Cin
Threads by month
- ----- 2025 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2024 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2023 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2022 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2021 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2020 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2019 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2018 -----
- December
- November
- 8 participants
- 1802 discussions
26 May '25
I have yet to get my system ffmpeg vulkan video transcoding to work on
my machines, while benchmarking works.
Therefore I hope get verified if my code line is correct, and how to fix
possible syntax errors.
-----------------
Dell XPS 13-9370 ultrabook
ffmpeg version 7.1.1 built with gcc 14 (SUSE Linux)
inxi -CGSz
System:
Kernel: 6.12.24-1.0.2.sr20250402-longterm arch: x86_64 bits: 64
Desktop: GNOME v: 48.0 Distro: openSUSE Tumbleweed-Slowroll 20250402
CPU:
Info: quad core model: Intel Core i7-8550U bits: 64 type: MT MCP cache:
L2: 1024 KiB
Speed (MHz): avg: 900 min/max: 400/4000 cores: 1: 900 2: 900 3: 900
4: 900
5: 900 6: 900 7: 900 8: 900
Graphics:
Device-1: Intel UHD Graphics 620 driver: i915 v: kernel
Device-2: Realtek Integrated_Webcam_HD driver: uvcvideo type: USB
Display: wayland server: X.org v: 1.21.1.15 with: Xwayland v: 24.1.6
compositor: gnome-shell driver: gpu: i915 resolution: 3840x2160~60Hz
API: OpenGL v: 4.6 vendor: intel mesa v: 25.0.3 renderer: Mesa Intel UHD
Graphics 620 (KBL GT2)
API: Vulkan v: 1.4.309 drivers: N/A surfaces: xcb,xlib,wayland
API: EGL Message: EGL data requires eglinfo. Check --recommends.
Info: Tools: api: glxinfo,vulkaninfo x11: xprop,xrandr
=========================
export VK_ICD_FILENAMES=/usr/share/vulkan/icd.d/intel_icd.x86_64.json
export ANV_VIDEO_DECODE=1
export ANV_VIDEO_ENCODE=1
Transcoding fails:
ffmpeg -hide_banner -init_hw_device vulkan=vk:0 -filter_hw_device vk -i
hdv09_04_h264_qsv_8b420.mp4 -c:v hevc_vulkan test_hevc.mp4
Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'hdv09_04_h264_qsv_8b420.mp4':
Metadata:
major_brand : isom
minor_version : 512
compatible_brands: isomiso2avc1mp41
encoder : Lavf61.1.100
Duration: 00:03:58.88, start: 0.000000, bitrate: 7671 kb/s
Stream #0:0[0x1](und): Video: h264 (High) (avc1 / 0x31637661),
yuvj420p(pc, smpte170m/unknown/unknown, progressive), 1440x1080 [SAR 4:3
DAR 16:9], 7536 kb/s, 25 fps, 25 tbr, 12800 tbn (default)
Metadata:
handler_name : VideoHandler
vendor_id : [0][0][0][0]
Stream #0:1[0x2](eng): Audio: aac (LC) (mp4a / 0x6134706D), 48000 Hz,
stereo, fltp, 129 kb/s (default)
Metadata:
handler_name : SoundHandler
vendor_id : [0][0][0][0]
Stream mapping:
Stream #0:0 -> #0:0 (h264 (native) -> hevc (hevc_vulkan))
Stream #0:1 -> #0:1 (aac (native) -> aac (native))
Press [q] to stop, [?] for help
Impossible to convert between the formats supported by the filter
'Parsed_null_0' and the filter 'auto_scale_0'
[vf#0:0 @ 0x561e923dd540] Error reinitializing filters!
[vf#0:0 @ 0x561e923dd540] Task finished with error code: -38 (Function
not implemented)
[vf#0:0 @ 0x561e923dd540] Terminating thread with return code -38
(Function not implemented)
[vost#0:0/hevc_vulkan @ 0x561e9243b200] Could not open encoder before EOF
[vost#0:0/hevc_vulkan @ 0x561e9243b200] Task finished with error code:
-22 (Invalid argument)
[vost#0:0/hevc_vulkan @ 0x561e9243b200] Terminating thread with return
code -22 (Invalid argument)
[out#0/mp4 @ 0x561e924a7a00] Nothing was written into output file,
because at least one of its streams received no packets.
frame= 0 fps=0.0 q=0.0 Lsize= 0KiB time=N/A bitrate=N/A speed=N/A
[aac @ 0x561e92cc4f80] Qavg: 208.809
Conversion failed!
=====================
Benchmarking works:
ffmpeg -hide_banner -init_hw_device "vulkan=vk:0" -hwaccel vulkan
-hwaccel_output_format vulkan -i hdv09_04_h264_qsv_8b420.mp4 -f null -
-benchmark
Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'hdv09_04_h264_qsv_8b420.mp4':
Metadata:
major_brand : isom
minor_version : 512
compatible_brands: isomiso2avc1mp41
encoder : Lavf61.1.100
Duration: 00:03:58.88, start: 0.000000, bitrate: 7671 kb/s
Stream #0:0[0x1](und): Video: h264 (High) (avc1 / 0x31637661),
yuvj420p(pc, smpte170m/unknown/unknown, progressive), 1440x1080 [SAR 4:3
DAR 16:9], 7536 kb/s, 25 fps, 25 tbr, 12800 tbn (default)
Metadata:
handler_name : VideoHandler
vendor_id : [0][0][0][0]
Stream #0:1[0x2](eng): Audio: aac (LC) (mp4a / 0x6134706D), 48000 Hz,
stereo, fltp, 129 kb/s (default)
Metadata:
handler_name : SoundHandler
vendor_id : [0][0][0][0]
Stream mapping:
Stream #0:0 -> #0:0 (h264 (native) -> wrapped_avframe (native))
Stream #0:1 -> #0:1 (aac (native) -> pcm_s16le (native))
Press [q] to stop, [?] for help
Output #0, null, to 'pipe:':
Metadata:
major_brand : isom
minor_version : 512
compatible_brands: isomiso2avc1mp41
encoder : Lavf61.7.100
Stream #0:0(und): Video: wrapped_avframe, vulkan(pc,
smpte170m/unknown/unknown, progressive), 1440x1080 [SAR 4:3 DAR 16:9],
q=2-31, 200 kb/s, 25 fps, 25 tbn (default)
Metadata:
handler_name : VideoHandler
vendor_id : [0][0][0][0]
encoder : Lavc61.19.101 wrapped_avframe
Stream #0:1(eng): Audio: pcm_s16le, 48000 Hz, stereo, s16, 1536 kb/s
(default)
Metadata:
handler_name : SoundHandler
vendor_id : [0][0][0][0]
encoder : Lavc61.19.101 pcm_s16le
4
11
According to Arch Wiki
https://wiki.archlinux.org/title/Hardware_video_acceleration
you need
MPEG-4 Part 2 is disabled by default due to VAAPI limitations. Set the
environment
variable <https://wiki.archlinux.org/title/Environment_variable>
VAAPI_MPEG4_ENABLED=true to try to use it anyway.
I tried with cingg/ffmpeg7 and it seemed to work, but ffmpeg7 had all our
patches including patch8
2
1
So I was experimenting with Vulkan decoding in cinelerra-gg.
After some fight I get build using ffmpeg git
commit 038314bc6be2f35a82e9fba2228bcac2e4fee648 for ffmpeg
here is bunch of errors like this:
[hevc @ 0x6f7465c0] Could not find ref with POC 296
[hevc @ 0x6f7465c0] Error constructing the frame RPS.
[hevc @ 0x6f7465c0] Skipping invalid undecodable NALU: 9
[hevc @ 0x6f72ad00] Could not find ref with POC 296
[hevc @ 0x6f72ad00] Error constructing the frame RPS.
[hevc @ 0x6f72ad00] Skipping invalid undecodable NALU: 8
[hevc @ 0x6f77b3c0] Could not find ref with POC 298
[hevc @ 0x6f77b3c0] Error constructing the frame RPS.
[hevc @ 0x6f77b3c0] Skipping invalid undecodable NALU: 8
[hevc @ 0x6f789480] Could not find ref with POC 300
[hevc @ 0x6f789480] Error constructing the frame RPS.
[hevc @ 0x6f789480] Skipping invalid undecodable NALU: 1
[hevc @ 0x6f7465c0] Could not find ref with POC 300
[hevc @ 0x6f7465c0] Error constructing the frame RPS.
[hevc @ 0x6f7465c0] Skipping invalid undecodable NALU: 3
[hevc @ 0x6f72ad00] Could not find ref with POC 300
[hevc @ 0x6f72ad00] Error constructing the frame RPS.
[hevc @ 0x6f72ad00] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f77b3c0] Could not find ref with POC 302
[hevc @ 0x6f77b3c0] Error constructing the frame RPS.
[hevc @ 0x6f77b3c0] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f789480] Could not find ref with POC 304
[hevc @ 0x6f789480] Error constructing the frame RPS.
[hevc @ 0x6f789480] Skipping invalid undecodable NALU: 1
[hevc @ 0x6f7465c0] Could not find ref with POC 304
[hevc @ 0x6f7465c0] Error constructing the frame RPS.
[hevc @ 0x6f7465c0] Skipping invalid undecodable NALU: 3
[hevc @ 0x6f72ad00] Could not find ref with POC 304
[hevc @ 0x6f72ad00] Error constructing the frame RPS.
[hevc @ 0x6f72ad00] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f77b3c0] Could not find ref with POC 306
[hevc @ 0x6f77b3c0] Error constructing the frame RPS.
[hevc @ 0x6f77b3c0] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f789480] Could not find ref with POC 308
[hevc @ 0x6f789480] Error constructing the frame RPS.
[hevc @ 0x6f789480] Skipping invalid undecodable NALU: 1
[hevc @ 0x6f7465c0] Could not find ref with POC 308
[hevc @ 0x6f7465c0] Error constructing the frame RPS.
[hevc @ 0x6f7465c0] Skipping invalid undecodable NALU: 3
[hevc @ 0x6f72ad00] Could not find ref with POC 308
[hevc @ 0x6f72ad00] Error constructing the frame RPS.
[hevc @ 0x6f72ad00] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f77b3c0] Could not find ref with POC 310
[hevc @ 0x6f77b3c0] Error constructing the frame RPS.
[hevc @ 0x6f77b3c0] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f789480] Could not find ref with POC 312
[hevc @ 0x6f789480] Error constructing the frame RPS.
[hevc @ 0x6f789480] Skipping invalid undecodable NALU: 1
[hevc @ 0x6f7465c0] Could not find ref with POC 312
[hevc @ 0x6f7465c0] Error constructing the frame RPS.
[hevc @ 0x6f7465c0] Skipping invalid undecodable NALU: 3
[hevc @ 0x6f72ad00] Could not find ref with POC 312
[hevc @ 0x6f72ad00] Error constructing the frame RPS.
[hevc @ 0x6f72ad00] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f77b3c0] Could not find ref with POC 314
[hevc @ 0x6f77b3c0] Error constructing the frame RPS.
[hevc @ 0x6f77b3c0] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f789480] Could not find ref with POC 316
[hevc @ 0x6f789480] Error constructing the frame RPS.
[hevc @ 0x6f789480] Skipping invalid undecodable NALU: 1
[hevc @ 0x6f7465c0] Could not find ref with POC 316
[hevc @ 0x6f7465c0] Error constructing the frame RPS.
