d91
2015-05-08 12:23:27 +05:30
Parag Salasakar
a8a9c2bb45 Merge "mips msa vp9 idct 32x32 optimization" 2015-05-08 04:27:44 +00:00
James Zern
7e55ff1593 build_intra_predictors*: reduce left_col size
this should only need to be the size of the largest block, i.e., 32, not
64.

Change-Id: Ib8cb2424771fdd2a64c55379597248b2722a5ceb
2015-05-07 16:16:42 -07:00
James Zern
fd3658b0e4 replace DECLARE_ALIGNED_ARRAY w/DECLARE_ALIGNED
this macro was used inconsistently and only differs in behavior from
DECLARE_ALIGNED when an alignment attribute is unavailable. this macro
is used with calls to assembly, while generic c-code doesn't rely on it,
so in a c-only build without an alignment attribute the code will
function as expected.

Change-Id: Ie9d06d4028c0de17c63b3a27e6c1b0491cc4ea79
2015-05-07 11:55:08 -07:00
Johann
76a08210b6 Merge "Move shared SAD code to vpx_dsp" 2015-05-07 18:33:06 +00:00
hkuang
086934136b Merge "Remove an unnecessary check." 2015-05-07 15:51:11 +00:00
Marco
97307af21a Merge "Remvoe EIGHTTAP_SHARP filter check for non-rd mode." 2015-05-07 15:40:11 +00:00
paulwilkins
aecb1770d5 Merge "Image size restriction to rd auto partition search." 2015-05-07 14:12:14 +00:00
Parag Salasakar
1601c1385a mips msa vp9 idct 32x32 optimization
average improvement ~4x-6x

Change-Id: Idaba7e49fbd7f388caee0d73773ccf6e4807ef17
2015-05-07 12:42:23 +05:30
hkuang
7153b822ed Add more sse2 code for intra prediction.
vp9_dc_left_predictor_16x16
vp9_dc_top_predictor_32x32
vp9_dc_left_predictor_32x32
vp9_dc_128_predictor_32x32

Change-Id: Ib9861deefd01c3527235b92ff6b3d571ef6b4bc6
2015-05-06 17:17:00 -07:00
Marco
76fe5dfc67 Remvoe EIGHTTAP_SHARP filter check for non-rd mode.
Using EIGHTTAP and EIGHTTAP_SMOOTH seem sufficient.
Hard to see any visual gain from allowing EIGHTTAP_SHARP, and it is
rarely selected.

PSNR/SSIM metrics go up by ~0.18/0.14%.

Change-Id: I96fa0d98f9321b913e3ebcd464d4ff3c63018791
2015-05-06 17:08:34 -07:00
Johann
d5d9289800 Move shared SAD code to vpx_dsp
Create a new component, vpx_dsp, for code that can be shared
between codecs. Move the SAD code into the component.

This reduces the size of vpxenc/dec by 36k on x86_64 builds.

Change-Id: I73f837ddaecac6b350bf757af0cfe19c4ab9327a
2015-05-06 16:58:20 -07:00
Adrian Grange
0af5ff49bd Reduce border extension by 1 line
The code was using one row too many.

Change-Id: Ie0c05d02c22ae3d0d13d3b6565c40de3bc4fa17a
2015-05-06 14:44:24 -07:00
hkuang
240767b29d Remove an unnecessary check.
Change-Id: Id0f224ac4667dd173363b0f05711678448291d4e
2015-05-06 14:15:00 -07:00
hkuang
92b199061a Correct the inter prediction coordinate calculation which greatly reduced the
times of border extension.

Change-Id: I8e5bd590cc696ee71cfe1f4cc66c12fb24aaf44e
2015-05-06 10:55:56 -07:00
hkuang
623e6eed5e Merge "Optimize the read_partition." 2015-05-06 17:29:52 +00:00
Yunqing Wang
7dbdada49f Merge "Add intra mode early termination in non-rd mode" 2015-05-06 17:12:11 +00:00
Parag Salasakar
d1cdda88bd Merge "mips msa vp9 idct 16x16 optimization" 2015-05-06 06:40:56 +00:00
Yunqing Wang
36eabb1c3c Add intra mode early termination in non-rd mode
Added the intra mode early termination in order to
speed up the mode search in non-rd case since we
started to include more intra modes in the search
list. Borg tests(rtc set) showed a 0.048% PSNR gain
and 0.061 SSIM gain. No speed change.

