1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-18 03:19:31 +02:00
Commit Graph

11291 Commits

Author SHA1 Message Date
Lynne
5f8feddd6a
lavfi/testsrc_vulkan: fix memory leaks 2023-08-28 22:29:36 +02:00
Lynne
3ef1e50c92
lavfi/transpose_vulkan: fix memory leaks 2023-08-28 22:29:36 +02:00
Lynne
d0ab2e2f98
lavfi/scale_vulkan: fix memory leaks 2023-08-28 22:29:35 +02:00
Lynne
0f2ae1ba69
lavfi/nlmeans_vulkan: fix memory leaks 2023-08-28 22:29:35 +02:00
Lynne
9944e96c61
lavfi/chromaber_vulkan: fix memory leaks 2023-08-28 22:29:34 +02:00
Lynne
a4673c9dff
lavfi/bwdif_vulkan: fix memory leaks 2023-08-28 22:29:34 +02:00
Lynne
b6cc53092a
lavfi/avgblur_vulkan: fix memory leaks 2023-08-28 22:29:34 +02:00
Niklas Haas
3c9dc009b6 lavfi/vf_libplacebo: add extra_opts AVDictionary
Can be used to configure libplacebo's underlying raw options, which
sometimes includes new or advanced / in-depth settings not (yet) exposed
by vf_libplacebo.
2023-08-27 13:37:03 +02:00
Niklas Haas
816983d951 lavfi/vf_libplacebo: switch to new pl_options struct
This new upstream struct simplifies params struct management by allowing
them to all be contained in a single dynamically allocated struct. This
commit switches to the new API in a backwards-compatible way.

The only nontrivial change that was required was to handle
`sigmoid_params` in a way consistent with the rest of the params
structs, instead of setting it directly to the upstream default.
2023-08-27 13:37:03 +02:00
Zhao Zhili
2387328fa2 avfilter/vf_transpose_vt: fix declaration-after-statement
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-08-26 16:31:43 +08:00
Zhao Zhili
7cca9c07ca avfilter/vf_transpose_vt: fix output frame dimension
Create a new hardware frame context when necessary.

Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-08-26 16:31:43 +08:00
Zhao Zhili
9b543881ef avfilter/vf_scale_vt: fix output frame dimension
The output frame dimension is incorrect because it shares hardware
frame context with input.

Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-08-26 16:31:43 +08:00
Wenbin Chen
e79bd1f1b1 lavfi/dnn: Add OpenVINO API 2.0 support
OpenVINO API 2.0 was released in March 2022, which introduced new
features.
This commit implements current OpenVINO features with new 2.0 APIs. And
will add other features in API 2.0.
Please add installation path, which include openvino.pc, to
PKG_CONFIG_PATH mannually for new OpenVINO libs config.

Signed-off-by: Ting Fu <ting.fu@intel.com>
Signed-off-by: Wenbin Chen <wenbin.chen@intel.com>
2023-08-26 14:12:10 +08:00
Marvin Scholz
15a89a9422 lavfi/vf_xfade_vulkan: add wipes
Add the wipetl, wipetr, wipebl, wipebr effects.
2023-08-24 22:50:57 +02:00
Marvin Scholz
85c7c79a62 lavfi/vf_xfade_vulkan: add pixelize 2023-08-24 22:50:57 +02:00
Marvin Scholz
98bd8a7f5c lavfi/vf_xfade_vulkan: add dissolve 2023-08-24 22:50:57 +02:00
Marvin Scholz
8d3fd489e2 lavfi/vf_xfade_vulkan: reindent after last commit 2023-08-24 22:50:57 +02:00
Marvin Scholz
95e366fb59 lavfi/vf_xfade_vulkan: add circleopen/circleclose 2023-08-24 22:50:57 +02:00
Paul B Mahol
52ebb15ec1 avfilter/af_afade: add 4 more curves 2023-08-24 22:20:06 +02:00
Evgeny Pavlov
cb1479faca avfilter/vf_ssim: Fix x86 assembly code for SSIM calculation
This commit fixes bug 

The code had several bugs related to post-loop compensation code:
- test assembly instruction performs bitwise AND operation and
generate flags used by jz branch instruction. Wrong test condition
leads to incorrect branching
- Incorrect compensation code for some branches

Signed-off-by: Evgeny Pavlov <lucenticus@gmail.com>
2023-08-21 17:04:51 +02:00
Wenbin Chen
9ae4863cc5 libavfilter/vf_vpp_qsv: Fix a timestamp bug when framerate is not set
If user doesn't set framerate when he creates a filter, the filter uses
default framerate {0, 1}. This causes error when setting timebase to
1/framerate. Now change it to pass inlink->time_base to outlink when
framerate is not set.
This patch fixes ticket:  

Signed-off-by: Wenbin Chen <wenbin.chen@intel.com>
2023-08-17 12:06:09 +08:00
L. E. Segovia
ddc1cd5cdd configure: Set WIN32_LEAN_AND_MEAN at configure time
Including winsock2.h or windows.h without WIN32_LEAN_AND_MEAN cause
bzlib.h to parse as nonsense, due to an instance of #define char small
in rpcndr.h.