[hevc @ 0x6f7465c0] Skipping invalid undecodable NALU: 3
[hevc @ 0x6f72ad00] Could not find ref with POC 316
[hevc @ 0x6f72ad00] Error constructing the frame RPS.
[hevc @ 0x6f72ad00] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f77b3c0] Could not find ref with POC 318
[hevc @ 0x6f77b3c0] Error constructing the frame RPS.
[hevc @ 0x6f77b3c0] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f789480] Could not find ref with POC 320
[hevc @ 0x6f789480] Error constructing the frame RPS.
[hevc @ 0x6f789480] Skipping invalid undecodable NALU: 1
[hevc @ 0x6f7465c0] Could not find ref with POC 320
[hevc @ 0x6f7465c0] Error constructing the frame RPS.
[hevc @ 0x6f7465c0] Skipping invalid undecodable NALU: 3
[hevc @ 0x6f72ad00] Could not find ref with POC 320
[hevc @ 0x6f72ad00] Error constructing the frame RPS.
[hevc @ 0x6f72ad00] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f77b3c0] Could not find ref with POC 322
[hevc @ 0x6f77b3c0] Error constructing the frame RPS.
[hevc @ 0x6f77b3c0] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f789480] Could not find ref with POC 324
[hevc @ 0x6f789480] Error constructing the frame RPS.
[hevc @ 0x6f789480] Skipping invalid undecodable NALU: 1
[hevc @ 0x6f7465c0] Could not find ref with POC 324
[hevc @ 0x6f7465c0] Error constructing the frame RPS.
[hevc @ 0x6f7465c0] Skipping invalid undecodable NALU: 3
[hevc @ 0x6f72ad00] Could not find ref with POC 324
[hevc @ 0x6f72ad00] Error constructing the frame RPS.
[hevc @ 0x6f72ad00] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f77b3c0] Could not find ref with POC 326
[hevc @ 0x6f77b3c0] Error constructing the frame RPS.
[hevc @ 0x6f77b3c0] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f789480] Could not find ref with POC 328
[hevc @ 0x6f789480] Error constructing the frame RPS.
[hevc @ 0x6f789480] Skipping invalid undecodable NALU: 1
[hevc @ 0x6f7465c0] Could not find ref with POC 328
[hevc @ 0x6f7465c0] Error constructing the frame RPS.
[hevc @ 0x6f7465c0] Skipping invalid undecodable NALU: 3
[hevc @ 0x6f72ad00] Could not find ref with POC 328
[hevc @ 0x6f72ad00] Error constructing the frame RPS.
[hevc @ 0x6f72ad00] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f77b3c0] Could not find ref with POC 330
[hevc @ 0x6f77b3c0] Error constructing the frame RPS.
[hevc @ 0x6f77b3c0] Skipping invalid undecodable NALU: 2
[hevc @ 0x6f789480] Could not find ref with POC 332
[hevc @ 0x6f789480] Error constructing the frame RPS.
[hevc @ 0x6f789480] Skipping invalid undecodable NALU: 1
[hevc @ 0x6f7465c0] Could not find ref with POC 332
[hevc @ 0x6f7465c0] Error constructing the frame RPS.
[hevc @ 0x6f7465c0] Skipping invalid undecodable NALU: 3
file itself:
https://drive.google.com/file/d/1Ic9DZXMSo07EJMqCFaQRKSSrSw6y1mYv/view
But I see some kind of first frame corruption I never saw on 7.0 build
https://paste.pics/860707666060d195485e07ea7b057ce8
and speed kinda down from 58 fps to just 50-51 relative to ffmpeg libs 7.0?
hw:
01:00.0 VGA compatible controller: Advanced Micro Devices, Inc.
[AMD/ATI] Lexa PRO [Radeon 540/540X/550/550X / RX 540X/550/550X] (rev
c7)
mesa:
OpenGL renderer string: AMD Radeon RX 550 / 550 Series (radeonsi,
polaris12, ACO, DRM 3.61, 6.12.26-x64)
OpenGL core profile version string: 4.6 (Core Profile) Mesa
25.2.0-devel (git-ef63e3e4d2)
cingg home git:
git://git.cinelerra-gg.org/goodguy/cinelerra.git
I set two evn. variables specific to my system, but mostly I wanted to
get libplacebo and libzimg for HDR->SDR conversion
EXTRA_LIBS := -lOpenCL -lSvtAv1Enc -lvpl -ldav1d -lxvidcore -lass
-lbluray -lsnappy -lzimg -lplacebo -lshaderc_shared -lnuma -lva
-lva-x11 `pkg-config --libs x11` -lva-drm -ldl
FFMPEG_EXTRA_CFG := --enable-libvpl --disable-doc --enable-opencl
--enable-libsvtav1 --enable-frei0r --enable-libdav1d --enable-libzimg
--enable-libxvid --enable-libass --enable-libbluray --enable-libsnappy
--enable-libshaderc --enable-libplacebo --enable-vulkan
--disable-debug --extra-cflags=-I/usr/local/include/vpl
--extra-cflags=-I/usr/include/svt-av1 --extra-ldflags=" -lva
`pkg-config --libs libva` -lva-x11 -lva-drm -ldl"
cingg configure:
./configure --with-git-ffmpeg=https://git.ffmpeg.org/ffmpeg.git
--with-single-user --enable-libsvtav1 --disable-dav1d
I used bunch of patches (attached, done by Phyllis and me) and
manually removed posprocess.a ref. from our configure.ac for now
there is still some problem with attaching ffmpeg filters
per-input-file, but main Vulkan decode problem still visible without
any filters
I run just-compiled cin binary like this:
RADV_PERFTEST=video_decode bin/cin
~/K38_sdcard1/Documents/iPhone11_4K-recorder_59.940HDR10.mov
and got corruption as visible on screenshot at the start of this email
Playing image forward cleans up this corruption, but returning to
timeline beginning bring it back.
3
11
You rarely see this Amiga software in action, but there seems to be few
longer recordings of it on this channel
https://m.youtube.com/watch?v=gNS6TxrK_NU
2
1
I think while this script was added as default shell command it was never
installed.
Attached patch fixes this.
Can anyone test please? ;)
2
3
forum
https://forum.doom9.org/showthread.php?t=176091
it says those LUTs should be usable for ffmpeg's lut3d and therefore for
CinGG?
github (each *.cube can be several mb's big)
https://github.com/FranceBB/LinearTransformation
author says they were verified on professional equipment...
list includes PAL and NTSC variants for rec. 601....
3
6
Even if our code compiles it does not work anymore ...
[out @ 0xe1821180] Option 'pix_fmts' is not a runtime option and so
cannot be set after the object has been initialized
int ret = 0; char args[BCTEXTLEN];
AVPixelFormat pix_fmt = (AVPixelFormat)avpar->format;
snprintf(args, sizeof(args),
"video_size=%dx%d:pix_fmt=%i:time_base=%d/%d:pixel_aspect=%d/%d",
avpar->width, avpar->height,(int)pix_fmt,
st->time_base.num, st->time_base.den, sa_num, sa_den);
if( ret >= 0 ) {
filt_ctx = 0;
ret = insert_filter("buffer", args, "in");
buffersrc_ctx = filt_ctx;
}
if( ret >= 0 )
ret = flip(theta);
AVFilterContext *fsrc = filt_ctx;
if( ret >= 0 ) {
filt_ctx = 0;
ret = insert_filter("buffersink", 0, "out");
buffersink_ctx = filt_ctx;
}
if( ret >= 0 ) {
ret = av_opt_set_bin(buffersink_ctx, "pix_fmts",
(uint8_t*)&pix_fmt, sizeof(pix_fmt),
AV_OPT_SEARCH_CHILDREN);
}
if( ret >= 0 )
ret = config_filters(filter_spec, fsrc);
else
ff_err(ret, "FFVideoStream::create_filter");
return ret >= 0 ? 0 : -1;
=========
There is no video filtering example in doc/examples.
How I supposed to know what should be put there now?
2
1
14 May '25
вт, 6 мая 2025 г., 02:27 Michael Niedermayer <michael(a)niedermayer.cc>:
> This will be available in https://github.com/michaelni/libpostproc
> either as a separate library or a ffmpeg source plugin whatever turns
> out more convenient to maintain
>
Congratulations, you broke building cinelerra-gg with ffmpeg.git despite
our best efforts :/
Why all this code movement?!
For whom it "simple"?
>
> Sponsored-by: Sovereign Tech Fund
> Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc>
> ---
> LICENSE.md | 2 -
> MAINTAINERS | 2 -
> configure | 13 +-
> doc/filters.texi | 175 --
> fftools/ffprobe.c | 3 -
> fftools/opt_common.c | 3 -
> libavfilter/Makefile | 1 -
> libavfilter/allfilters.c | 1 -
> libavfilter/vf_pp.c | 191 --
> libavutil/avutil.h | 1 -
> libpostproc/Makefile | 23 -
> libpostproc/libpostproc.v | 7 -
> libpostproc/postprocess.c | 992 ------
> libpostproc/postprocess.h | 109 -
> libpostproc/postprocess_altivec_template.c | 1214 --------
> libpostproc/postprocess_internal.h | 185 --
> libpostproc/postprocess_template.c | 3293 --------------------
> libpostproc/postprocres.rc | 55 -
> libpostproc/tests/.gitignore | 3 -
> libpostproc/tests/blocktest.c | 113 -
> libpostproc/tests/stripetest.c | 115 -
> libpostproc/tests/temptest.c | 106 -
> libpostproc/tests/test_utils.c | 38 -
> libpostproc/tests/test_utils.h | 27 -
> libpostproc/version.c | 45 -
> libpostproc/version.h | 48 -
> libpostproc/version_major.h | 33 -
> tests/Makefile | 1 -
> tests/fate/filter-video.mak | 16 -
> tests/fate/libpostproc.mak | 14 -
> tests/ref/fate/blocktest | 1800 -----------
> tests/ref/fate/filter-pp | 10 -
> tests/ref/fate/filter-pp1 | 1 -
> tests/ref/fate/filter-pp2 | 1 -
> tests/ref/fate/filter-pp3 | 1 -
> tests/ref/fate/filter-pp4 | 1 -
> tests/ref/fate/filter-pp5 | 1 -
> tests/ref/fate/filter-pp6 | 1 -
> tests/ref/fate/filter-qp | 1 -
> tests/ref/fate/stripetest | 360 ---
> tests/ref/fate/temptest | 336 --
> tools/gen-rc | 1 -
> 42 files changed, 1 insertion(+), 9342 deletions(-)
> delete mode 100644 libavfilter/vf_pp.c
> delete mode 100644 libpostproc/Makefile
> delete mode 100644 libpostproc/libpostproc.v
> delete mode 100644 libpostproc/postprocess.c
> delete mode 100644 libpostproc/postprocess.h
> delete mode 100644 libpostproc/postprocess_altivec_template.c
> delete mode 100644 libpostproc/postprocess_internal.h
> delete mode 100644 libpostproc/postprocess_template.c
> delete mode 100644 libpostproc/postprocres.rc
> delete mode 100644 libpostproc/tests/.gitignore
> delete mode 100644 libpostproc/tests/blocktest.c
> delete mode 100644 libpostproc/tests/stripetest.c
> delete mode 100644 libpostproc/tests/temptest.c
> delete mode 100644 libpostproc/tests/test_utils.c
> delete mode 100644 libpostproc/tests/test_utils.h
> delete mode 100644 libpostproc/version.c
> delete mode 100644 libpostproc/version.h
> delete mode 100644 libpostproc/version_major.h
> delete mode 100644 tests/fate/libpostproc.mak
> delete mode 100644 tests/ref/fate/blocktest
> delete mode 100644 tests/ref/fate/filter-pp
> delete mode 100644 tests/ref/fate/filter-pp1
> delete mode 100644 tests/ref/fate/filter-pp2
> delete mode 100644 tests/ref/fate/filter-pp3
> delete mode 100644 tests/ref/fate/filter-pp4
> delete mode 100644 tests/ref/fate/filter-pp5
> delete mode 100644 tests/ref/fate/filter-pp6
> delete mode 100644 tests/ref/fate/filter-qp
> delete mode 100644 tests/ref/fate/stripetest
> delete mode 100644 tests/ref/fate/temptest
>
> diff --git a/LICENSE.md b/LICENSE.md
> index 613070e1b63..371b0913ce7 100644
> --- a/LICENSE.md
> +++ b/LICENSE.md
> @@ -12,7 +12,6 @@ configure to activate them. In this case, FFmpeg's
> license changes to GPL v2+.