Change-Id: I6f255fe534dc50b736e6a66a726ad458eb9b4443
2015-05-05 16:31:36 -07:00
hkuang
4c1a8be29d Optimize the read_partition.
Change-Id: I5a796425ce5706824a2fc17c6f24f983c5b9e43b
2015-05-05 15:51:04 -07:00
James Zern
ccae5d99d2 fix and enable vp9_dc_128_predictor_16x16
widen the loads and stores to 128-bit.

this was added, but not enabled in:
493a857 Add some sse2 code for intra prediction.

Change-Id: I277d7db608a7db7d75cc0bde86f48fa66ad487e4
2015-05-05 11:40:13 -07:00
hkuang
e47811ef8f Merge "Add some sse2 code for intra prediction." 2015-05-05 17:11:07 +00:00
paulwilkins
af76953448 Merge "Remove CONSTRAIN_NEIGHBORING_MIN_MAX." 2015-05-05 09:32:11 +00:00
paulwilkins
4cd65e4f19 Merge "Adjust ARF min and max interval." 2015-05-05 09:31:38 +00:00
Parag Salasakar
60052b618f mips msa vp9 idct 16x16 optimization
average improvement ~4x-6x

Change-Id: I55e95b7f2ba403dff11813958dc7c73a900dd022
2015-05-05 12:37:06 +05:30
Marco
b9a72d3c4d Allow for H and V intra modes for non-rd mode.
For non-rd mode (speed >=5): use mask based on prediction block size, and
(for non-screen content mode) allow for checking horiz and vert intra modes
for blocks sizes < 16x16.

Avg psnr/ssim metrics go up by about ~0.2%.

Only allowing H/V intra on block sizes below 16x16 for now, to keep
encoding time increase very small, and also when allowing H/V on 16x16 blocks,
metrics went down on a few clips which need to be further examined.

Change-Id: I8ae0bc8cb2a964f9709612c76c5661acaab1381e
2015-05-04 09:48:41 -07:00
James Zern
670b2c09ce vp9_idct_intrin_sse2: cosmetics: reindent
+ fix some whitespace

Change-Id: Id61b739282014288a7e5d3c17a9d6448d9d4cda2
2015-05-01 16:07:54 -07:00
Yunqing Wang
d31256cd38 Merge "Reduce intra_cost_penalty for BLOCK_8X8" 2015-05-01 18:29:38 +00:00
Yunqing Wang
57fefd5f9a Merge "Adjust the vbp early termination threshold slightly" 2015-05-01 18:29:25 +00:00
Scott LaVarnway
7b0f24fc21 FIX: Use correct above/left in read_intra_frame_mode_info
Should be using xd->above_mi and xd->left_mi.

Change-Id: Ifab83512db5491a955a3ed44a2d6e3b25b5ae5a5
2015-05-01 08:23:47 -07:00
paulwilkins
4a7dcf8eb2 Image size restriction to rd auto partition search.
Impose a limit on the rd auto partition search based on
the image format. Smaller formats require that the search
includes includes a smaller minimum block size.

This change is intended to mitigate the visual impact of
ringing in some problem clips, for smaller image formats.

Change-Id: Ie039e5f599ee079bbef5d272f3e40e2e27d8f97b
2015-05-01 16:16:02 +01:00
paulwilkins
287b0c6da9 Remove CONSTRAIN_NEIGHBORING_MIN_MAX.
Remove one of the auto partition size cases.
This case can behaves badly in some types of animated content
and was only used for the rd encode path. A subsequent patch
will add additional checks to help further improve visual quality.

Change-Id: I0ebd8da3d45ab8501afa45d7959ced8c2d60ee4e
2015-05-01 15:15:16 +01:00
paulwilkins
e0786c280e Adjust ARF min and max interval.
Previously limit on max interval  set to 0.5 seconds.
Though this helped some low frame rate material it
appears to be a bit too aggressive for some 24 and 25 fps
content. This patch relaxes the limit to 0.75 seconds.

The patch also adds a new minimum interval variable
to replace the current hard wired value. This allows us
to impose a limit on the maximum number of primary
arfs per second for high frame rate (e.g. 50 & 60fps)
content. This is to address concerns regarding playback
performance on some platforms if there is a high base
frame rate and very frequent arfs.