See:

https://stackoverflow.com/a/27794577

Signed-off-by: L. E. Segovia <amy@amyspark.me>
Signed-off-by: Martin Storsjö <martin@martin.st>
2023-08-14 22:57:28 +03:00
Paul B Mahol
c704901324 Bump minor libavfilter version and add Changelog entry 2023-08-14 11:24:12 +02:00
Paul B Mahol
24c013369d avfilter/af_asdr: remove wrong scaling from sdr, and fix sisdr formula 2023-08-14 11:19:56 +02:00
Paul B Mahol
10110a30b5 avfilter/af_asdr: use single structure for sums 2023-08-14 11:19:56 +02:00
Paul B Mahol
e41d52216c avfilter: add asisdr filter 2023-08-14 11:19:56 +02:00
Paul B Mahol
951def850a avfilter: add apsnr filter 2023-08-14 11:19:55 +02:00
Paul B Mahol
a5b334e18e avfilter/af_astats: fix reporting of noise_floor for small number of samples 2023-08-13 14:38:28 +02:00
Paul B Mahol
90afcc2bfd avfilter/af_astats: drop usage of histogram for noise floor calculation 2023-08-13 14:24:47 +02:00
Paul B Mahol
80fdf51b8f avfilter/af_asdr: add fltp sample format support 2023-08-13 02:33:32 +02:00
Paul B Mahol
37b4837584 avfilter/af_asdr: stop recreating output timestamps 2023-08-13 02:33:30 +02:00
Paul B Mahol
9c694804fd avfilter/af_adeclick: stop returning EINVAL error
Instead clip relevant parameters.
2023-08-12 11:27:32 +02:00
Paul B Mahol
3e04f53571 avfilter/af_adeclick: stop outputing pointless message 2023-08-12 00:26:50 +02:00
Paul B Mahol
3d85892052 avfilter/af_adeclick: fix window generation
Stops adding invalid sinusoids in overlap-add mode.
2023-08-11 23:59:19 +02:00
Paul B Mahol
c7bfc826c3 avfilter/af_vibrato: refactor code
Move some code out of inner loop.
2023-08-08 17:45:32 +02:00
Paul B Mahol
482a22e10f avfilter/avf_showcwt: reduce latency/delay in some cases 2023-08-07 22:24:59 +02:00
Paul B Mahol
5ff88bb7bc avfilter/avf_showcwt: simplify 2023-08-07 22:24:59 +02:00
Paul B Mahol
de71928383 avfilter/avf_showcwt: add two more options 2023-08-07 22:24:58 +02:00
Paul B Mahol
d295b6b693 avfilter: fix highshelf zdf coefficients 2023-08-07 11:57:28 +02:00
Andreas Rheinhardt
19ffa2ff2d avfilter: Remove unnecessary formats.h inclusions
A filter needs formats.h iff it uses FILTER_QUERY_FUNC();
since lots of filters have been switched to use something
else than FILTER_QUERY_FUNC, they don't need it any more,
but removing this header has been forgotten.
This commit does this; files with formats.h inclusion went down
from 304 to 139 here (it were 449 before the preceding commit).

While just at it, also improve the other headers a bit.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2023-08-07 09:21:13 +02:00
Andreas Rheinhardt
19af142d45 avfilter/internal: Don't include formats.h
internal.h doesn't rely on it; instead include it directly
in every user that needs it (a filter needing it is basically
equivalent to it using FILTER_QUERY_FUNC, i.e. a majority of
filters doesn't need it).

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2023-08-07 09:21:13 +02:00
Andreas Rheinhardt
571d0359fa avfilter/vf_colorconstancy: Remove redundant CONFIG check
This file is only used by the greyedge filter and therefore
only compiled if said filter is enabled. This also allows
to remove a config_components.h inclusion, avoiding unnecessary
rebuilds.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2023-08-07 09:21:13 +02:00
Andreas Rheinhardt
2f62a433f2 avfilter: Deduplicate default video inputs/outputs
Lots of video filters use a very simple input or output:
An array with a single AVFilterPad whose name is "default"
and whose type is AVMEDIA_TYPE_VIDEO; everything else is unset.

Given that we never use pointer equality for inputs or outputs*,
we can simply use a single AVFilterPad instead of dozens; this
even saves .data.rel.ro (8312B here) as well as relocations.

*: In fact, several filters (like the filters in vf_lut.c)
already use the same outputs; furthermore, ff_filter_alloc()
duplicates the input and output pads so that we do not even
work with the pads directly.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2023-08-07 09:21:13 +02:00
Andreas Rheinhardt
6d15643173 avfilter/internal: Don't include video.h
internal.h does not depend on video.h (and should not depend on it)
and therefore should not include video.h at all; instead all users
of video.h should include it directly.

Doing so also avoids unnecessary video.h inclusions in files that
don't need it, like most audio filters.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2023-08-07 09:21:13 +02:00
Andreas Rheinhardt
50ea7389ec avfilter: Deduplicate default audio inputs/outputs
Lots of audio filters use very simple inputs or outputs:
An array with a single AVFilterPad whose name is "default"
and whose type is AVMEDIA_TYPE_AUDIO; everything else is unset.

Given that we never use pointer equality for inputs or outputs*,
we can simply use a single AVFilterPad instead of dozens; this
even saves .data.rel.ro (4784B here) as well as relocations.

*: In fact, several filters (like the filters in af_biquads.c)
already use the same inputs; furthermore, ff_filter_alloc()
duplicates the input and output pads so that we do not even
work with the pads directly.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2023-08-07 09:21:13 +02:00
Andreas Rheinhardt
1062a6d6c4 avfilter/avfilter: Make functions only used here static
Reviewed-by: Hendrik Leppkes <h.leppkes@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2023-08-04 11:49:26 +02:00
Andreas Rheinhardt
9732684ba2 avfilter/avfilter: Also deprecate variable name
Otherwise the var_names and the corresponding enum will be off
and e.g. the array holding the variable values will be too small.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2023-08-04 11:47:32 +02:00
Haihao Xiang
84fb7b9459 lavfi/vf_vpp_qsv: perform conversion from HDR to SDR
option tonemap is added to disable / enable tonemapping. By default
tonemapping is not performed.

