virtualenv/FFmpeg
1
0
Fork 0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-12 19:18:44 +02:00
Code Issues Releases Activity
94,469 Commits 37 Branches 400 Tags 463 MiB
652ea23cb3
Commit Graph

7587 Commits

Author SHA1 Message Date
Andriy Gelman
f60b1211b2 lavfi/zmq: Avoid mem copy past the end of input buffer 2019-08-01 19:10:05 +02:00
Rodger Combs
a0c1970781 lavfi/vf_thumbnail_cuda: fix operator precedence bug 
Discovered via a warning when building with clang

Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
 2019-07-30 15:17:22 +02:00
Guo, Yejun
ccbab41039 dnn: convert tf.pad to native model in python script, and load/execute it in the c code. 
since tf.pad is enabled, the conv2d(valid) changes back to its original behavior.

Signed-off-by: Guo, Yejun <yejun.guo@intel.com>
Signed-off-by: Pedro Arthur <bygrandao@gmail.com>
 2019-07-29 12:34:19 -03:00
Guo, Yejun
df8db34552 dnn: add layer pad which is equivalent to tf.pad 
the reason to add this layer first is that vf_sr uses it in its
tensorflow model, and the next plan is to update the python script
to convert tf.pad into native model.

Signed-off-by: Guo, Yejun <yejun.guo@intel.com>
Signed-off-by: Pedro Arthur <bygrandao@gmail.com>
 2019-07-29 12:34:19 -03:00
Gyan Doshi
43891ea8ab avfilter/fade: don't allow nb_frames == 0 
In filter init, there's a division by nb_frames but value isn't checked.
 2019-07-29 12:46:59 +05:30
Mark Thompson
20fed2f0ab lavfi: addroi filter 
This can be used to add region of interest side data to video frames.
 2019-07-28 22:34:33 +01:00
Guo, Yejun
1b9064e3f4 libavfilter/dnn: move dnn files from libavfilter to libavfilter/dnn 
it is expected that there will be more files to support native mode,
so put all the dnn codes under libavfilter/dnn

The main change of this patch is to move the file location, see below:
modified:   libavfilter/Makefile
new file:   libavfilter/dnn/Makefile
renamed:    libavfilter/dnn_backend_native.c -> libavfilter/dnn/dnn_backend_native.c
renamed:    libavfilter/dnn_backend_native.h -> libavfilter/dnn/dnn_backend_native.h
renamed:    libavfilter/dnn_backend_tf.c -> libavfilter/dnn/dnn_backend_tf.c
renamed:    libavfilter/dnn_backend_tf.h -> libavfilter/dnn/dnn_backend_tf.h
renamed:    libavfilter/dnn_interface.c -> libavfilter/dnn/dnn_interface.c

Signed-off-by: Guo, Yejun <yejun.guo@intel.com>
Signed-off-by: Pedro Arthur <bygrandao@gmail.com>
 2019-07-26 13:07:43 -03:00
Paul B Mahol
73afea3ccc avfilter/af_dynaudnorm: add more descriptive aliases for options 2019-07-24 15:20:47 +01:00
Paul B Mahol
3883c9d147 avfilter/vf_ciescope: add DCI-P3 2019-07-22 22:56:21 +01:00
Limin Wang
b696caba1a avformat/f_select: add support for more pixel formats for scene change score calculations 
This avoids automatic conversions to rgb if scene change score is used in the expression.

Below is the tested results for the new added pixel format without autoscale to rgb24:
1. AV_PIX_FMT_YUVJ420P
time ./ffprobe -of compact=p=0 -show_entries frame=pkt_pts:frame_tags -bitexact -f lavfi \
"sws_flags=+accurate_rnd+bitexact;movie=../fate-suite/svq3/Vertical400kbit.sorenson3.mov,select=gt(scene\,.25)"

master:
pkt_pts=1620|tag:lavfi.scene_score=1.000000
pkt_pts=4140|tag:lavfi.scene_score=0.875036
pkt_pts=5800|tag:lavfi.scene_score=1.000000
pkt_pts=6720|tag:lavfi.scene_score=0.461625
pkt_pts=8160|tag:lavfi.scene_score=1.000000
pkt_pts=9760|tag:lavfi.scene_score=1.000000
pkt_pts=14080|tag:lavfi.scene_score=0.838916
pkt_pts=15700|tag:lavfi.scene_score=1.000000
pkt_pts=18500|tag:lavfi.scene_score=0.474948
pkt_pts=20040|tag:lavfi.scene_score=0.379700
pkt_pts=21760|tag:lavfi.scene_score=1.000000
./ffprobe -of compact=p=0 -show_entries frame=pkt_pts:frame_tags -bitexact -f  0.71s user 0.01s system 99% cpu 0.721 total