Change-Id: I373e8b6b2a8ef522eced6c6d2cceb234ff763fcf
2015-05-01 15:11:49 +01:00
James Zern
c77b1f5acd vp9: RECON_AND_STORE4X4: remove dest offset
offsetting by a variable stride prevents instruction reordering,
resulting in poor assembly

Change-Id: Id62d6b3299cdd23f8c44f97b630abf4fea241446
2015-04-30 19:14:17 -07:00
James Zern
778845da05 vp9_idct_intrin_*: RECON_AND_STORE: remove dest offset
offsetting by a variable stride prevents instruction reordering,
resulting in poor assembly.
additionally reroll 16x16/32x32 loops to reduce register spill with this
new format

Change-Id: I0635b8ba21ecdb88116e927dbdab53acdf256e11
2015-04-30 19:14:17 -07:00
Yaowu Xu
2061359fcf Merge "Remove vp9_idct16x16_10_add_ssse3()" 2015-04-30 23:13:33 +00:00
James Zern
0ae1e4a95a Merge "vp9_decodeframe: simplify compare_tile_buffers" 2015-04-30 23:05:42 +00:00
hkuang
493a8579f1 Add some sse2 code for intra prediction.
Change-Id: I16c0a62e52dab62837c547345df31e7518620ed4
2015-04-30 15:42:57 -07:00
Yunqing Wang
4907c29904 Reduce intra_cost_penalty for BLOCK_8X8
This patch reduced the BLOCK_8X8's intra_cost_penalty, which
allows 8x8 blocks to conduct intra mode search. Borg test
result(rtc set): 0.077% PSNR gain, 0.228% SSIM gain. No speed
changes.

Change-Id: Icfe90c4f6969de24bda8ecacbd3da50330bf22b2
2015-04-30 11:03:06 -07:00
Yaowu Xu
47767609fe Remove vp9_idct16x16_10_add_ssse3()
The rotation computation using 2X of cos(pi/16) has a potential to
overflow 32 bit, this commit disable the function to allow further
investigation and optimization.

Change-Id: I4a9803bc71303d459cb1ec5bbd7c4aaf8968e5cf
2015-04-30 09:07:30 -07:00
Yunqing Wang
fd90ce2711 Merge "Improve golden frame refreshing in non-rd mode" 2015-04-30 15:57:55 +00:00
Yunqing Wang
a257e469e1 Adjust the vbp early termination threshold slightly
Calculated cpi->vbp_threshold_sad from this frame's dequant value.
The encoding quality and speed didn't change much. Borg test
result: PSNR: -0.002%, SSIM: -0.003%.

Change-Id: I97c9826986f39582f29910d637d08a69c90afdee
2015-04-30 08:51:02 -07:00
Parag Salasakar
95cb130f32 Merge "mips msa vp9 copy and avg convolve optimization" 2015-04-30 04:39:13 +00:00
Yaowu Xu
d45870be8d Merge "Disable ssse3 version idct16x16_256_add()" 2015-04-30 03:09:23 +00:00
James Zern
9e81112df2 vp9_decodeframe: simplify compare_tile_buffers
return the difference between the 2 buffer sizes rather than exactly
-1/0/1.

Change-Id: Idf1ccff7088b31845470bcc71bea5927b0598cc7
2015-04-29 17:42:30 -07:00
Yaowu Xu
486a73a9ce Disable ssse3 version idct16x16_256_add()
The version is currently producing different result from c version
for some input. Disable the use of it for now to allow time for
investigation the source of mismatch.

Change-Id: Id039455494ee531db4886a9f1fa4761174ef6df3
2015-04-29 16:58:59 -07:00
Yunqing Wang
d31698b0e0 Improve golden frame refreshing in non-rd mode
The default golden frame interval was doubled. After encoding a
frame, the background motion was measured. If the motion was high,
the current frame was set as the golden frame. Currently, the
changes were applied only while aq-mode 3 was on.

Borg tests(rtc set) showed a 0.226% PSNR gain and 0.312% SSIM gain.
No speed changes.

Change-Id: Id1e2793cc5be37e8a9bacec1380af6f36182f9b1
2015-04-29 16:43:43 -07:00
Parag Salasakar
2301d10f73 mips msa vp9 copy and avg convolve optimization
average improvement ~3x-5x

Change-Id: I422e4c33ea7e6d6783ba40029438ccf21b0e76bb
2015-04-29 12:28:17 +05:30
James Zern
f58011ada5 vpx_mem: remove vpx_memset
vestigial. replace instances with memset() which they already were being
defined to.

Change-Id: Ie030cfaaa3e890dd92cf1a995fcb1927ba175201
2015-04-28 20:00:59 -07:00
James Zern
f274c2199b vpx_mem: remove vpx_memcpy
vestigial. replace instances with memcpy() which they already were being
defined to.

Change-Id: Icfd1b0bc5d95b70efab91b9ae777ace1e81d2d7c
2015-04-28 19:59:41 -07:00
James Zern
fbd3b89488 vpx_mem: remove vpx_memmove
vestigial. replace instances with memmove() which they already were
being defined to.