Signed-off-by: Haihao Xiang <haihao.xiang@intel.com>
2023-08-04 10:27:55 +08:00
Haihao Xiang
b66aecf1b4 lavfi/vf_vpp_qsv: set color properties for output
User may set color range / matrix coefficient set / primaries / transfer
characteristics for output.

Signed-off-by: Haihao Xiang <haihao.xiang@intel.com>
2023-08-04 10:27:55 +08:00
Haihao Xiang
b42ea03f8b lavfi/vf_vpp_qsv: take input color properties into account
Signed-off-by: Haihao Xiang <haihao.xiang@intel.com>
2023-08-04 10:27:55 +08:00
Haihao Xiang
96b0b3f67a lavfi/qsvvpp: add set_frame_ext_params callback
This allows we add mfxExtBuffer per frame later.

Signed-off-by: Haihao Xiang <haihao.xiang@intel.com>
2023-08-04 10:27:55 +08:00
Haihao Xiang
567e78b283 lavfi/vf_vpp_qsv: fix duration in pass-through mode
Signed-off-by: Haihao Xiang <haihao.xiang@intel.com>
2023-07-31 09:09:20 +08:00
Paul B Mahol
8bbaecb5de avfilter/avf_showcwt: improve range detection for morlet scales 2023-07-30 19:57:50 +02:00
Paul B Mahol
48cd5561e1 avfilter/avf_showcwt: rename option value 2023-07-30 18:14:16 +02:00
Paul B Mahol
bd23b1d0ac avfilter/avf_showcwt: add iscale option 2023-07-30 18:14:16 +02:00
Paul B Mahol
6f552b9c5e avfilter/avf_showcwt: improve caching of some arrays 2023-07-30 18:14:15 +02:00
Paul B Mahol
4f8b0beeba avfilter/avf_showcwt: reduce excessive operations 2023-07-30 18:14:15 +02:00
Paul B Mahol
37d3000e3e avfilter/avf_showcwt: fix truncation of low Hz kernels 2023-07-29 17:16:42 +02:00
James Almer
075dfc271b avfilter/qsvvpp: remove usage of deprecated fields
Added by mistake in 88b3841149.

Reviewed-by: Xiang, Haihao <haihao.xiang@intel.com>
Signed-off-by: James Almer <jamrial@gmail.com>
2023-07-29 09:25:58 -03:00
Paul B Mahol
44813d66e2 avfilter/avf_showcwt: make kernels symmetrical 2023-07-29 11:30:16 +02:00
Paul B Mahol
6f6907fd20 avfilter/avf_showcwt: fix very low freqs leakage 2023-07-29 11:30:16 +02:00
Paul B Mahol
8c67e13473 avfilter/avf_showcwt: remove not needed fabsf() 2023-07-28 22:32:31 +02:00
Paul B Mahol
7d6dd90f3a avfilter/avf_showcwt: unbreak highest frequencies
The kernel was truncated too much.
2023-07-28 22:32:31 +02:00
Paul B Mahol
dca54600f0 avfilter/avf_showcwt: keep kernels precision as high as possible 2023-07-28 22:32:31 +02:00
Paul B Mahol
a7a46aff46 avfilter/avf_showcwt: remove excessive operations 2023-07-28 14:02:35 +02:00
Paul B Mahol
32a2268b6e avfilter/avf_showcwt: switch to overlap-save
And do both left and right zero-padding for forward FFT.
2023-07-28 14:02:34 +02:00
Paul B Mahol
830e522cde avfilter/avf_showcwt: reduce memory usage even more 2023-07-28 14:02:33 +02:00
Paul B Mahol
60244d0bec avfilter/af_afir: fix possible array read-overflow 2023-07-24 11:01:21 +02:00
Paul B Mahol
9fb2efe3b3 avfilter/avf_showcwt: make read only variable const 2023-07-24 11:01:21 +02:00
Wenbin Chen
13f49a7c62 libavfilter/vf_dnn_detect: bbox index may bigger than bbox number
Fix a bug that queried bbox index may bigger than bbox's total number.

Signed-off-by: Wenbin Chen <wenbin.chen@intel.com>
2023-07-24 09:34:11 +08:00
Niklas Haas
e9e37ea2f1 avfilter/vf_libplacebo: fix skip_aa option range
Max incorrectly specified as 0, should be 1.
2023-07-23 17:50:41 +02:00
Zhao Zhili
89f5124d0a avfilter: add transpose_vt for videotoolbox pix_fmt
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-07-23 11:52:28 +08:00
Zhao Zhili
c2c96c4c24 avfilter: add scale_vt for videotoolbox pix_fmt
For example,

./ffmpeg -hwaccel videotoolbox \
	-hwaccel_output_format videotoolbox_vld \
	-i ios-265.mov \
	-c:v hevc_videotoolbox \
	-profile:v main \
	-b:v 3M \
	-vf scale_vt=w=iw/2:h=ih/2:color_matrix=bt709:color_primaries=bt709:color_transfer=bt709 \
	-c:a copy \
	-tag:v hvc1 \
	/tmp/test.mp4