patch applied:
pkt_pts=1620|tag:lavfi.scene_score=1.000000
pkt_pts=4140|tag:lavfi.scene_score=0.668643
pkt_pts=5800|tag:lavfi.scene_score=0.996721
pkt_pts=6720|tag:lavfi.scene_score=0.357390
pkt_pts=8160|tag:lavfi.scene_score=0.886268
pkt_pts=9760|tag:lavfi.scene_score=0.926219
pkt_pts=14080|tag:lavfi.scene_score=0.650033
pkt_pts=15700|tag:lavfi.scene_score=1.000000
pkt_pts=18500|tag:lavfi.scene_score=0.316402
pkt_pts=20040|tag:lavfi.scene_score=0.269509
pkt_pts=21760|tag:lavfi.scene_score=1.000000
./ffprobe -of compact=p=0 -show_entries frame=pkt_pts:frame_tags -bitexact -f  0.19s user 0.01s system 81% cpu 0.240 total

2. AV_PIX_FMT_YUV420P
time ./ffprobe -of compact=p=0 -show_entries frame=pkt_pts:frame_tags -bitexact -f lavfi \
"sws_flags=+accurate_rnd+bitexact;movie=../Passengers_Breakfast_1080-sdr.mkv,select=gt(scene\,.2)"
master:
pkt_pts=3587|tag:lavfi.scene_score=0.462364
pkt_pts=4838|tag:lavfi.scene_score=0.419519
pkt_pts=6548|tag:lavfi.scene_score=0.397027
pkt_pts=9968|tag:lavfi.scene_score=0.419245
pkt_pts=12471|tag:lavfi.scene_score=0.413084
pkt_pts=16225|tag:lavfi.scene_score=0.506370
pkt_pts=19645|tag:lavfi.scene_score=0.507538
pkt_pts=22314|tag:lavfi.scene_score=0.504319
pkt_pts=24817|tag:lavfi.scene_score=0.417544
pkt_pts=25651|tag:lavfi.scene_score=0.413916
pkt_pts=26652|tag:lavfi.scene_score=0.487707
18.58s user 0.07s system 99% cpu 18.663 total

patch applied:
pkt_pts=3587|tag:lavfi.scene_score=0.272173
pkt_pts=4838|tag:lavfi.scene_score=0.247841
pkt_pts=6548|tag:lavfi.scene_score=0.233134
pkt_pts=9968|tag:lavfi.scene_score=0.247253
pkt_pts=12471|tag:lavfi.scene_score=0.244129
pkt_pts=16225|tag:lavfi.scene_score=0.302531
pkt_pts=19645|tag:lavfi.scene_score=0.303560
pkt_pts=22314|tag:lavfi.scene_score=0.301861
pkt_pts=24817|tag:lavfi.scene_score=0.249331
pkt_pts=25651|tag:lavfi.scene_score=0.247096
pkt_pts=26652|tag:lavfi.scene_score=0.287728
10.90s user 0.06s system 99% cpu 10.967 total

3. AV_PIX_FMT_YUV422P

time ./ffprobe -of compact=p=0 -show_entries frame=pkt_pts:frame_tags -bitexact -f lavfi \
"sws_flags=+accurate_rnd+bitexact;movie=../Passengers_Breakfast_1080-sdr.mkv,format=yuv422p,select=gt(scene\,.2)"
master:

patched applied:
pkt_pts=3587|tag:lavfi.scene_score=0.224017
pkt_pts=4838|tag:lavfi.scene_score=0.204225
pkt_pts=9968|tag:lavfi.scene_score=0.204636
pkt_pts=12471|tag:lavfi.scene_score=0.202772
pkt_pts=16225|tag:lavfi.scene_score=0.248765
pkt_pts=19645|tag:lavfi.scene_score=0.250144
pkt_pts=22314|tag:lavfi.scene_score=0.248802
pkt_pts=24817|tag:lavfi.scene_score=0.208362
pkt_pts=25651|tag:lavfi.scene_score=0.205777
pkt_pts=26652|tag:lavfi.scene_score=0.230742