>
> Specifically, the GPL parts of FFmpeg are:
>
> -- libpostproc
> - optional x86 optimization in the files
> - `libavcodec/x86/flac_dsp_gpl.asm`
> - `libavcodec/x86/idct_mmx.c`
> @@ -45,7 +44,6 @@ Specifically, the GPL parts of FFmpeg are:
> - `vf_owdenoise.c`
> - `vf_perspective.c`
> - `vf_phase.c`
> - - `vf_pp.c`
> - `vf_pp7.c`
> - `vf_pullup.c`
> - `vf_repeatfields.c`
> diff --git a/MAINTAINERS b/MAINTAINERS
> index f58936db619..7f7faa05959 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -134,8 +134,6 @@ Generic Parts:
> ratecontrol.c [2] Michael Niedermayer
> simple IDCT:
> simple_idct.c, simple_idct.h [2] Michael Niedermayer
> - postprocessing:
> - libpostproc/* [2] Michael Niedermayer
> table generation:
> tableprint.c, tableprint.h Reimar Doeffinger
> fixed point FFT:
> diff --git a/configure b/configure
> index 6c23a38be1f..2e69b3c56c3 100755
> --- a/configure
> +++ b/configure
> @@ -130,7 +130,6 @@ Component options:
> --disable-avformat disable libavformat build
> --disable-swresample disable libswresample build
> --disable-swscale disable libswscale build
> - --disable-postproc disable libpostproc build
> --disable-avfilter disable libavfilter build
> --disable-pthreads disable pthreads [autodetect]
> --disable-w32threads disable Win32 threads [autodetect]
> @@ -2072,7 +2071,6 @@ LIBRARY_LIST="
> avdevice
> avfilter
> swscale
> - postproc
> avformat
> avcodec
> swresample
> @@ -3969,7 +3967,6 @@ pan_filter_deps="swresample"
> perspective_filter_deps="gpl"
> phase_filter_deps="gpl"
> pp7_filter_deps="gpl"
> -pp_filter_deps="gpl postproc"
> prewitt_opencl_filter_deps="opencl"
> procamp_vaapi_filter_deps="vaapi"
> program_opencl_filter_deps="opencl"
> @@ -4086,8 +4083,6 @@ avfilter_suggest="libm stdatomic spirv_compiler"
> avformat_deps="avcodec avutil"
> avformat_suggest="libm network zlib stdatomic"
> avutil_suggest="clock_gettime ffnvcodec gcrypt libm libdrm libmfx opencl
> openssl user32 vaapi vulkan videotoolbox corefoundation corevideo coremedia
> bcrypt stdatomic"
> -postproc_deps="avutil gpl"
> -postproc_suggest="libm stdatomic"
> swresample_deps="avutil"
> swresample_suggest="libm libsoxr stdatomic"
> swscale_deps="avutil"
> @@ -7533,7 +7528,7 @@ void (^block)(void);
> EOF
>
> # add some linker flags
> -check_ldflags
> -Wl,-rpath-link=:libpostproc:libswresample:libswscale:libavfilter:libavdevice:libavformat:libavcodec:libavutil
> +check_ldflags
> -Wl,-rpath-link=:libswresample:libswscale:libavfilter:libavdevice:libavformat:libavcodec:libavutil
> enabled rpath && add_ldexeflags -Wl,-rpath,$libdir && add_ldsoflags
> -Wl,-rpath,$libdir
> test_ldflags -Wl,-Bsymbolic && append SHFLAGS -Wl,-Bsymbolic
>
> @@ -7898,7 +7893,6 @@ enabled fsync_filter && prepend avfilter_deps
> "avformat"
> enabled mcdeint_filter && prepend avfilter_deps "avcodec"
> enabled movie_filter && prepend avfilter_deps "avformat avcodec"
> enabled pan_filter && prepend avfilter_deps "swresample"
> -enabled pp_filter && prepend avfilter_deps "postproc"
> enabled qrencode_filter && prepend avfilter_deps "swscale"
> enabled qrencodesrc_filter && prepend avfilter_deps "swscale"
> enabled removelogo_filter && prepend avfilter_deps "avformat avcodec
> swscale"
> @@ -7950,9 +7944,6 @@ expand_deps(){
> reorder_by ${1}_deps LIBRARY_LIST # linking order is expected later
> }
>
> -#we have to remove gpl from the deps here as some code assumes all lib
> deps are libs
> -postproc_deps="$(filter_out 'gpl' $postproc_deps)"
> -
> map 'expand_deps $v' $LIBRARY_LIST
>
> if test "$quiet" != "yes"; then
> @@ -8043,7 +8034,6 @@ echo "optimize for size ${small-no}"
> echo "optimizations ${optimizations-no}"
> echo "static ${static-no}"
> echo "shared ${shared-no}"
> -echo "postprocessing support ${postproc-no}"
> echo "network support ${network-no}"
> echo "threading support ${thread_type-no}"
> echo "safe bitstream reader ${safe_bitstream_reader-no}"
> @@ -8420,7 +8410,6 @@ extralibs_avcodec="$avcodec_extralibs"
> extralibs_avformat="$avformat_extralibs"
> extralibs_avdevice="$avdevice_extralibs"
> extralibs_avfilter="$avfilter_extralibs"
> -extralibs_postproc="$postproc_extralibs"
> extralibs_swscale="$swscale_extralibs"
> extralibs_swresample="$swresample_extralibs"
> EOF
> diff --git a/doc/filters.texi b/doc/filters.texi
> index a9ec077ef04..679b71f2906 100644
> --- a/doc/filters.texi
> +++ b/doc/filters.texi
> @@ -19387,181 +19387,6 @@ Set window Y position, relative offset on Y axis.
>
> This filter supports same @ref{commands} as options.
>
> -@section pp
> -
> -Enable the specified chain of postprocessing subfilters using
> libpostproc. This
> -library should be automatically selected with a GPL build
> (@code{--enable-gpl}).
> -Subfilters must be separated by '/' and can be disabled by prepending a
> '-'.
> -Each subfilter and some options have a short and a long name that can be
> used
> -interchangeably, i.e. dr/dering are the same.
> -
> -The filters accept the following options:
> -
> -@table @option
> -@item subfilters
> -Set postprocessing subfilters string.
> -@end table
> -
> -All subfilters share common options to determine their scope:
> -
> -@table @option
> -@item a/autoq
> -Honor the quality commands for this subfilter.
> -
> -@item c/chrom
> -Do chrominance filtering, too (default).
> -
> -@item y/nochrom
> -Do luma filtering only (no chrominance).
> -
> -@item n/noluma
> -Do chrominance filtering only (no luma).
> -@end table
> -
> -These options can be appended after the subfilter name, separated by a
> '|'.
> -
> -Available subfilters are:
> -
> -@table @option
> -@item hb/hdeblock[|difference[|flatness]]
> -Horizontal deblocking filter
> -@table @option
> -@item difference
> -Difference factor where higher values mean more deblocking (default:
> @code{32}).
> -@item flatness
> -Flatness threshold where lower values mean more deblocking (default:
> @code{39}).
> -@end table
> -
> -@item vb/vdeblock[|difference[|flatness]]
> -Vertical deblocking filter
> -@table @option
> -@item difference
> -Difference factor where higher values mean more deblocking (default:
> @code{32}).
> -@item flatness
> -Flatness threshold where lower values mean more deblocking (default:
> @code{39}).
> -@end table
> -
> -@item ha/hadeblock[|difference[|flatness]]
> -Accurate horizontal deblocking filter
> -@table @option
> -@item difference
> -Difference factor where higher values mean more deblocking (default:
> @code{32}).
> -@item flatness
> -Flatness threshold where lower values mean more deblocking (default:
> @code{39}).
> -@end table
> -
> -@item va/vadeblock[|difference[|flatness]]
> -Accurate vertical deblocking filter
> -@table @option
> -@item difference
> -Difference factor where higher values mean more deblocking (default:
> @code{32}).
> -@item flatness
> -Flatness threshold where lower values mean more deblocking (default:
> @code{39}).
> -@end table
> -@end table
> -
> -The horizontal and vertical deblocking filters share the difference and
> -flatness values so you cannot set different horizontal and vertical
> -thresholds.
> -
> -@table @option
> -@item h1/x1hdeblock
> -Experimental horizontal deblocking filter
> -
> -@item v1/x1vdeblock
> -Experimental vertical deblocking filter
> -
> -@item dr/dering
> -Deringing filter
> -
> -@item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise
> reducer
> -@table @option
> -@item threshold1
> -larger -> stronger filtering
> -@item threshold2
> -larger -> stronger filtering
> -@item threshold3
> -larger -> stronger filtering
> -@end table
> -
> -@item al/autolevels[:f/fullyrange], automatic brightness / contrast
> correction
> -@table @option
> -@item f/fullyrange
> -Stretch luma to @code{0-255}.
> -@end table
> -
> -@item lb/linblenddeint
> -Linear blend deinterlacing filter that deinterlaces the given block by
> -filtering all lines with a @code{(1 2 1)} filter.
> -
> -@item li/linipoldeint
> -Linear interpolating deinterlacing filter that deinterlaces the given
> block by
> -linearly interpolating every second line.
> -
> -@item ci/cubicipoldeint
> -Cubic interpolating deinterlacing filter deinterlaces the given block by
> -cubically interpolating every second line.
> -
> -@item md/mediandeint
> -Median deinterlacing filter that deinterlaces the given block by applying
> a
> -median filter to every second line.
> -
> -@item fd/ffmpegdeint
> -FFmpeg deinterlacing filter that deinterlaces the given block by
> filtering every
> -second line with a @code{(-1 4 2 4 -1)} filter.
> -
> -@item l5/lowpass5
> -Vertically applied FIR lowpass deinterlacing filter that deinterlaces the
> given
> -block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
> -
> -@item fq/forceQuant[|quantizer]
> -Overrides the quantizer table from the input with the constant quantizer
> you
> -specify.