Input: hevc (Main 10) (hvc1 / 0x31637668), yuv420p10le(tv, bt2020nc/bt2020/arib-std-b67), 3840x2160
Output: hevc (Main) (hvc1 / 0x31637668), yuv420p(tv, bt709, progressive), 1920x1080
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-07-23 11:51:57 +08:00
Paul B Mahol
b2ceed406b avfilter/avf_showcwt: reduce memory usage 2023-07-22 22:27:57 +02:00
Paul B Mahol
ffa3edbf51 avfilter/avf_showcwt: extend min frequency accepted range
Also always check that min frequency is lower than max frequency.
2023-07-22 18:05:38 +02:00
Paul B Mahol
c24c7bcc95 avfilter/avf_showcwt: no need for big forward FFT with small sample rates 2023-07-22 18:05:38 +02:00
Paul B Mahol
e6168e43a2 avfilter/avf_showcwt: improve analysis
Make inverse FFT step always power of 2 in size.
2023-07-22 18:05:38 +02:00
Paul B Mahol
74474a551b avfilter/avf_showcwt: fix index storing 2023-07-21 00:44:01 +02:00
Paul B Mahol
86a1904f41 avfilter/avf_showcwt: move variables out of loop 2023-07-21 00:43:47 +02:00
Paul B Mahol
3dc93acc56 avfilter/vf_pseudocolor: add four more presets 2023-07-21 00:43:46 +02:00
Tobias Rapp
6747cda5ca avfilter/vf_overlay: Add support for yuv444p10 pixel format 2023-07-20 16:49:05 +02:00
Paul B Mahol
72390c7d80 avfilter/vf_pseudocolor: add heat preset 2023-07-19 23:42:03 +02:00
Paul B Mahol
5d8879208c avfilter/vf_pseudocolor: add cool preset 2023-07-19 23:42:01 +02:00
Paul B Mahol
c183f840fa avfilter/af_axcorrelate: add another algorithm for calculation
Rewrite EOF logic while here.
2023-07-15 23:41:09 +02:00
Timo Rothenpieler
3c73200be2 avfilter/vsrc_ddagrab: calculate pointer position on rotated screens 2023-07-15 18:48:49 +02:00
Timo Rothenpieler
da94c6a00c avfilter/vsrc_ddagrab: account for mouse-only frames during probing 2023-07-15 17:06:58 +02:00
Niklas Haas
138d3ac15d lavfi/vf_libplacebo: test for renderable pixel formats
Requires a new upstream function to test not for *import* support on a
given output pixel format, but also whether we can render to it.

Fixes: https://github.com/haasn/libplacebo/issues/173
2023-07-15 13:37:54 +02:00
Paul B Mahol
ef3b5789ec avfilter/af_atempo: improve RE_MALLOC_OR_FAIL macro
Make use of third parameter of av_calloc() call.
2023-07-15 10:52:15 +02:00
Paul B Mahol
374184a4dc avfilter/split: fix EOF passing to inlink 2023-07-14 00:46:45 +02:00
Paul B Mahol
7430ee2b8b avfilter/avf_showcwt: add rotation option 2023-07-13 00:11:36 +02:00
Paul B Mahol
422bfdc4ae avfilter/avf_showcwt: add bar display support 2023-07-13 00:11:34 +02:00
Anton Khirnov
fcc33ada47 lavfi: make sure frame SAR matches the link value 2023-07-11 19:28:18 +02:00
Anton Khirnov
bbff143b07 lavfi/src_movie: stop using AV_CODEC_FLAG_COPY_OPAQUE
That feature is overkill for a constant pointer to AVFilterLink which
can be stored in AVCodecContext.opaque (indirectly, because the link is
not allocated yet at the time the codec is opened).

This also avoids leaking non-NULL AVFrame.opaque to callers.
2023-07-11 19:14:23 +02:00
Paul B Mahol
f3896a148e avfilter/avf_showcwt: fix frame slide output 2023-07-10 18:57:14 +02:00
Paul B Mahol
1324698ff3 avfilter/avf_showcwt: simplify EOF checking 2023-07-10 18:57:13 +02:00
Paul B Mahol
2f2f76def7 avfilter/avf_showcwt: optimize multiplication with kernel 2023-07-10 18:57:12 +02:00
Paul B Mahol
587189cdb9 avfilter/avf_showcwt: simplify cache handling 2023-07-10 18:57:10 +02:00
Paul B Mahol
2c4fa968b6 avfilter/avf_showcwt: use av_rescale() 2023-07-10 18:57:08 +02:00
Paul B Mahol
fcbee7efdf avfilter/avf_showcwt: add qdrt frequency scale 2023-07-09 20:53:55 +02:00
Paul B Mahol
aee8689ba5 avfilter/avf_showcwt: add cbrt frequency scale 2023-07-09 20:53:54 +02:00
Paul B Mahol
7f23b72dde avfilter/avf_showcwt: add sqrt frequency scale 2023-07-09 20:53:52 +02:00
Paul B Mahol
20b8688092 avfilter/adynamicequalizer_template: fix peak filter 2023-07-09 11:58:22 +02:00
Paul B Mahol
889c2529cc avfilter/af_adynamicequalizer: simplify calculations 2023-07-09 11:58:03 +02:00
Niklas Haas
13e9899014 lavfi/vf_libplacebo: add contrast recovery options
New upstream option. Enabled by default in the high-quality preset
upstream, so enable it by default here.
2023-07-08 12:54:49 +02:00
Niklas Haas
2e83ba001d lavfi/vf_libplacebo: deprecate hybrid_mix option
Deprecated upstream in libplacebo v6.292.
2023-07-08 12:41:45 +02:00
John Cox
f00222e81f avfilter/vf_bwdif: Add neon for filter_line3
Signed-off-by: John Cox <jc@kynesim.co.uk>
Signed-off-by: Martin Storsjö <martin@martin.st>
2023-07-06 00:21:05 +03:00
John Cox
697533e76d avfilter/vf_bwdif: Add a filter_line3 method for optimisation
Add an optional filter_line3 to the available optimisations.

filter_line3 is equivalent to filter_line, memcpy, filter_line

filter_line shares quite a number of loads and some calculations in
common with its next iteration and testing shows that using aarch64
neon filter_line3s performance is 30% better than two filter_lines
and a memcpy.

Adds a test for vf_bwdif filter_line3 to checkasm

Rounds job start lines down to a multiple of 4. This means that if
filter_line3 exists then filter_line will not sometimes be called
once at the end of a slice depending on thread count. The final slice
may do up to 3 extra lines but filter_edge is faster than filter_line
so it is unlikely to create any noticable thread load variation.