4.  AV_PIX_FMT_YUV420P10
time ./ffprobe -of compact=p=0 -show_entries frame=pkt_pts:frame_tags -bitexact -f lavfi \
"sws_flags=+accurate_rnd+bitexact;movie=../Passengers_Breakfast_4k.mkv,select=gt(scene\,.2)"

master:
pkt_pts=3587|tag:lavfi.scene_score=0.269890
pkt_pts=4838|tag:lavfi.scene_score=0.248957
pkt_pts=6548|tag:lavfi.scene_score=0.234619
pkt_pts=9969|tag:lavfi.scene_score=0.224912
pkt_pts=12471|tag:lavfi.scene_score=0.225158
pkt_pts=16225|tag:lavfi.scene_score=0.289809
pkt_pts=19645|tag:lavfi.scene_score=0.285013
pkt_pts=22314|tag:lavfi.scene_score=0.280295
pkt_pts=24817|tag:lavfi.scene_score=0.206486
pkt_pts=25651|tag:lavfi.scene_score=0.208556
pkt_pts=26652|tag:lavfi.scene_score=0.249577
./ffprobe -of compact=p=0 -show_entries frame=pkt_pts:frame_tags -bitexact -f  76.03s user 0.22s system 99% cpu 1:16.27 total

patch applied
pkt_pts=3587|tag:lavfi.scene_score=0.269890
pkt_pts=4838|tag:lavfi.scene_score=0.248957
pkt_pts=6548|tag:lavfi.scene_score=0.234619
pkt_pts=9969|tag:lavfi.scene_score=0.224912
pkt_pts=12471|tag:lavfi.scene_score=0.225158
pkt_pts=16225|tag:lavfi.scene_score=0.289809
pkt_pts=19645|tag:lavfi.scene_score=0.285013
pkt_pts=22314|tag:lavfi.scene_score=0.280295
pkt_pts=24817|tag:lavfi.scene_score=0.206486
pkt_pts=25651|tag:lavfi.scene_score=0.208556
pkt_pts=26652|tag:lavfi.scene_score=0.249577
./ffprobe -of compact=p=0 -show_entries frame=pkt_pts:frame_tags -bitexact -f  50.27s user 0.20s system 99% cpu 50.476 total

5. AV_PIX_FMT_RGBA, AV_PIX_FMT_ABGR, AV_PIX_FMT_BGRA, AV_PIX_FMT_GRAY8
are tested by format= with the fate sample: Vertical400kbit.sorenson3.mov like below:
time ./ffprobe -of compact=p=0 -show_entries frame=pkt_pts:frame_tags -bitexact -f lavfi \
"sws_flags=+accurate_rnd+bitexact;movie=../fate-suite/svq3/Vertical400kbit.sorenson3.mov,format=rgba,select=gt(scene\,.25)"

patch applied:
pkt_pts=1620|tag:lavfi.scene_score=1.000000
pkt_pts=4140|tag:lavfi.scene_score=0.656277
pkt_pts=5800|tag:lavfi.scene_score=1.000000
pkt_pts=6720|tag:lavfi.scene_score=0.346218
pkt_pts=8160|tag:lavfi.scene_score=0.987686
pkt_pts=9760|tag:lavfi.scene_score=1.000000
pkt_pts=14080|tag:lavfi.scene_score=0.629187
pkt_pts=15700|tag:lavfi.scene_score=1.000000
pkt_pts=18500|tag:lavfi.scene_score=0.356211
pkt_pts=20040|tag:lavfi.scene_score=0.284775
pkt_pts=21760|tag:lavfi.scene_score=1.000000

6. AV_PIX_FMT_YUVJ422P
time ./ffprobe -of compact=p=0 -show_entries frame=pkt_pts:frame_tags -bitexact -f lavfi \
"sws_flags=+accurate_rnd+bitexact;movie=../fate-suite/svq3/Vertical400kbit.sorenson3.mov,format=yuvj422p,select=gt(scene\,.25)"

patch applied:
pkt_pts=1620|tag:lavfi.scene_score=0.838281
pkt_pts=4140|tag:lavfi.scene_score=0.541382
pkt_pts=5800|tag:lavfi.scene_score=0.780588
pkt_pts=6720|tag:lavfi.scene_score=0.298274
pkt_pts=8160|tag:lavfi.scene_score=0.699106
pkt_pts=9760|tag:lavfi.scene_score=0.730136
pkt_pts=14080|tag:lavfi.scene_score=0.537742
pkt_pts=15700|tag:lavfi.scene_score=0.811946
pkt_pts=18500|tag:lavfi.scene_score=0.263382
pkt_pts=21760|tag:lavfi.scene_score=0.880773

Reviewed-by: Marton Balint <cus@passwd.hu>
Signed-off-by: Limin Wang <lance.lmwang@gmail.com>
 2019-07-22 22:54:51 +02:00
Limin Wang
53462cea2f avformat/f_select: support scenecut with more pixel formats 
This patch does not make other pixel formats usable yet to make sure the test
result is the same with rgb32 format.