> -@table @option
> -@item quantizer
> -Quantizer to use
> -@end table
> -
> -@item de/default
> -Default pp filter combination (@code{hb|a,vb|a,dr|a})
> -
> -@item fa/fast
> -Fast pp filter combination (@code{h1|a,v1|a,dr|a})
> -
> -@item ac
> -High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
> -@end table
> -
> -@subsection Examples
> -
> -@itemize
> -@item
> -Apply horizontal and vertical deblocking, deringing and automatic
> -brightness/contrast:
> -@example
> -pp=hb/vb/dr/al
> -@end example
> -
> -@item
> -Apply default filters without brightness/contrast correction:
> -@example
> -pp=de/-al
> -@end example
> -
> -@item
> -Apply default filters and temporal denoiser:
> -@example
> -pp=default/tmpnoise|1|2|3
> -@end example
> -
> -@item
> -Apply deblocking on luma only, and switch vertical deblocking on or off
> -automatically depending on available CPU time:
> -@example
> -pp=hb|y/vb|a
> -@end example
> -@end itemize
> -
> @section pp7
> Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
> similar to spp = 6 with 7 point DCT, where only the center sample is
> diff --git a/fftools/ffprobe.c b/fftools/ffprobe.c
> index d980d4e64fe..f33531fd84c 100644
> --- a/fftools/ffprobe.c
> +++ b/fftools/ffprobe.c
> @@ -62,8 +62,6 @@
> #include "libswscale/version.h"
> #include "libswresample/swresample.h"
> #include "libswresample/version.h"
> -#include "libpostproc/postprocess.h"
> -#include "libpostproc/version.h"
> #include "libavfilter/version.h"
> #include "textformat/avtextformat.h"
> #include "cmdutils.h"
> @@ -2573,7 +2571,6 @@ static void
> ffprobe_show_library_versions(AVTextFormatContext *tfc)
> SHOW_LIB_VERSION(avfilter, AVFILTER);
> SHOW_LIB_VERSION(swscale, SWSCALE);
> SHOW_LIB_VERSION(swresample, SWRESAMPLE);
> - SHOW_LIB_VERSION(postproc, POSTPROC);
> avtext_print_section_footer(tfc);
> }
>
> diff --git a/fftools/opt_common.c b/fftools/opt_common.c
> index 2ac3fd4fb36..c2f6b9de2a7 100644
> --- a/fftools/opt_common.c
> +++ b/fftools/opt_common.c
> @@ -60,8 +60,6 @@
> #include "libswresample/swresample.h"
> #include "libswresample/version.h"
>
> -#include "libpostproc/postprocess.h"
> -#include "libpostproc/version.h"
>
> enum show_muxdemuxers {
> SHOW_DEFAULT,
> @@ -191,7 +189,6 @@ static void print_all_libs_info(int flags, int level)
> PRINT_LIB_INFO(avfilter, AVFILTER, flags, level);
> PRINT_LIB_INFO(swscale, SWSCALE, flags, level);
> PRINT_LIB_INFO(swresample, SWRESAMPLE, flags, level);
> - PRINT_LIB_INFO(postproc, POSTPROC, flags, level);
> }
>
> static void print_program_info(int flags, int level)
> diff --git a/libavfilter/Makefile b/libavfilter/Makefile
> index 7c0d879ec9c..0effe4127ff 100644
> --- a/libavfilter/Makefile
> +++ b/libavfilter/Makefile
> @@ -431,7 +431,6 @@ OBJS-$(CONFIG_PHOTOSENSITIVITY_FILTER) +=
> vf_photosensitivity.o
> OBJS-$(CONFIG_PIXDESCTEST_FILTER) += vf_pixdesctest.o
> OBJS-$(CONFIG_PIXELIZE_FILTER) += vf_pixelize.o
> OBJS-$(CONFIG_PIXSCOPE_FILTER) += vf_datascope.o
> -OBJS-$(CONFIG_PP_FILTER) += vf_pp.o qp_table.o
> OBJS-$(CONFIG_PP7_FILTER) += vf_pp7.o qp_table.o
> OBJS-$(CONFIG_PREMULTIPLY_FILTER) += vf_premultiply.o
> framesync.o
> OBJS-$(CONFIG_PREWITT_FILTER) += vf_convolution.o
> diff --git a/libavfilter/allfilters.c b/libavfilter/allfilters.c
> index 740d9ab265c..5ea33cdf01b 100644
> --- a/libavfilter/allfilters.c
> +++ b/libavfilter/allfilters.c
> @@ -404,7 +404,6 @@ extern const FFFilter ff_vf_photosensitivity;
> extern const FFFilter ff_vf_pixdesctest;
> extern const FFFilter ff_vf_pixelize;
> extern const FFFilter ff_vf_pixscope;
> -extern const FFFilter ff_vf_pp;
> extern const FFFilter ff_vf_pp7;
> extern const FFFilter ff_vf_premultiply;
> extern const FFFilter ff_vf_prewitt;
> diff --git a/libavfilter/vf_pp.c b/libavfilter/vf_pp.c
> deleted file mode 100644
> index 9e9903eb974..00000000000
> --- a/libavfilter/vf_pp.c
> +++ /dev/null
> @@ -1,191 +0,0 @@
> -/*
> - * Copyright (c) 2002 A'rpi
> - * Copyright (C) 2012 Clément Bœsch
> - *
> - * This file is part of FFmpeg.
> - *
> - * FFmpeg is free software; you can redistribute it and/or modify
> - * it under the terms of the GNU General Public License as published by
> - * the Free Software Foundation; either version 2 of the License, or
> - * (at your option) any later version.
> - *
> - * FFmpeg is distributed in the hope that it will be useful,
> - * but WITHOUT ANY WARRANTY; without even the implied warranty of
> - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> - * GNU General Public License for more details.
> - *
> - * You should have received a copy of the GNU General Public License along
> - * with FFmpeg; if not, write to the Free Software Foundation, Inc.,
> - * 51 Franklin Street
> <https://www.google.com/maps/search/51+Franklin+Street?entry=gmail&source=g>,
> Fifth Floor, Boston, MA 02110-1301 USA.
> - */
> -
> -/**
> - * @file
> - * libpostproc filter, ported from MPlayer.
> - */
> -
> -#include "libavutil/avassert.h"
> -#include "libavutil/mem.h"
> -#include "libavutil/opt.h"
> -
> -#include "filters.h"
> -#include "qp_table.h"
> -#include "video.h"
> -
> -#include "libpostproc/postprocess.h"
> -
> -typedef struct PPFilterContext {
> - const AVClass *class;
> - char *subfilters;
> - int mode_id;
> - pp_mode *modes[PP_QUALITY_MAX + 1];
> - void *pp_ctx;
> -} PPFilterContext;
> -
> -#define OFFSET(x) offsetof(PPFilterContext, x)
> -#define FLAGS AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_VIDEO_PARAM
> -static const AVOption pp_options[] = {
> - { "subfilters", "set postprocess subfilters", OFFSET(subfilters),
> AV_OPT_TYPE_STRING, {.str="de"}, .flags = FLAGS },
> - { NULL }
> -};
> -
> -AVFILTER_DEFINE_CLASS(pp);
> -
> -static av_cold int pp_init(AVFilterContext *ctx)
> -{
> - int i;
> - PPFilterContext *pp = ctx->priv;
> -
> - for (i = 0; i <= PP_QUALITY_MAX; i++) {
> - pp->modes[i] = pp_get_mode_by_name_and_quality(pp->subfilters, i);
> - if (!pp->modes[i])
> - return AVERROR_EXTERNAL;
> - }
> - pp->mode_id = PP_QUALITY_MAX;
> - return 0;
> -}
> -
> -static int pp_process_command(AVFilterContext *ctx, const char *cmd,
> const char *args,
> - char *res, int res_len, int flags)
> -{
> - PPFilterContext *pp = ctx->priv;
> -
> - if (!strcmp(cmd, "quality")) {
> - pp->mode_id = av_clip(strtol(args, NULL, 10), 0, PP_QUALITY_MAX);
> - return 0;
> - }
> - return AVERROR(ENOSYS);
> -}
> -
> -static const enum AVPixelFormat pix_fmts[] = {
> - AV_PIX_FMT_YUV420P, AV_PIX_FMT_YUVJ420P,
> - AV_PIX_FMT_YUV422P, AV_PIX_FMT_YUVJ422P,
> - AV_PIX_FMT_YUV411P,
> - AV_PIX_FMT_GBRP,
> - AV_PIX_FMT_YUV444P, AV_PIX_FMT_YUVJ444P,
> - AV_PIX_FMT_YUV440P, AV_PIX_FMT_YUVJ440P,
> - AV_PIX_FMT_GRAY8,
> - AV_PIX_FMT_NONE
> -};
> -
> -static int pp_config_props(AVFilterLink *inlink)
> -{
> - int flags = PP_CPU_CAPS_AUTO;
> - PPFilterContext *pp = inlink->dst->priv;
> -
> - switch (inlink->format) {
> - case AV_PIX_FMT_GRAY8:
> - case AV_PIX_FMT_YUVJ420P:
> - case AV_PIX_FMT_YUV420P: flags |= PP_FORMAT_420; break;
> - case AV_PIX_FMT_YUVJ422P:
> - case AV_PIX_FMT_YUV422P: flags |= PP_FORMAT_422; break;
> - case AV_PIX_FMT_YUV411P: flags |= PP_FORMAT_411; break;
> - case AV_PIX_FMT_GBRP:
> - case AV_PIX_FMT_YUVJ444P:
> - case AV_PIX_FMT_YUV444P: flags |= PP_FORMAT_444; break;
> - case AV_PIX_FMT_YUVJ440P:
> - case AV_PIX_FMT_YUV440P: flags |= PP_FORMAT_440; break;
> - default: av_assert0(0);
> - }
> -
> - pp->pp_ctx = pp_get_context(inlink->w, inlink->h, flags);
> - if (!pp->pp_ctx)
> - return AVERROR(ENOMEM);
> - return 0;
> -}
> -
> -static int pp_filter_frame(AVFilterLink *inlink, AVFrame *inbuf)
> -{
> - AVFilterContext *ctx = inlink->dst;
> - PPFilterContext *pp = ctx->priv;
> - AVFilterLink *outlink = ctx->outputs[0];
> - const int aligned_w = FFALIGN(outlink->w, 8);
> - const int aligned_h = FFALIGN(outlink->h, 8);
> - AVFrame *outbuf;
> - int qstride = 0;
> - int8_t *qp_table = NULL;
> - int ret;
> -
> - outbuf = ff_get_video_buffer(outlink, aligned_w, aligned_h);
> - if (!outbuf) {
> - av_frame_free(&inbuf);
> - return AVERROR(ENOMEM);
> - }
> - av_frame_copy_props(outbuf, inbuf);
> - outbuf->width = inbuf->width;
> - outbuf->height = inbuf->height;
> -
> - ret = ff_qp_table_extract(inbuf, &qp_table, &qstride, NULL, NULL);
> - if (ret < 0) {
> - av_frame_free(&inbuf);
> - av_frame_free(&outbuf);
> - return ret;
> - }
> -
> - pp_postprocess((const uint8_t **)inbuf->data, inbuf->linesize,
> - outbuf->data, outbuf->linesize,
> - aligned_w, outlink->h,
> - qp_table,