Change-Id: If396d3f9e3cf79c0ee5d7429615ef3d6b2a34afa
2015-04-28 19:59:40 -07:00
Frank Galligan
2be50a1c9c Merge "WIP: Use LUT for y_dequant/uv_dequant" 2015-04-28 16:12:10 +00:00
Yunqing Wang
bfce02971e Merge "Fix debugmodes file to print modes and MVs correctly" 2015-04-28 15:47:46 +00:00
Scott LaVarnway
afcb62b414 WIP: Use LUT for y_dequant/uv_dequant
instead of calculating every block.

Change-Id: Ib19ff2546be8441f8755ae971ba2910f29412029
2015-04-28 07:52:06 -07:00
Yunqing Wang
297b2b99de Fix debugmodes file to print modes and MVs correctly
This patch fixed the issues in debugmodes file because of the recent
changes in MODE_INFO struct.

Change-Id: I4df83379ecc887c1f009d4a8329c9809c5b299d6
2015-04-27 17:09:38 -07:00
Yaowu Xu
b3e411e481 Add validation of UV partition size
For color sampling format other than 420, valid partion size in Y may
not work for UV plane. This commit adds validation of UV partition
size before select the partition choice.

This fixes a crash for real time encoding of 422 input.

Change-Id: I1fe3282accfd58625e8b5e6a4c8d2c84199751b6
2015-04-24 12:34:18 -07:00
Jim Bankoski
a6e9ae9066 Adds worst frame metrics for a bunch of metrics.
Change-Id: Ieaccc36ed1bee024bb644a9cfaafdaaa65d31772
2015-04-22 06:45:56 -07:00
paulwilkins
e07b141da0 Merge "Modified test for auto key frame detection." 2015-04-22 02:29:17 -07:00
paulwilkins
5d8877a944 Merge "Limit arf interval for low fpf clips." 2015-04-22 02:25:38 -07:00
Parag Salasakar
1c9af9833d Merge "mips msa vp9 convolve8 horiz optimization" 2015-04-21 22:08:25 -07:00
Jim Bankoski
3b35e962e2 Merge "Adds a new temporal consistency metric to libvpx." 2015-04-21 16:11:11 -07:00
Johann
931c0a954f Merge "Rename neon convolve avg file" 2015-04-21 15:45:29 -07:00
Johann
66b9933b8d Rename neon convolve avg file
Some build systems use just the basename for object files.

Change-Id: I333e1107ee866f3906cc46476ef8d04c6200a8a0
2015-04-21 14:18:17 -07:00
Scott LaVarnway
8b17f7f4eb Revert "Remove mi_grid_* structures."
(see I3a05cf1610679fed26e0b2eadd315a9ae91afdd6)

For the test clip used, the decoder performance improved by ~2%.
This is also an intermediate step towards adding back the
mode_info streams.

Change-Id: Idddc4a3f46e4180fbebddc156c4bbf177d5c2e0d
2015-04-21 11:16:45 -07:00
Jim Bankoski
ee87e20d53 Adds a new temporal consistency metric to libvpx.
Change-Id: Id61699ebf57ae4f8af96a468740c852b2f45f8e1
2015-04-21 10:05:37 -07:00
Yaowu Xu
924d06a075 Merge "Resolve configuration conflict" 2015-04-21 08:00:49 -07:00
paulwilkins
3606b78108 Modified test for auto key frame detection.
The existing test was triggering a lot of false positives on some types
of animated material with very plain backgrounds. These were triggering
code designed to catch key frames in letter box format clips.

This patch tightens up the criteria and imposes a minimum requirement
on the % blocks coded intra in the first pass and the ratio between the
% coded intra and the modified inter % after discounting neutral (flat)
blocks that are coded equally well either way.

On a particular problem animation clip this change eliminated a large
number of false positives including some cases where the old code
selected kf several times in a row. Marginal false negatives are less
damaging typically to compression and in the problem clip there are now
a couple of cases where "visual" scene cuts are ignored because of well
correlated content across the scene cut.

Replaced some magic numbers related to this with #defines and added
explanatory comments.

Change-Id: Ia3d304ac60eb7e4323e3817eaf83b4752cd63ecf
2015-04-21 12:50:11 +01:00
Parag Salasakar
ca90d4fd96 mips msa vp9 convolve8 horiz optimization
average improvement ~6x-8x

Change-Id: I7c91eec41aada3b0a5231dda7869b3b968f3ad18
2015-04-21 12:31:26 +05:30