Signed-off-by: John Cox <jc@kynesim.co.uk>
Signed-off-by: Martin Storsjö <martin@martin.st>
2023-07-06 00:21:05 +03:00
John Cox
94cb94a2c0 avfilter/vf_bwdif: Add neon for filter_line
Exports C filter_line needed for tail fixup of neon code
Adds neon for filter_line

Signed-off-by: John Cox <jc@kynesim.co.uk>
Signed-off-by: Martin Storsjö <martin@martin.st>
2023-07-06 00:21:05 +03:00
John Cox
8130df83e0 avfilter/vf_bwdif: Add neon for filter_edge
Adds clip and spatial macros for aarch64 neon
Exports C filter_edge needed for tail fixup of neon code
Adds neon for filter_edge

Signed-off-by: John Cox <jc@kynesim.co.uk>
Signed-off-by: Martin Storsjö <martin@martin.st>
2023-07-06 00:21:05 +03:00
John Cox
5075cfb4e6 avfilter/vf_bwdif: Add neon for filter_intra
Adds an outline for aarch neon functions
Adds common macros and consts for aarch64 neon
Exports C filter_intra needed for tail fixup of neon code
Adds neon for filter_intra

Signed-off-by: John Cox <jc@kynesim.co.uk>
Signed-off-by: Martin Storsjö <martin@martin.st>
2023-07-06 00:21:05 +03:00
Paul B Mahol
3bbb1c023b avfilter/avf_showcwt: fix offset to take into initial delay
Syncs audio with video.
2023-07-05 23:02:08 +02:00
Paul B Mahol
0cdfe46f4e avfilter/avf_showcwt: fix output frame pts if inlink timebase does not match sample rate 2023-07-05 22:56:05 +02:00
Paul B Mahol
a196d02e86 avfilter/af_aresample: switch to activate 2023-07-05 21:42:22 +02:00
Haihao Xiang
ca939ebb4f lavfi/tonemap_vaapi: remove HDR metadata
The HDR metadata should be removed after HDR to SDR conversion,
otherwise the output frame still has HDR side data.

Signed-off-by: Haihao Xiang <haihao.xiang@intel.com>
2023-07-05 09:34:19 +08:00
Paul B Mahol
8622dcb39b avfilter/adynamicequalizer_template: refactor and improve output 2023-07-05 01:53:04 +02:00
Lynne
9ff834c2a0
bwdif_vulkan: clamp the temporarily interpolated sample spatially
This makes the filter output match that of the C version.
It was left intentionally while we figured out if it was better
or not, and while it makes certain samples better, it makes static
samples jump around slightly.
2023-07-04 00:52:19 +02:00
Andreas Rheinhardt
652add3a6d avfilter/vf_ccrepack: Constify filter
The discrepancy between the definition and the declaration
in allfilters.c is actually UB.

Reviewed-by: James Almer <jamrial@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2023-06-29 20:22:51 +02:00
Fei Wang
13a10fe892 lavfi/{denoise, procamp, scale, sharpness}_vaapi: Add passthrough mode
Signed-off-by: Fei Wang <fei.w.wang@intel.com>
2023-06-25 10:01:51 +08:00
Fei Wang
f2d45bc565 lavfi/vaapi: Add some debug message
Signed-off-by: Fei Wang <fei.w.wang@intel.com>
2023-06-25 10:01:51 +08:00
Fei Wang
f565da8346 lavfi/vaapi: Add function to get surface ID from AVFrame
Signed-off-by: Fei Wang <fei.w.wang@intel.com>
2023-06-25 10:01:51 +08:00
Niklas Haas
5fdb12d6a0 lavfi/vf_libplacebo: simplify SAR normalization
The old logic was trying to be excessively clever in "deducing" that the
user wanted to stretch/scale the image when ow/oh differed from iw/ih
aspect ratio. But this is almost surely unintended except in
pathological cases, and in those cases users should simply disable
normalize_sar and do all the stretching/scaling logic themselves. This
is especially important in multi-input mode, where the canvas may be
vastly different from the input dimensions of any stream. Also, passing
through input 0 SAR in multi-input mode is arbitrary and nearly useless,
so again force output SAR to 1:1 here.
2023-06-23 15:18:25 +02:00
Niklas Haas
e998d0a4f7 lavfi/vf_libplacebo: allow pos_w to depend on pos_h
Ditto for crop_w/crop_h. Requires simply repeating the evaluation of the
first variable to be evaluated.
2023-06-20 17:09:58 +02:00
Niklas Haas
f998a618ba lavfi/vf_libplacebo: add nb_inputs option
To control the number of inputs.
2023-06-20 17:09:58 +02:00
Niklas Haas
9049ea6763 lavfi/vf_libplacebo: set time_base/frame_rate dynamically
Use the gcd of all input timebases to ensure PTS accuracy. For the
framerate, just pick the highest of all the inputs, under the assumption
that we will render frames with approximately this frequency. Of course,
this is not 100% accurate, in particular if the input frames are badly
misaligned. But this field is informational to begin with.