> - qstride,
> - pp->modes[pp->mode_id],
> - pp->pp_ctx,
> - outbuf->pict_type | (qp_table ? PP_PICT_TYPE_QP2 : 0));
> -
> - av_frame_free(&inbuf);
> - av_freep(&qp_table);
> - return ff_filter_frame(outlink, outbuf);
> -}
> -
> -static av_cold void pp_uninit(AVFilterContext *ctx)
> -{
> - int i;
> - PPFilterContext *pp = ctx->priv;
> -
> - for (i = 0; i <= PP_QUALITY_MAX; i++)
> - pp_free_mode(pp->modes[i]);
> - if (pp->pp_ctx)
> - pp_free_context(pp->pp_ctx);
> -}
> -
> -static const AVFilterPad pp_inputs[] = {
> - {
> - .name = "default",
> - .type = AVMEDIA_TYPE_VIDEO,
> - .config_props = pp_config_props,
> - .filter_frame = pp_filter_frame,
> - },
> -};
> -
> -const FFFilter ff_vf_pp = {
> - .p.name = "pp",
> - .p.description = NULL_IF_CONFIG_SMALL("Filter video using
> libpostproc."),
> - .p.priv_class = &pp_class,
> - .p.flags = AVFILTER_FLAG_SUPPORT_TIMELINE_GENERIC,
> - .priv_size = sizeof(PPFilterContext),
> - .init = pp_init,
> - .uninit = pp_uninit,
> - FILTER_INPUTS(pp_inputs),
> - FILTER_OUTPUTS(ff_video_default_filterpad),
> - FILTER_PIXFMTS_ARRAY(pix_fmts),
> - .process_command = pp_process_command,
> -};
> diff --git a/libavutil/avutil.h b/libavutil/avutil.h
> index ee709fbb2ab..c8ae114ab6f 100644
> --- a/libavutil/avutil.h
> +++ b/libavutil/avutil.h
> @@ -41,7 +41,6 @@
> * @li @ref lavd "libavdevice" special devices muxing/demuxing library
> * @li @ref lavu "libavutil" common utility library
> * @li @ref lswr "libswresample" audio resampling, format conversion and
> mixing
> - * @li @ref lpp "libpostproc" post processing library
> * @li @ref libsws "libswscale" color conversion and scaling library
> *
> * @section ffmpeg_versioning Versioning and compatibility
> diff --git a/libpostproc/Makefile b/libpostproc/Makefile
> deleted file mode 100644
> index d78fc0277f4..00000000000
> --- a/libpostproc/Makefile
> +++ /dev/null
> @@ -1,23 +0,0 @@
> -NAME = postproc
> -DESC = FFmpeg postprocessing library
> -FFLIBS = avutil
> -
> -HEADERS = postprocess.h \
> - version.h \
> - version_major.h \
> -
> -OBJS = postprocess.o \
> - version.o \
> -
> -TESTOBJS = tests/test_utils.o \
> -
> -# Windows resource file
> -SHLIBOBJS-$(HAVE_GNU_WINDRES) += postprocres.o
> -
> -TESTPROGS = blocktest \
> - stripetest \
> - temptest \
> -
> -$(SUBDIR)tests/blocktest$(EXESUF): $(SUBDIR)tests/test_utils.o
> -$(SUBDIR)tests/stripetest$(EXESUF): $(SUBDIR)tests/test_utils.o
> -$(SUBDIR)tests/temptest$(EXESUF): $(SUBDIR)tests/test_utils.o
> diff --git a/libpostproc/libpostproc.v b/libpostproc/libpostproc.v
> deleted file mode 100644
> index 27381c6aca3..00000000000
> --- a/libpostproc/libpostproc.v
> +++ /dev/null
> @@ -1,7 +0,0 @@
> -LIBPOSTPROC_MAJOR {
> - global:
> - postproc_*;
> - pp_*;
> - local:
> - *;
> -};
> diff --git a/libpostproc/postprocess.c b/libpostproc/postprocess.c
> deleted file mode 100644
> index eab22289192..00000000000
> --- a/libpostproc/postprocess.c
> +++ /dev/null
> @@ -1,992 +0,0 @@
> -/*
> - * Copyright (C) 2001-2003 Michael Niedermayer (michaelni(a)gmx.at)
> - *
> - * AltiVec optimizations (C) 2004 Romain Dolbeau <romain(a)dolbeau.org>
> - *
> - * This file is part of FFmpeg.
> - *
> - * FFmpeg is free software; you can redistribute it and/or modify
> - * it under the terms of the GNU General Public License as published by
> - * the Free Software Foundation; either version 2 of the License, or
> - * (at your option) any later version.
> - *
> - * FFmpeg is distributed in the hope that it will be useful,
> - * but WITHOUT ANY WARRANTY; without even the implied warranty of
> - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> - * GNU General Public License for more details.
> - *
> - * You should have received a copy of the GNU General Public License
> - * along with FFmpeg; if not, write to the Free Software
> - * Foundation, Inc., 51 Franklin Street
> <https://www.google.com/maps/search/51+Franklin+Street?entry=gmail&source=g>,
> Fifth Floor, Boston, MA 02110-1301 USA
> - */
> -
> -/**
> - * @file
> - * postprocessing.
> - */
> -
> -/*
> - C MMX MMX2 AltiVec
> -isVertDC Ec Ec Ec
> -isVertMinMaxOk Ec Ec Ec
> -doVertLowPass E e Ec
> -doVertDefFilter Ec Ec e Ec
> -isHorizDC Ec Ec Ec
> -isHorizMinMaxOk a E Ec
> -doHorizLowPass E e Ec
> -doHorizDefFilter Ec Ec e Ec
> -do_a_deblock Ec E Ec
> -deRing E e Ecp
> -Vertical RKAlgo1 E a
> -Horizontal RKAlgo1 a
> -Vertical X1# a E
> -Horizontal X1# a E
> -LinIpolDeinterlace e E
> -CubicIpolDeinterlace a e
> -LinBlendDeinterlace e E
> -MedianDeinterlace# E Ec Ec
> -TempDeNoiser# E e Ec
> -
> -# more or less selfinvented filters so the exactness is not too meaningful
> -E = Exact implementation
> -e = almost exact implementation (slightly different rounding,...)
> -a = alternative / approximate impl
> -c = checked against the other implementations (-vo md5)
> -p = partially optimized, still some work to do
> -*/
> -
> -/*
> -TODO:
> -reduce the time wasted on the mem transfer
> -unroll stuff if instructions depend too much on the prior one
> -move YScale thing to the end instead of fixing QP
> -write a faster and higher quality deblocking filter :)
> -make the mainloop more flexible (variable number of blocks at once
> - (the if/else stuff per block is slowing things down)
> -compare the quality & speed of all filters
> -split this huge file
> -optimize c versions
> -try to unroll inner for(x=0 ... loop to avoid these damn if(x ... checks
> -...
> -*/
> -
> -//Changelog: use git log
> -
> -#include <stddef.h>
> -#include <stdlib.h>
> -#include <string.h>
> -
> -#include "config.h"
> -#include "libavutil/common.h"
> -#include "libavutil/cpu.h"
> -#include "libavutil/intreadwrite.h"
> -#include "libavutil/mem.h"
> -//#undef HAVE_MMXEXT_INLINE
> -//#undef HAVE_MMX_INLINE
> -//#undef ARCH_X86
> -//#define DEBUG_BRIGHTNESS
> -#include "postprocess.h"
> -#include "postprocess_internal.h"
> -#include "libavutil/avstring.h"
> -
> -#define GET_MODE_BUFFER_SIZE 500
> -#define OPTIONS_ARRAY_SIZE 10
> -#define BLOCK_SIZE 8
> -#define TEMP_STRIDE 8
> -//#define NUM_BLOCKS_AT_ONCE 16 //not used yet
> -
> -#define DERING_THRESHOLD 20
> -
> -#if ARCH_X86 && HAVE_INLINE_ASM
> -DECLARE_ASM_CONST(8, uint64_t, w05)= 0x0005000500050005LL;
> -DECLARE_ASM_CONST(8, uint64_t, w04)= 0x0004000400040004LL;
> -DECLARE_ASM_CONST(8, uint64_t, w20)= 0x0020002000200020LL;
> -DECLARE_ASM_CONST(8, uint64_t, b00)= 0x0000000000000000LL;
> -DECLARE_ASM_CONST(8, uint64_t, b01)= 0x0101010101010101LL;
> -DECLARE_ASM_CONST(8, uint64_t, b08)= 0x0808080808080808LL;
> -DECLARE_ASM_CONST(8, uint64_t, b80)= 0x8080808080808080LL;
> -#endif
> -
> -static const struct PPFilter filters[]=
> -{
> - {"hb", "hdeblock", 1, 1, 3, H_DEBLOCK},
> - {"vb", "vdeblock", 1, 2, 4, V_DEBLOCK},
> -/* {"hr", "rkhdeblock", 1, 1, 3, H_RK1_FILTER},
> - {"vr", "rkvdeblock", 1, 2, 4, V_RK1_FILTER},*/
> - {"h1", "x1hdeblock", 1, 1, 3, H_X1_FILTER},
> - {"v1", "x1vdeblock", 1, 2, 4, V_X1_FILTER},
> - {"ha", "ahdeblock", 1, 1, 3, H_A_DEBLOCK},
> - {"va", "avdeblock", 1, 2, 4, V_A_DEBLOCK},
> - {"dr", "dering", 1, 5, 6, DERING},
> - {"al", "autolevels", 0, 1, 2, LEVEL_FIX},
> - {"lb", "linblenddeint", 1, 1, 4, LINEAR_BLEND_DEINT_FILTER},
> - {"li", "linipoldeint", 1, 1, 4, LINEAR_IPOL_DEINT_FILTER},
> - {"ci", "cubicipoldeint", 1, 1, 4, CUBIC_IPOL_DEINT_FILTER},
> - {"md", "mediandeint", 1, 1, 4, MEDIAN_DEINT_FILTER},
> - {"fd", "ffmpegdeint", 1, 1, 4, FFMPEG_DEINT_FILTER},
> - {"l5", "lowpass5", 1, 1, 4, LOWPASS5_DEINT_FILTER},
> - {"tn", "tmpnoise", 1, 7, 8, TEMP_NOISE_FILTER},
> - {"fq", "forcequant", 1, 0, 0, FORCE_QUANT},
> - {"be", "bitexact", 1, 0, 0, BITEXACT},
> - {"vi", "visualize", 1, 0, 0, VISUALIZE},
> - {NULL, NULL,0,0,0,0} //End Marker
> -};
> -
> -static const char * const replaceTable[]=
> -{
> - "default", "hb:a,vb:a,dr:a",
> - "de", "hb:a,vb:a,dr:a",
> - "fast", "h1:a,v1:a,dr:a",
> - "fa", "h1:a,v1:a,dr:a",
> - "ac", "ha:a:128:7,va:a,dr:a",
> - NULL //End Marker
> -};
> -
> -/* The horizontal functions exist only in C because the MMX
> - * code is faster with vertical filters and transposing. */
> -
> -/**
> - * Check if the given 8x8 Block is mostly "flat"
> - */
> -static inline int isHorizDC_C(const uint8_t src[], int stride, const
> PPContext *c)
> -{
> - int numEq= 0;
> - int y;
> - const int dcOffset= ((c->nonBQP*c->ppMode.baseDcDiff)>>8) + 1;
> - const int dcThreshold= dcOffset*2 + 1;
> -
> - for(y=0; y<BLOCK_SIZE; y++){
> - numEq += ((unsigned)(src[0] - src[1] + dcOffset)) < dcThreshold;