Reviewed-by: Marton Balint <cus@passwd.hu>
Signed-off-by: Limin Wang <lance.lmwang@gmail.com>
 2019-07-22 22:54:51 +02:00
Michael Niedermayer
80bb65fafa Bump minor versions again on master to keep 4.2 versions separate from master 
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
 2019-07-21 18:36:31 +02:00
Michael Niedermayer
22db337a40 Bump minor versions to separate 4.2 from master 
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
 2019-07-21 18:36:18 +02:00
Jun Zhao
7eec3d22fc lavfi/showinfo: support regions of interest sidedata 
support regions of interest sidedata

Signed-off-by: Jun Zhao <barryjzhao@tencent.com>
 2019-07-21 10:05:27 +08:00
Paul B Mahol
97698b15a2 avfilter/vf_hysteresis: use time_base from framesync 
Fixes non-monotonous timestamps.
 2019-07-14 19:14:10 +02:00
Paul B Mahol
54fc8d370e avfilter/vf_remap: add option to control output format 2019-07-14 18:20:18 +02:00
Paul B Mahol
be822eb6f8 avfilter/vf_remap: use time_base from framesync 
Fixes non-monotonous timestamps.
 2019-07-14 18:20:18 +02:00
Paul B Mahol
077bf9058f avfilter/vf_displace: use time_base from framesync 
Fixes non-monotonous timestamps.
 2019-07-14 18:20:18 +02:00
Paul B Mahol
6882427841 avfilter/vf_midequalizer: always use framesync timestamps 2019-07-14 17:45:42 +02:00
Paul B Mahol
74236f675e avfilter/vf_maskedmerge: always use framesync timestamps 2019-07-14 17:38:26 +02:00
Paul B Mahol
3d78ad43f7 avfilter/vf_threshold: use time_base from framesync 
Fixes non-monotonous timestamps.
 2019-07-14 17:28:04 +02:00
Paul B Mahol
37e69b40df avfilter/vf_blend: add 9bit support 2019-07-14 17:00:58 +02:00
Paul B Mahol
80dacbedba avfilter/af_aiir: calculate group delay too 2019-07-13 15:45:06 +02:00
Paul B Mahol
60e6db2f5b avfilter/af_aiir: do not ignore k option for audio filtering 
Previously it was used only for displaying frequency response.
 2019-07-13 12:29:52 +02:00
Paul B Mahol
74d4fd0822 avfilter/avf_showfreqs: make selecting window size simpler 
The previous solution was very bad.
 2019-07-10 16:03:34 +02:00
Paul B Mahol
57a2688fe3 avfilter/af_afftfilt: make selecting window size simpler 
Next step after this one will be adding support for more window sizes.
 2019-07-10 15:57:38 +02:00
Paul B Mahol
9e78c73d86 avfilter/vf_readeia608: implement lowpass operation prior to processing lines 2019-07-08 19:21:59 +02:00
Paul B Mahol
dc481105a1 avfilter/vf_tinterlace: re-enable lowpass option 2019-07-08 17:57:31 +02:00
Paul B Mahol
2a801e8856 avfilter/af_aiir: implement mix option 2019-07-08 16:48:10 +02:00
Paul B Mahol
034a9d2507 avfilter/af_biquads: clip gain picked from command to sane values 2019-07-08 16:29:15 +02:00
Paul B Mahol
7b2d39fc27 avfilter/af_biquads: implement mix option to all filters 2019-07-08 16:20:57 +02:00
Calvin Walton
2bd21b9609 vf_drawtext: Add pkt_pos, pkt_duration, pkt_size as variables 
Changes to vf_drawtext.c written by
Calvin Walton <calvin.walton@kepstin.ca>

Changes to filters.texi written by
greg Luce <electron.rotoscope@gmail.com>
with lots of help from Moritz Barsnick and Gyan

Fixes #7947.
 2019-07-07 11:12:07 +05:30
Paul B Mahol
b9a6106842 avfilter/af_afftfilt: remove invalid scaling 2019-07-06 12:01:38 +02:00
Jun Zhao
3ed0b3f6f7 lavfi/scale: Add sub-options for in_color_matrix/out_color_matrix 
Add sub-options for in_color_matrix/out_color_matrix

Signed-off-by: Jun Zhao <barryjzhao@tencent.com>
 2019-07-06 13:12:13 +08:00
Paul B Mahol
0f39ef4db2 avfilter/vf_lut: fix regression with >8bit planar filtering 2019-07-05 12:55:02 +02:00
Yonglin Luo
664a27ea40 libavfilter/vf_colorspace.c: fix demarcation point of gamma linearize function 
The linearize function (usually refered to EOTF) is the inverse of
delinearize function (usually referred to OETF). Demarcation point of
EOTF should be beta*delta, but the actual value used now in the source
code is beta.