Importantly, it covers the "common" case of combining high FPS and low
FPS streams with aligned frames.
2023-06-20 17:09:58 +02:00
Niklas Haas
1b2c6c9a03 lavfi/vf_libplacebo: also skip cache if in FPS == out FPS
Fixes an oversight in the previous code which should have been >=, not >.
2023-06-20 17:09:58 +02:00
Niklas Haas
618b72d4b4 lavfi/vf_libplacebo: skip cache selectively per-input
It may be the case that we want to skip the single frame cache for some
inputs but not others.
2023-06-20 17:09:58 +02:00
Niklas Haas
015c3b659e lavfi/vf_libplacebo: set format list for all inputs 2023-06-20 17:09:58 +02:00
Niklas Haas
5989719e87 lavfi/vf_libplacebo: add in_idx variable
To allow placing an input dynamically, as a function of the input index.
2023-06-20 17:09:58 +02:00
Niklas Haas
7645c8df66 lavfi/vf_libplacebo: make input-dependent vars dynamic
Because these can differ based on the input, for multiple inputs.
2023-06-20 17:09:58 +02:00
Niklas Haas
d625010acd lavfi/vf_libplacebo: generalize frame update to multiple inputs
In the event that some frame mixes are OK while others are not, the
priority goes:

1. Errors in updating any frame -> return error
2. Any input incomplete -> request frames and return
3. Any inputs OK -> ignore EOF streams and render remaining inputs
4. No inputs OK -> set output to most recent status

This logic ensures that we can continue rendering the remaining streams,
no matter which streams reach their end of life, until we have no
streams left at which point we forward the last EOF.
2023-06-20 17:09:58 +02:00
Niklas Haas
6c41c3f928 lavfi/vf_libplacebo: only drain actually used PTS
When combining multiple inputs, the output PTS may be less than the PTS
of the input. In this case, the current's code assumption of always
draining one value from the FIFO is incorrect. Replace by a smarter
function which drains only those PTS values that were actually consumed.
2023-06-20 17:09:58 +02:00
Niklas Haas
502b699a50 lavfi/vf_libplacebo: determine PTS of next frame from any input
When combining multiple inputs with different PTS and durations, in
input-timed mode, we emit one output frame for every input frame PTS,
from *any* input. So when combining a low FPS stream with a high FPS
stream, the output framerate would match the higher FPS, independent of
which order they are specified in.
2023-06-20 17:09:58 +02:00
Niklas Haas
0eb37c2419 lavfi/vf_libplacebo: handle multiple inputs
This commit still relies on a single input for PTS determination, to be
changed in the next commit.
2023-06-20 17:09:58 +02:00
Niklas Haas
56e550b264 lavfi/vf_libplacebo: support blending multiple inputs
Subsequent inputs require frame blending to be enabled, in order to not
overwrite the existing frame contents.

For output metadata, we implicitly copy the metadata of the *first*
available stream (falling back to the second stream if the first has
already reached EOF, and so on). This is done to resolve any conflicts
between inputs with differing metadata. So when e.g. input 1 is HDR and
output 2 is SDR, the output will be HDR, and vice versa. This logic
could probablly be improved by dynamically determining some "superior"
set of metadata, but I don't want to handle that complexity in this
series.
2023-06-20 17:09:58 +02:00
Niklas Haas
0c66d912ba lavfi/vf_libplacebo: keep track of latest status globally
This field will effectively hold the most recent status set by any
input. Currently functionally equivalent to input->status, but will
change soon.
2023-06-20 17:09:57 +02:00
Niklas Haas
4e4a66b8c4 lavfi/vf_libplacebo: replace s->input by dynamic array
For now, hard-coded to 1 element.
2023-06-20 17:09:57 +02:00
Niklas Haas
a7775bd32c lavfi/vf_libplacebo: use correct link in update_crops()
Instead of hard-coding input 0, pass the per-input structure and use the
link contained inside it.
2023-06-20 17:09:57 +02:00
Niklas Haas
93c7e8c0ae lavfi/vf_libplacebo: factor out ref frame logic
Instead of finding the ref frame in output_frame() and then passing its
signature to update_crops(), pull out the logic and invoke it a second
time inside update_crops().

This may seem wasteful at present, but will actually become required in
the future, since update_crops() runs on *every* input, and needs values
specific to that input (which the signature isn't), while output_frame()
is only interested in a single input. It's much easier to just split the
logic cleanly.
2023-06-20 17:09:57 +02:00
Niklas Haas
666c8aa4d7 lavif/vf_libplacebo: remove pl_frame_mix from output_frame_mix
Instead, rename this function to `output_frame` and make it pull the
`pl_frame_mix` from the input structure. Step towards handling multiple
inputs.
2023-06-20 17:09:57 +02:00
Niklas Haas
ad445a7030 lavfi/vf_libplacebo: move temporary vars into per-input struct
Including the queue status, because these will need to be re-queried
inside output_frame_mix when that function is refactored to handle
multiple inputs.
2023-06-20 17:09:57 +02:00
Niklas Haas
a4197196a0 lavfi/vf_libplacebo: cosmetic
Assign local variable 'in' for 's->input' and replace 'inlink' by
'in->link' to avoid hard-coding ID 0 in more than one place.
2023-06-20 17:09:57 +02:00
Niklas Haas
b5d48111ad lavfi/vf_libplacebo: move input handling to separate function
To be re-used once we support more than one input.
2023-06-20 17:09:57 +02:00
Niklas Haas
c3b17ccc62 lavfi/vf_libplacebo: move input-specific state to struct
In anticipation of a refactor which will enable multiple input support.

Note: the renderer is also input-specific because it maintains a frame
cache, HDR peak detection state and mixing cache, all of which are tied
to a specific input stream.
2023-06-20 17:09:57 +02:00
Niklas Haas
7be4434c88 lavfi/vf_libplacebo: drop redundant case
If the input queue is EOF, then the s->status check should already have
covered it, and prevented the code from getting this far.