> - numEq += ((unsigned)(src[1] - src[2] + dcOffset)) < dcThreshold;
> - numEq += ((unsigned)(src[2] - src[3] + dcOffset)) < dcThreshold;
> - numEq += ((unsigned)(src[3] - src[4] + dcOffset)) < dcThreshold;
> - numEq += ((unsigned)(src[4] - src[5] + dcOffset)) < dcThreshold;
> - numEq += ((unsigned)(src[5] - src[6] + dcOffset)) < dcThreshold;
> - numEq += ((unsigned)(src[6] - src[7] + dcOffset)) < dcThreshold;
> - src+= stride;
> - }
> - return numEq > c->ppMode.flatnessThreshold;
> -}
> -
> -/**
> - * Check if the middle 8x8 Block in the given 8x16 block is flat
> - */
> -static inline int isVertDC_C(const uint8_t src[], int stride, const
> PPContext *c)
> -{
> - int numEq= 0;
> - int y;
> - const int dcOffset= ((c->nonBQP*c->ppMode.baseDcDiff)>>8) + 1;
> - const int dcThreshold= dcOffset*2 + 1;
> -
> - src+= stride*4; // src points to begin of the 8x8 Block
> - for(y=0; y<BLOCK_SIZE-1; y++){
> - numEq += ((unsigned)(src[0] - src[0+stride] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[1] - src[1+stride] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[2] - src[2+stride] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[3] - src[3+stride] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[4] - src[4+stride] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[5] - src[5+stride] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[6] - src[6+stride] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[7] - src[7+stride] + dcOffset)) <
> dcThreshold;
> - src+= stride;
> - }
> - return numEq > c->ppMode.flatnessThreshold;
> -}
> -
> -static inline int isHorizMinMaxOk_C(const uint8_t src[], int stride, int
> QP)
> -{
> - int i;
> - for(i=0; i<2; i++){
> - if((unsigned)(src[0] - src[5] + 2*QP) > 4*QP) return 0;
> - src += stride;
> - if((unsigned)(src[2] - src[7] + 2*QP) > 4*QP) return 0;
> - src += stride;
> - if((unsigned)(src[4] - src[1] + 2*QP) > 4*QP) return 0;
> - src += stride;
> - if((unsigned)(src[6] - src[3] + 2*QP) > 4*QP) return 0;
> - src += stride;
> - }
> - return 1;
> -}
> -
> -static inline int isVertMinMaxOk_C(const uint8_t src[], int stride, int
> QP)
> -{
> - int x;
> - src+= stride*4;
> - for(x=0; x<BLOCK_SIZE; x+=4){
> - if((unsigned)(src[ x + 0*stride] - src[ x + 5*stride] + 2*QP) >
> 4*QP) return 0;
> - if((unsigned)(src[1+x + 2*stride] - src[1+x + 7*stride] + 2*QP) >
> 4*QP) return 0;
> - if((unsigned)(src[2+x + 4*stride] - src[2+x + 1*stride] + 2*QP) >
> 4*QP) return 0;
> - if((unsigned)(src[3+x + 6*stride] - src[3+x + 3*stride] + 2*QP) >
> 4*QP) return 0;
> - }
> - return 1;
> -}
> -
> -static inline int horizClassify_C(const uint8_t src[], int stride, const
> PPContext *c)
> -{
> - if( isHorizDC_C(src, stride, c) ){
> - return isHorizMinMaxOk_C(src, stride, c->QP);
> - }else{
> - return 2;
> - }
> -}
> -
> -static inline int vertClassify_C(const uint8_t src[], int stride, const
> PPContext *c)
> -{
> - if( isVertDC_C(src, stride, c) ){
> - return isVertMinMaxOk_C(src, stride, c->QP);
> - }else{
> - return 2;
> - }
> -}
> -
> -static inline void doHorizDefFilter_C(uint8_t dst[], int stride, const
> PPContext *c)
> -{
> - int y;
> - for(y=0; y<BLOCK_SIZE; y++){
> - const int middleEnergy= 5*(dst[4] - dst[3]) + 2*(dst[2] - dst[5]);
> -
> - if(FFABS(middleEnergy) < 8*c->QP){
> - const int q=(dst[3] - dst[4])/2;
> - const int leftEnergy= 5*(dst[2] - dst[1]) + 2*(dst[0] -
> dst[3]);
> - const int rightEnergy= 5*(dst[6] - dst[5]) + 2*(dst[4] -
> dst[7]);
> -
> - int d= FFABS(middleEnergy) - FFMIN( FFABS(leftEnergy),
> FFABS(rightEnergy) );
> - d= FFMAX(d, 0);
> -
> - d= (5*d + 32) >> 6;
> - d*= FFSIGN(-middleEnergy);
> -
> - if(q>0)
> - {
> - d = FFMAX(d, 0);
> - d = FFMIN(d, q);
> - }
> - else
> - {
> - d = FFMIN(d, 0);
> - d = FFMAX(d, q);
> - }
> -
> - dst[3]-= d;
> - dst[4]+= d;
> - }
> - dst+= stride;
> - }
> -}
> -
> -/**
> - * Do a horizontal low pass filter on the 10x8 block (dst points to
> middle 8x8 Block)
> - * using the 9-Tap Filter (1,1,2,2,4,2,2,1,1)/16 (C version)
> - */
> -static inline void doHorizLowPass_C(uint8_t dst[], int stride, const
> PPContext *c)
> -{
> - int y;
> - for(y=0; y<BLOCK_SIZE; y++){
> - const int first= FFABS(dst[-1] - dst[0]) < c->QP ? dst[-1] :
> dst[0];
> - const int last= FFABS(dst[8] - dst[7]) < c->QP ? dst[8] : dst[7];
> -
> - int sums[10];
> - sums[0] = 4*first + dst[0] + dst[1] + dst[2] + 4;
> - sums[1] = sums[0] - first + dst[3];
> - sums[2] = sums[1] - first + dst[4];
> - sums[3] = sums[2] - first + dst[5];
> - sums[4] = sums[3] - first + dst[6];
> - sums[5] = sums[4] - dst[0] + dst[7];
> - sums[6] = sums[5] - dst[1] + last;
> - sums[7] = sums[6] - dst[2] + last;
> - sums[8] = sums[7] - dst[3] + last;
> - sums[9] = sums[8] - dst[4] + last;
> -
> - dst[0]= (sums[0] + sums[2] + 2*dst[0])>>4;
> - dst[1]= (sums[1] + sums[3] + 2*dst[1])>>4;
> - dst[2]= (sums[2] + sums[4] + 2*dst[2])>>4;
> - dst[3]= (sums[3] + sums[5] + 2*dst[3])>>4;
> - dst[4]= (sums[4] + sums[6] + 2*dst[4])>>4;
> - dst[5]= (sums[5] + sums[7] + 2*dst[5])>>4;
> - dst[6]= (sums[6] + sums[8] + 2*dst[6])>>4;
> - dst[7]= (sums[7] + sums[9] + 2*dst[7])>>4;
> -
> - dst+= stride;
> - }
> -}
> -
> -/**
> - * Experimental Filter 1 (Horizontal)
> - * will not damage linear gradients
> - * Flat blocks should look like they were passed through the
> (1,1,2,2,4,2,2,1,1) 9-Tap filter
> - * can only smooth blocks at the expected locations (it cannot smooth
> them if they did move)
> - * MMX2 version does correct clipping C version does not
> - * not identical with the vertical one
> - */
> -static inline void horizX1Filter(uint8_t *src, int stride, int QP)
> -{
> - int y;
> - static uint64_t lut[256];
> - if(!lut[255])
> - {
> - int i;
> - for(i=0; i<256; i++)
> - {
> - int v= i < 128 ? 2*i : 2*(i-256);
> -/*
> -//Simulate 112242211 9-Tap filter
> - uint64_t a= (v/16) & 0xFF;
> - uint64_t b= (v/8) & 0xFF;
> - uint64_t c= (v/4) & 0xFF;
> - uint64_t d= (3*v/8) & 0xFF;
> -*/
> -//Simulate piecewise linear interpolation
> - uint64_t a= (v/16) & 0xFF;
> - uint64_t b= (v*3/16) & 0xFF;
> - uint64_t c= (v*5/16) & 0xFF;
> - uint64_t d= (7*v/16) & 0xFF;
> - uint64_t A= (0x100 - a)&0xFF;
> - uint64_t B= (0x100 - b)&0xFF;
> - uint64_t C= (0x100 - c)&0xFF;
> - uint64_t D= (0x100 - c)&0xFF;
> -
> - lut[i] = (a<<56) | (b<<48) | (c<<40) | (d<<32) |
> - (D<<24) | (C<<16) | (B<<8) | (A);
> - //lut[i] = (v<<32) | (v<<24);
> - }
> - }
> -
> - for(y=0; y<BLOCK_SIZE; y++){
> - int a= src[1] - src[2];
> - int b= src[3] - src[4];
> - int c= src[5] - src[6];
> -
> - int d= FFMAX(FFABS(b) - (FFABS(a) + FFABS(c))/2, 0);
> -
> - if(d < QP){
> - int v = d * FFSIGN(-b);
> -
> - src[1] +=v/8;
> - src[2] +=v/4;
> - src[3] +=3*v/8;
> - src[4] -=3*v/8;
> - src[5] -=v/4;
> - src[6] -=v/8;
> - }
> - src+=stride;
> - }
> -}
> -
> -/**
> - * accurate deblock filter
> - */
> -static av_always_inline void do_a_deblock_C(uint8_t *src, int step,
> - int stride, const PPContext
> *c, int mode)
> -{
> - int y;
> - const int QP= c->QP;
> - const int dcOffset= ((c->nonBQP*c->ppMode.baseDcDiff)>>8) + 1;
> - const int dcThreshold= dcOffset*2 + 1;
> -
> - src+= step*4; // src points to begin of the 8x8 Block
> - for(y=0; y<8; y++){
> - int numEq= 0;
> -
> - numEq += ((unsigned)(src[-1*step] - src[0*step] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[ 0*step] - src[1*step] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[ 1*step] - src[2*step] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[ 2*step] - src[3*step] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[ 3*step] - src[4*step] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[ 4*step] - src[5*step] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[ 5*step] - src[6*step] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[ 6*step] - src[7*step] + dcOffset)) <
> dcThreshold;
> - numEq += ((unsigned)(src[ 7*step] - src[8*step] + dcOffset)) <
> dcThreshold;
> - if(numEq > c->ppMode.flatnessThreshold){
> - int min, max, x;
> -
> - if(src[0] > src[step]){
> - max= src[0];
> - min= src[step];
> - }else{
> - max= src[step];
> - min= src[0];
> - }
> - for(x=2; x<8; x+=2){
> - if(src[x*step] > src[(x+1)*step]){
> - if(src[x *step] > max) max= src[ x *step];
> - if(src[(x+1)*step] < min) min= src[(x+1)*step];
> - }else{
> - if(src[(x+1)*step] > max) max= src[(x+1)*step];
> - if(src[ x *step] < min) min= src[ x *step];
> - }
> - }
> - if(max-min < 2*QP){
> - const int first= FFABS(src[-1*step] - src[0]) < QP ?