For ITU Rec.709, they are 0.081 (0.018*4.5) and 0.018 respectively
(beta = 0.018 and delta = 4.5), and they correspond to pixel value 5
and 21 for an 8-bit image. Linearized result of pixel within that range
(5-21) will be different, but this commit will make linearize function
of the filter more accurate in the mathematical sense.

Signed-off-by: Yonglin Luo <vincenluo@tencent.com>
Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
 2019-07-03 13:57:11 -04:00
Paul B Mahol
1ac643f066 avfilter/vf_lut2: use time_base from framesync 
Fixes non-monotonous timestamps.
 2019-07-03 14:54:05 +02:00
Paul B Mahol
cc5c636780 avfilter/vf_midequalizer: use time_base from framesync 
Fixes non-monotonous timestamps.
 2019-07-03 14:54:05 +02:00
Paul B Mahol
d49d7d238f avfilter/vf_blend: use time_base from framesync 
Fixes non-monotonous timestamps.
 2019-07-03 14:54:05 +02:00
Paul B Mahol
d71dafb2ab avfilter/vf_maskedclamp: use time_base from framesync 
Fixes non-monotonous timestamps.
 2019-07-03 14:54:05 +02:00
Paul B Mahol
8a0636a93b avfilter/vf_maskedmerge: use time_base from framesync 
Fixes non-monotonous timestamps.
 2019-07-03 14:54:05 +02:00
Paul B Mahol
93a73df54d avfilter/af_deesser: remove extra ; 2019-07-03 08:53:35 +02:00
Paul B Mahol
bd5e92ef8a avfilter: add deesser audio filter 2019-07-02 19:02:54 +02:00
Paul B Mahol
1b262004d3 avfilter/af_silenceremove: use input timestamp for timestamp recalculation 2019-07-02 19:01:42 +02:00
Zhong Li
dd662bbdd2 lavf/qsv_scale: add scaling modes support 
low_power mode will use a fixed HW engine (SFC), thus can offload EU usage.
high quality mode will take EU usage (AVS sampler).

Performance and EU usage (Render usage) comparsion on Intel(R) Xeon(R) CPU E3-1225 v5 @ 3.30GHz:

High quality mode : ffmpeg -hwaccel qsv -c:v h264_qsv -i bbb_sunflower_1080p_30fps_normal_2000frames.h264 \
-vf scale_qsv=w=1280:h=736:mode=hq -f null -
fps=389
RENDER usage: 28.10 (provided by MSDK metrics_monitor)

Low Power mode: ffmpeg -hwaccel qsv -c:v h264_qsv -i ~/bbb_sunflower_1080p_30fps_normal_2000frames.h264 \
-vf scale_qsv=w=1280:h=736:mode=low_power -f null -
fps=343
RENDER usage: 0.00

Low power mode (SFC) may be disabled if not supported by
MSDK/Driver/HW, and replaced by AVS mode interanlly.

Signed-off-by: Zhong Li <zhong.li@intel.com>
 2019-06-25 11:33:12 +08:00
Jun Zhao
ebcf4d354f lavfi/af_asetnsamples: Remove the redundant condition check 
Redundant condition: '!A || B' is equivalent to '!A || (A && B)' but
more clearly.

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Jun Zhao <barryjzhao@tencent.com>
 2019-06-18 23:27:07 +08:00
Limin Wang
c9c1711f49 libavfilter/vf_blend.c: remove duplicate code with macro-defined function 
Signed-off-by: Limin Wang <lance.lmwang@gmail.com>
 2019-06-14 13:37:45 +02:00
Limin Wang
24321b8751 libavfilter/vf_blend.c: remove duplicate code with same function 
Signed-off-by: Limin Wang <lance.lmwang@gmail.com>
 2019-06-14 13:37:35 +02:00
Zhong Li
4208b428c4 lavf/qsv_vpp: add frame format option 
1. Currently output format is hard-coded as NV12, thus means
   CSC is always done for not NV12 input such as P010.
   Follow original input format as default output.
2. Add an option to specify output format.

Signed-off-by: Zhong Li <zhong.li@intel.com>
 2019-06-14 10:27:33 +08:00
James Almer
b8f1542dcb avfilter/vf_gblur: add missing preprocessor check 
Fixes compilation on x86_32

Signed-off-by: James Almer <jamrial@gmail.com>
 2019-06-12 10:54:59 -03:00