If we still hit this case for some reason, it's probably a bug. Better
to hit the AVERROR_BUG branch.
2023-06-20 17:09:57 +02:00
Haihao Xiang
88b3841149 lavfi/qsvvpp: use the right picture struct for vpp initilaization
Signed-off-by: Haihao Xiang <haihao.xiang@intel.com>
2023-06-20 09:52:43 +08:00
Haihao Xiang
bcdea21675 lavfi/qsvvpp: postpone vpp session initialization
So there is a chance to update vpp parameters per frame

Signed-off-by: Haihao Xiang <haihao.xiang@intel.com>
2023-06-20 09:52:43 +08:00
Haihao Xiang
cf21ca9efb lavfi/qsvvpp: store a copy of the sequence parameters
We will postpone the vpp session initialization to when input and output
frames are ready, this copy of the sequence parameters will be used to
initialize vpp session.

Signed-off-by: Haihao Xiang <haihao.xiang@intel.com>
2023-06-20 09:52:43 +08:00
Haihao Xiang
579711c2cc lavfi/qsvvpp: check the parameters before initializing vpp session
According to the description about MFXVideoVPP_Query [1], we may call
MFXVideoVPP_Query to check the validity of the parameters for vpp
session, use the corrected values to initialize the session.

[1] https://spec.oneapi.io/versions/latest/elements/oneVPL/source/API_ref/VPL_func_vid_vpp.html#mfxvideovpp-query

Signed-off-by: Haihao Xiang <haihao.xiang@intel.com>
2023-06-20 09:52:43 +08:00
Haihao Xiang
5189b51939 lavfi/qsvvpp: copy metadata fields from src to dst
Signed-off-by: Haihao Xiang <haihao.xiang@intel.com>
2023-06-20 09:52:43 +08:00
Haihao Xiang
07139b6deb lavfi/qsvvpp: track the runtime version in vpp context
We may check whether a feature is supported via the runtime version in
future.

Signed-off-by: Haihao Xiang <haihao.xiang@intel.com>
2023-06-20 09:52:43 +08:00
Paul B Mahol
ea190b8a23 avfilter/vf_drawtext: use flags type of option for text_align 2023-06-19 22:35:53 +02:00
yethie
01420316d2 avfilter/vf_drawtext: add support for commands 2023-06-19 22:35:53 +02:00
yethie
692d37d2e9 avfilter/vf_drawtext: add y_align option
The new y_align option specifies if the user provided y value
is referred to the top of the text, to the font baseline or to the
top of the font
2023-06-19 22:35:53 +02:00
yethie
5f2d907de2 avfilter/vf_drawtext: implement text alignment
Text can now be aligned vertically (top, middle, bottom) and horizontally (left, center, right)
relative to the background box.
2023-06-19 22:35:53 +02:00
yethie
01d843f6d0 avfilter/vf_drawtext: add expression variables to control font ascent/descent
The following new variables can be used in the x and y expressions: font_a, font_d, top_a, bottom_d.
2023-06-19 22:35:53 +02:00
yethie
fac7bb06d8 avfilter/vf_drawtext: add options for size of the background box 2023-06-19 22:35:53 +02:00
yethie
79e39978d8 avfilter/vf_drawtext: extend boxborderw option
Adds support for different border size in each direction.
2023-06-19 22:35:53 +02:00
yethie
1eeb59a209 avfilter/vf_drawtext: improve glyph shaping and positioning
- text is now shaped using libharfbuz
- glyphs position is now accurate to 1/4 pixel in both directions
- the default line height is now the one defined in the font

Adds libharfbuzz dependency.
2023-06-19 22:35:52 +02:00
yethie
ba00ed33e5 avfilter/vf_drawtext: cosmetics 2023-06-19 22:35:52 +02:00
Paul B Mahol
8e21f32fe7 avfilter/vf_displace: add support for commands 2023-06-18 11:11:46 +02:00
Paul B Mahol
54dc06e09e avfilter/vf_displace: add slice threading support 2023-06-18 11:11:46 +02:00
Paul B Mahol
494ebfdbfa avfilter/af_drmeter: misc cleanups 2023-06-18 02:21:29 +02:00
Paul B Mahol
238ec0c31b avfilter/af_drmeter: fix error in picking 2nd peak 2023-06-18 02:21:29 +02:00
Paul B Mahol
c44fe10160 avfilter/af_afade: stop using ff_outlink_get_status on inputs 2023-06-17 22:36:31 +02:00
Paul B Mahol
a756ae41ed avfilter/af_apad: switch to activate
Fixes EOF PTS reporting.
Also allowing previous/next filters in graph to release no longer needed resources.
2023-06-17 21:33:01 +02:00
Paul B Mahol
4583a24c65 avfilter/vf_xfade: add reveal transitions 2023-06-17 19:48:10 +02:00
Paul B Mahol
8fe25b1203 avfilter/vf_xfade: add cover transitions 2023-06-17 19:48:09 +02:00
Philip Langdale
73a2252f1d avfilter/vf_bwdif_cuda: CUDA accelerated bwdif deinterlacer
I've been sitting on this for 3 1/2 years now(!), and I finally got
around to fixing the loose ends and convincing myself that it was
correct. It follows the same basic structure as yadif_cuda, including
leaving out the edge handling, to avoid expensive branching.
2023-06-16 12:52:32 -07:00
Philip Langdale
3c07c2489d avfilter/scale_cuda: add support for rgb32/bgr32 conversions
As we are introducing two new formats and supporting conversions
between them, and also with the existing 0rgb32/0bgr32 formats, we get
a combinatorial explosion of kernels. I introduced a few new macros to
keep the things mostly managable.