> src[-1*step] : src[0];
> - const int last= FFABS(src[8*step] - src[7*step]) < QP ?
> src[8*step] : src[7*step];
> -
> - int sums[10];
> - sums[0] = 4*first + src[0*step] + src[1*step] +
> src[2*step] + 4;
> - sums[1] = sums[0] - first + src[3*step];
> - sums[2] = sums[1] - first + src[4*step];
> - sums[3] = sums[2] - first + src[5*step];
> - sums[4] = sums[3] - first + src[6*step];
> - sums[5] = sums[4] - src[0*step] + src[7*step];
> - sums[6] = sums[5] - src[1*step] + last;
> - sums[7] = sums[6] - src[2*step] + last;
> - sums[8] = sums[7] - src[3*step] + last;
> - sums[9] = sums[8] - src[4*step] + last;
> -
> - if (mode & VISUALIZE) {
> - src[0*step] =
> - src[1*step] =
> - src[2*step] =
> - src[3*step] =
> - src[4*step] =
> - src[5*step] =
> - src[6*step] =
> - src[7*step] = 128;
> - }
> - src[0*step]= (sums[0] + sums[2] + 2*src[0*step])>>4;
> - src[1*step]= (sums[1] + sums[3] + 2*src[1*step])>>4;
> - src[2*step]= (sums[2] + sums[4] + 2*src[2*step])>>4;
> - src[3*step]= (sums[3] + sums[5] + 2*src[3*step])>>4;
> - src[4*step]= (sums[4] + sums[6] + 2*src[4*step])>>4;
> - src[5*step]= (sums[5] + sums[7] + 2*src[5*step])>>4;
> - src[6*step]= (sums[6] + sums[8] + 2*src[6*step])>>4;
> - src[7*step]= (sums[7] + sums[9] + 2*src[7*step])>>4;
> - }
> - }else{
> - const int middleEnergy= 5*(src[4*step] - src[3*step]) +
> 2*(src[2*step] - src[5*step]);
> -
> - if(FFABS(middleEnergy) < 8*QP){
> - const int q=(src[3*step] - src[4*step])/2;
> - const int leftEnergy= 5*(src[2*step] - src[1*step]) +
> 2*(src[0*step] - src[3*step]);
> - const int rightEnergy= 5*(src[6*step] - src[5*step]) +
> 2*(src[4*step] - src[7*step]);
> -
> - int d= FFABS(middleEnergy) - FFMIN( FFABS(leftEnergy),
> FFABS(rightEnergy) );
> - d= FFMAX(d, 0);
> -
> - d= (5*d + 32) >> 6;
> - d*= FFSIGN(-middleEnergy);
> -
> - if(q>0){
> - d = FFMAX(d, 0);
> - d = FFMIN(d, q);
> - }else{
> - d = FFMIN(d, 0);
> - d = FFMAX(d, q);
> - }
> -
> - if ((mode & VISUALIZE) && d) {
> - d= (d < 0) ? 32 : -32;
> - src[3*step]= av_clip_uint8(src[3*step] - d);
> - src[4*step]= av_clip_uint8(src[4*step] + d);
> - d = 0;
> - }
> -
> - src[3*step]-= d;
> - src[4*step]+= d;
> - }
> - }
> -
> - src += stride;
> - }
> -}
> -
> -//Note: we have C and SSE2 version (which uses MMX(EXT) when advantageous)
> -//Plain C versions
> -//we always compile C for testing which needs bitexactness
> -#define TEMPLATE_PP_C 1
> -#include "postprocess_template.c"
> -
> -#if HAVE_ALTIVEC
> -#include "libavutil/ppc/util_altivec.h"
> -
> -# define TEMPLATE_PP_ALTIVEC 1
> -# include "postprocess_altivec_template.c"
> -# include "postprocess_template.c"
> -#endif
> -
> -#if ARCH_X86 && HAVE_INLINE_ASM
> -# if CONFIG_RUNTIME_CPUDETECT
> -# define TEMPLATE_PP_SSE2 1
> -# include "postprocess_template.c"
> -# else
> -# if HAVE_SSE2_INLINE
> -# define TEMPLATE_PP_SSE2 1
> -# include "postprocess_template.c"
> -# endif
> -# endif
> -#endif
> -
> -typedef void (*pp_fn)(const uint8_t src[], int srcStride, uint8_t dst[],
> int dstStride, int width, int height,
> - const int8_t QPs[], int QPStride, int isColor,
> PPContext *c2);
> -
> -static inline void postProcess(const uint8_t src[], int srcStride,
> uint8_t dst[], int dstStride, int width, int height,
> - const int8_t QPs[], int QPStride, int isColor, pp_mode *vm,
> pp_context *vc)
> -{
> - pp_fn pp = postProcess_C;
> - PPContext *c= (PPContext *)vc;
> - PPMode *ppMode= (PPMode *)vm;
> - c->ppMode= *ppMode; //FIXME
> -
> - if (!(ppMode->lumMode & BITEXACT)) {
> -#if CONFIG_RUNTIME_CPUDETECT
> -#if ARCH_X86 && HAVE_INLINE_ASM
> - // ordered per speed fastest first
> - if (c->cpuCaps & AV_CPU_FLAG_SSE2) pp = postProcess_SSE2;
> -#elif HAVE_ALTIVEC
> - if (c->cpuCaps & AV_CPU_FLAG_ALTIVEC) pp =
> postProcess_altivec;
> -#endif
> -#else /* CONFIG_RUNTIME_CPUDETECT */
> -#if HAVE_SSE2_INLINE
> - pp = postProcess_SSE2;
> -#elif HAVE_ALTIVEC
> - pp = postProcess_altivec;
> -#endif
> -#endif /* !CONFIG_RUNTIME_CPUDETECT */
> - }
> -
> - pp(src, srcStride, dst, dstStride, width, height, QPs, QPStride,
> isColor, c);
> -}
> -
> -/* -pp Command line Help
> -*/
> -const char pp_help[] =
> -"Available postprocessing filters:\n"
> -"Filters Options\n"
> -"short long name short long option Description\n"
> -"* * a autoq CPU power dependent
> enabler\n"
> -" c chrom chrominance filtering
> enabled\n"
> -" y nochrom chrominance filtering
> disabled\n"
> -" n noluma luma filtering disabled\n"
> -"hb hdeblock (2 threshold) horizontal deblocking
> filter\n"
> -" 1. difference factor: default=32, higher -> more deblocking\n"
> -" 2. flatness threshold: default=39, lower -> more deblocking\n"
> -" the h & v deblocking filters share these\n"
> -" so you can't set different thresholds for h / v\n"
> -"vb vdeblock (2 threshold) vertical deblocking
> filter\n"
> -"ha hadeblock (2 threshold) horizontal deblocking
> filter\n"
> -"va vadeblock (2 threshold) vertical deblocking
> filter\n"
> -"h1 x1hdeblock experimental h deblock
> filter 1\n"
> -"v1 x1vdeblock experimental v deblock
> filter 1\n"
> -"dr dering deringing filter\n"
> -"al autolevels automatic brightness /
> contrast\n"
> -" f fullyrange stretch luminance to
> (0..255)\n"
> -"lb linblenddeint linear blend
> deinterlacer\n"
> -"li linipoldeint linear interpolating
> deinterlace\n"
> -"ci cubicipoldeint cubic interpolating
> deinterlacer\n"
> -"md mediandeint median deinterlacer\n"
> -"fd ffmpegdeint ffmpeg deinterlacer\n"
> -"l5 lowpass5 FIR lowpass
> deinterlacer\n"
> -"de default hb:a,vb:a,dr:a\n"
> -"fa fast h1:a,v1:a,dr:a\n"
> -"ac ha:a:128:7,va:a,dr:a\n"
> -"tn tmpnoise (3 threshold) temporal noise reducer\n"
> -" 1. <= 2. <= 3. larger -> stronger
> filtering\n"
> -"fq forceQuant <quantizer> force quantizer\n"
> -"Usage:\n"
>
> -"<filterName>[:<option>[:<option>...]][[,|/][-]<filterName>[:<option>...]]...\n"
> -"long form example:\n"
> -"vdeblock:autoq/hdeblock:autoq/linblenddeint default,-vdeblock\n"
> -"short form example:\n"
> -"vb:a/hb:a/lb de,-vb\n"
> -"more examples:\n"
> -"tn:64:128:256\n"
> -"\n"
> -;
> -
> -pp_mode *pp_get_mode_by_name_and_quality(const char *name, int quality)
> -{
> - char temp[GET_MODE_BUFFER_SIZE];
> - char *p= temp;
> - static const char filterDelimiters[] = ",/";
> - static const char optionDelimiters[] = ":|";
> - struct PPMode *ppMode;
> - char *filterToken;
> -
> - if (!name) {
> - av_log(NULL, AV_LOG_ERROR, "pp: Missing argument\n");
> - return NULL;
> - }
> -
> - if (!strcmp(name, "help")) {
> - const char *p;
> - for (p = pp_help; strchr(p, '\n'); p = strchr(p, '\n') + 1) {
> - av_strlcpy(temp, p, FFMIN(sizeof(temp), strchr(p, '\n') - p +
> 2));
> - av_log(NULL, AV_LOG_INFO, "%s", temp);
> - }
> - return NULL;
> - }
> -
> - ppMode= av_malloc(sizeof(PPMode));
> - if (!ppMode)
> - return NULL;
> -
> - ppMode->lumMode= 0;
> - ppMode->chromMode= 0;
> - ppMode->maxTmpNoise[0]= 700;
> - ppMode->maxTmpNoise[1]= 1500;
> - ppMode->maxTmpNoise[2]= 3000;
> - ppMode->maxAllowedY= 234;
> - ppMode->minAllowedY= 16;
> - ppMode->baseDcDiff= 256/8;
> - ppMode->flatnessThreshold= 56-16-1;
> - ppMode->maxClippedThreshold= (AVRational){1,100};
> - ppMode->error=0;
> -
> - memset(temp, 0, GET_MODE_BUFFER_SIZE);
> - av_strlcpy(temp, name, GET_MODE_BUFFER_SIZE - 1);
> -
> - av_log(NULL, AV_LOG_DEBUG, "pp: %s\n", name);
> -
> - for(;;){
> - const char *filterName;
> - int q= 1000000; //PP_QUALITY_MAX;
> - int chrom=-1;
> - int luma=-1;
> - const char *option;
> - const char *options[OPTIONS_ARRAY_SIZE];
> - int i;
> - int filterNameOk=0;
> - int numOfUnknownOptions=0;
> - int enable=1; //does the user want us to enabled or disabled the
> filter
> - char *tokstate;
> -
> - filterToken= av_strtok(p, filterDelimiters, &tokstate);
> - if(!filterToken) break;
> - p+= strlen(filterToken) + 1; // p points to next filterToken
> - filterName= av_strtok(filterToken, optionDelimiters, &tokstate);
> - if (!filterName) {
> - ppMode->error++;
> - break;
> - }
> - av_log(NULL, AV_LOG_DEBUG, "pp: %s::%s\n", filterToken,
> filterName);
> -
> - if(*filterName == '-'){
> - enable=0;
> - filterName++;
> - }
> -
> - for(;;){ //for all options
> - option= av_strtok(NULL, optionDelimiters, &tokstate);
> - if(!option) break;
> -