The conversions are all simple, following existing patterns, with four
specific exceptions. When converting from 0rgb32/0bgr32 to rgb32/bgr32,
we need to ensure the alpha value is set to 1. In all other cases, it
can just be passed through, either to be used or ignored.
2023-06-16 10:40:21 -07:00
Paul B Mahol
12a073f72c avfilter/buffersrc: stop passing AV_NOPTS_VALUE on EOF 2023-06-16 11:46:09 +02:00
Anton Khirnov
4abf42ab96 lavfi/graphparser: be more verbose when filtergraph parsing fails
When an option could not be found, print its name and value. Note that
this is not done while applying the options in
avfilter_graph_segment_apply_opts() to give the caller the option of
handling the missing options in some other way.
2023-06-16 08:07:02 +02:00
Philip Langdale
ecc3f158e1 avfilter/vf_yadif_cuda: remove unnecessary private struct fields
I'm not sure why I originally did this, but there's no good reason to
put pointers to the cuda context and stream in the priv struct. They
are directly available in the device context that is already being
stored there.
2023-06-15 12:29:52 -07:00
Niklas Haas
d26ea03f60 lavfi/vf_libplacebo: fix typo in assert
Not caught because av_assert1 was compiled out on my end.
2023-06-13 15:28:32 +02:00
Zhao Zhili
32a749c7a6 avfilter/dnn_backend_openvino: fix log message
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-06-08 10:50:44 +08:00
Zhao Zhili
3a5d95e3fa avfilter/dnn_backend_tf: silence implicit cast warning
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-06-08 10:50:24 +08:00
Zhao Zhili
b0c0fedcda avfilter/dnn_backend_tf: fix use of uninitialized value
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-06-08 10:50:24 +08:00
Zhao Zhili
d9f41a343e avfilter/dnn_backend_tf: check TF_OperationOutputType return value
This also fixed a warning: implicit conversion from enumeration
type 'TF_DataType' (aka 'enum TF_DataType') to different
enumeration type 'DNNDataType'.

Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-06-08 10:50:24 +08:00
Zhao Zhili
f3495ef4f8 avfilter/dnn_backend_tf: remove unused define
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-06-08 10:50:23 +08:00
Zhao Zhili
e67af9e7f7 avfilter/vf_dnn_processing: replace magic number by enum value
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-06-08 10:50:23 +08:00
Zhao Zhili
e7e80e2083 avfilter/vf_dnn_detect: replace magic number by enum value
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-06-08 10:50:23 +08:00
Zhao Zhili
a887bc4ff1 avfilter/vf_dnn_classify: replace magic number by enum value
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-06-08 10:50:23 +08:00
Zhao Zhili
016f2f61c3 avfilter/dnn: add log context to ff_get_dnn_module
Print backend type when failed.

Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-06-08 10:50:23 +08:00
Zhao Zhili
505c43bb65 avfilter/dnn: refactor ff_get_dnn_module to remove allocation
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-06-08 10:50:23 +08:00
Zhao Zhili
3f52b7eedc avfilter/dnn: define each backend as a DNNModule
To avoid export multiple functions for each backend
implementation.

Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2023-06-08 10:50:23 +08:00
Paul B Mahol
bfd5090628 avfilter/af_rubberband: rewrite EOF handling
Also pass correct EOF timestamp.
2023-06-08 02:31:19 +02:00
Marvin Scholz
d8c7fea8ef lavfi/vf_blend_vulkan: fix leak on error 2023-06-07 23:59:06 +02:00
Marvin Scholz
dae393fe72 libavfilter: add vf_xfade_vulkan
This is an initial version of vf_xfade_vulkan based
on vf_xfade_opencl, for now only a subset of transitions
are supported.
2023-06-07 23:59:06 +02:00
Marvin Scholz
e62824835d lavfi/vf_xfade: rewrite activate inputs handling
The old code was not properly handling a bunch of edge-cases with
streams terminating earlier and also did not properly report back EOF
to the first input.

This fixes at least one case where the filter could stop doing
anything:

ffmpeg -f lavfi -i "color=blue:d=10" -f lavfi -i "color=aqua:d=0" -filter_complex "[0][1]xfade=duration=2:offset=0:transition=wiperight" -f null -
2023-06-07 08:56:24 +02:00
Paul B Mahol
9a9b56c870 avfilter/vf_mix: add fast filtering in case all tmix weights are same 2023-06-02 18:54:32 +02:00
Paul B Mahol
3cfd85ec20 avfilter/avf_showcqt: switch to activate
And make output frame durations always same.
2023-06-01 20:27:33 +02:00
Paul B Mahol
baa9fccf8d avfilter/vf_mergeplanes: remove hack for linesize calculation 2023-05-31 19:24:03 +02:00
Paul B Mahol
4d9afbeef5 avfilter/avf_aphasemeter: switch to activate 2023-05-30 20:09:47 +02:00
Paul B Mahol
3ea3a4ba16 avfilter/avf_aphasemeter: avoid using double 2023-05-30 20:09:46 +02:00
Paul B Mahol
862ebbf2ed avfilter/avf_abitscope: refactor bit counting for histogram
This helps compiler figure out it can unroll loop and give 4x speedup.
2023-05-30 19:21:13 +02:00
Paul B Mahol
1636fb415a avfilter/avf_ahistogram: refactor frame fill data code 2023-05-30 18:18:14 +02:00
Paul B Mahol
b01cfdb4ab avfilter/avf_showwaves: cleanup and simplify some draw calls 2023-05-30 17:00:30 +02:00
Sami Boukortt
06e8eedeba avfilter/vf_colorspace: fix message for missing/incorrect colorspace
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
2023-05-30 09:05:19 -04:00
Paul B Mahol
81f658b99d avfilter/f_ebur128: set output frame duration
Also produce output video timestamps that are consistent with
10 frames per second.
2023-05-30 14:08:40 +02:00
Paul B Mahol
447bdddeac avfilter/af_afir: do not use ff_outlink_get_status() on inlinks 2023-05-30 14:08:22 +02:00
Paul B Mahol
4ab1184fae avfilter/af_headphone: do not call ff_outlink_get_status() on inlink 2023-05-30 02:16:47 +02:00