> - av_log(NULL, AV_LOG_DEBUG, "pp: option: %s\n", option);
> - if(!strcmp("autoq", option) || !strcmp("a", option)) q=
> quality;
> - else if(!strcmp("nochrom", option) || !strcmp("y", option))
> chrom=0;
> - else if(!strcmp("chrom", option) || !strcmp("c", option))
> chrom=1;
> - else if(!strcmp("noluma", option) || !strcmp("n", option))
> luma=0;
> - else{
> - options[numOfUnknownOptions] = option;
> - numOfUnknownOptions++;
> - }
> - if(numOfUnknownOptions >= OPTIONS_ARRAY_SIZE-1) break;
> - }
> - options[numOfUnknownOptions] = NULL;
> -
> - /* replace stuff from the replace Table */
> - for(i=0; replaceTable[2*i]; i++){
> - if(!strcmp(replaceTable[2*i], filterName)){
> - size_t newlen = strlen(replaceTable[2*i + 1]);
> - int plen;
> - int spaceLeft;
> -
> - p--, *p=',';
> -
> - plen= strlen(p);
> - spaceLeft= p - temp + plen;
> - if(spaceLeft + newlen >= GET_MODE_BUFFER_SIZE - 1){
> - ppMode->error++;
> - break;
> - }
> - memmove(p + newlen, p, plen+1);
> - memcpy(p, replaceTable[2*i + 1], newlen);
> - filterNameOk=1;
> - }
> - }
> -
> - for(i=0; filters[i].shortName; i++){
> - if( !strcmp(filters[i].longName, filterName)
> - || !strcmp(filters[i].shortName, filterName)){
> - ppMode->lumMode &= ~filters[i].mask;
> - ppMode->chromMode &= ~filters[i].mask;
> -
> - filterNameOk=1;
> - if(!enable) break; // user wants to disable it
> -
> - if(q >= filters[i].minLumQuality && luma)
> - ppMode->lumMode|= filters[i].mask;
> - if(chrom==1 || (chrom==-1 && filters[i].chromDefault))
> - if(q >= filters[i].minChromQuality)
> - ppMode->chromMode|= filters[i].mask;
> -
> - if(filters[i].mask == LEVEL_FIX){
> - int o;
> - ppMode->minAllowedY= 16;
> - ppMode->maxAllowedY= 234;
> - for(o=0; options[o]; o++){
> - if( !strcmp(options[o],"fullyrange")
> - ||!strcmp(options[o],"f")){
> - ppMode->minAllowedY= 0;
> - ppMode->maxAllowedY= 255;
> - numOfUnknownOptions--;
> - }
> - }
> - }
> - else if(filters[i].mask == TEMP_NOISE_FILTER)
> - {
> - int o;
> - int numOfNoises=0;
> -
> - for(o=0; options[o]; o++){
> - char *tail;
> - ppMode->maxTmpNoise[numOfNoises]=
> - strtol(options[o], &tail, 0);
> - if(tail!=options[o]){
> - numOfNoises++;
> - numOfUnknownOptions--;
> - if(numOfNoises >= 3) break;
> - }
> - }
> - }
> - else if(filters[i].mask == V_DEBLOCK || filters[i].mask
> == H_DEBLOCK
> - || filters[i].mask == V_A_DEBLOCK || filters[i].mask
> == H_A_DEBLOCK){
> - int o;
> -
> - for(o=0; options[o] && o<2; o++){
> - char *tail;
> - int val= strtol(options[o], &tail, 0);
> - if(tail==options[o]) break;
> -
> - numOfUnknownOptions--;
> - if(o==0) ppMode->baseDcDiff= val;
> - else ppMode->flatnessThreshold= val;
> - }
> - }
> - else if(filters[i].mask == FORCE_QUANT){
> - int o;
> - ppMode->forcedQuant= 15;
> -
> - for(o=0; options[o] && o<1; o++){
> - char *tail;
> - int val= strtol(options[o], &tail, 0);
> - if(tail==options[o]) break;
> -
> - numOfUnknownOptions--;
> - ppMode->forcedQuant= val;
> - }
> - }
> - }
> - }
> - if(!filterNameOk) ppMode->error++;
> - ppMode->error += numOfUnknownOptions;
> - }
> -
> - av_log(NULL, AV_LOG_DEBUG, "pp: lumMode=%X, chromMode=%X\n",
> ppMode->lumMode, ppMode->chromMode);
> - if(ppMode->error){
> - av_log(NULL, AV_LOG_ERROR, "%d errors in postprocess string
> \"%s\"\n", ppMode->error, name);
> - av_free(ppMode);
> - return NULL;
> - }
> - return ppMode;
> -}
> -
> -void pp_free_mode(pp_mode *mode){
> - av_free(mode);
> -}
> -
> -static void reallocAlign(void **p, int size){
> - av_free(*p);
> - *p= av_mallocz(size);
> -}
> -
> -static void reallocBuffers(PPContext *c, int width, int height, int
> stride, int qpStride){
> - int mbWidth = (width+15)>>4;
> - int mbHeight= (height+15)>>4;
> - int i;
> -
> - c->stride= stride;
> - c->qpStride= qpStride;
> -
> - reallocAlign((void **)&c->tempDst, stride*24+32);
> - reallocAlign((void **)&c->tempSrc, stride*24);
> - reallocAlign((void **)&c->tempBlocks, 2*16*8);
> - reallocAlign((void **)&c->yHistogram, 256*sizeof(uint64_t));
> - for(i=0; i<256; i++)
> - c->yHistogram[i]= width*height/64*15/256;
> -
> - for(i=0; i<3; i++){
> - //Note: The +17*1024 is just there so I do not have to worry
> about r/w over the end.
> - reallocAlign((void **)&c->tempBlurred[i], stride*mbHeight*16 +
> 17*1024);
> - reallocAlign((void **)&c->tempBlurredPast[i],
> 256*((height+7)&(~7))/2 + 17*1024);//FIXME size
> - }
> -
> - reallocAlign((void **)&c->deintTemp, 2*width+32);
> - reallocAlign((void **)&c->nonBQPTable,
> qpStride*mbHeight*sizeof(int8_t));
> - reallocAlign((void **)&c->stdQPTable,
> qpStride*mbHeight*sizeof(int8_t));
> - reallocAlign((void **)&c->forcedQPTable, mbWidth*sizeof(int8_t));
> -}
> -
> -static const char * context_to_name(void * ptr) {
> - return "postproc";
> -}
> -
> -static const AVClass av_codec_context_class = { "Postproc",
> context_to_name, NULL };
> -
> -av_cold pp_context *pp_get_context(int width, int height, int cpuCaps){
> - PPContext *c= av_mallocz(sizeof(PPContext));
> - int stride= FFALIGN(width, 16); //assumed / will realloc if needed
> - int qpStride= (width+15)/16 + 2; //assumed / will realloc if needed
> -
> - if (!c)
> - return NULL;
> -
> - c->av_class = &av_codec_context_class;
> - if(cpuCaps&PP_FORMAT){
> - c->hChromaSubSample= cpuCaps&0x3;
> - c->vChromaSubSample= (cpuCaps>>4)&0x3;
> - }else{
> - c->hChromaSubSample= 1;
> - c->vChromaSubSample= 1;
> - }
> - if (cpuCaps & PP_CPU_CAPS_AUTO) {
> - c->cpuCaps = av_get_cpu_flags();
> - } else {
> - c->cpuCaps = 0;
> - if (cpuCaps & PP_CPU_CAPS_ALTIVEC) c->cpuCaps |=
> AV_CPU_FLAG_ALTIVEC;
> - }
> -
> - reallocBuffers(c, width, height, stride, qpStride);
> -
> - c->frameNum=-1;
> -
> - return c;
> -}
> -
> -av_cold void pp_free_context(void *vc){
> - PPContext *c = (PPContext*)vc;
> - int i;
> -
> - for(i=0; i<FF_ARRAY_ELEMS(c->tempBlurred); i++)
> - av_free(c->tempBlurred[i]);
> - for(i=0; i<FF_ARRAY_ELEMS(c->tempBlurredPast); i++)
> - av_free(c->tempBlurredPast[i]);
> -
> - av_free(c->tempBlocks);
> - av_free(c->yHistogram);
> - av_free(c->tempDst);
> - av_free(c->tempSrc);
> - av_free(c->deintTemp);
> - av_free(c->stdQPTable);
> - av_free(c->nonBQPTable);
> - av_free(c->forcedQPTable);
> -
> - memset(c, 0, sizeof(PPContext));
> -
> - av_free(c);
> -}
> -
> -void pp_postprocess(const uint8_t * src[3], const int srcStride[3],
> - uint8_t * dst[3], const int dstStride[3],
> - int width, int height,
> - const int8_t *QP_store, int QPStride,
> - pp_mode *vm, void *vc, int pict_type)
> -{
> - int mbWidth = (width+15)>>4;
> - int mbHeight= (height+15)>>4;
> - PPMode *mode = vm;
> - PPContext *c = vc;
> - int minStride= FFMAX(FFABS(srcStride[0]), FFABS(dstStride[0]));
> - int absQPStride = FFABS(QPStride);
> -
> - if (width < 16 || height < 16) {
> - av_log(c, AV_LOG_ERROR, "Postproc is designed to filter 16x16
> macroblock based formats, the minimum size is 1 macroblock\n");
> - return;
> - }
> -
> - // c->stride and c->QPStride are always positive
> - if(c->stride < minStride || c->qpStride < absQPStride)
> - reallocBuffers(c, width, height,
> - FFMAX(minStride, c->stride),
> - FFMAX(c->qpStride, absQPStride));
> -
> - if(!QP_store || (mode->lumMode & FORCE_QUANT)){
> - int i;
> - QP_store= c->forcedQPTable;
> - absQPStride = QPStride = 0;
> - if(mode->lumMode & FORCE_QUANT)
> - for(i=0; i<mbWidth; i++) c->forcedQPTable[i]=
> mode->forcedQuant;
> - else
> - for(i=0; i<mbWidth; i++) c->forcedQPTable[i]= 1;
> - }
> -
> - if(pict_type & PP_PICT_TYPE_QP2){
> - int i;
> - const int count= FFMAX(mbHeight * absQPStride, mbWidth);
> - for(i=0; i<(count>>2); i++){
> - AV_WN32(c->stdQPTable + (i<<2), AV_RN32(QP_store + (i<<2)) >>
> 1 & 0x7F7F7F7F);
> - }
> - for(i<<=2; i<count; i++){
> - c->stdQPTable[i] = QP_store[i]>>1;
> - }
> - QP_store= c->stdQPTable;
> - QPStride= absQPStride;
> - }
> -
> - if(0){
> - int x,y;
> - for(y=0; y<mbHeight; y++){
> - for(x=0; x<mbWidth; x++){
> - av_log(c, AV_LOG_INFO, "%2d ", QP_store[x + y*QPStride]);
> - }
> - av_log(c, AV_LOG_INFO, "\n");
> - }
> - av_log(c, AV_LOG_INFO, "\n");
> - }
> -
> - if((pict_type&7)!=3){
> - if (QPStride >= 0){
> - int i;
> - const int count= FFMAX(mbHeight * QPStride, mbWidth);
> - for(i=0; i<(count>>2); i++){
> - AV_WN32(c->nonBQPTable + (i<<2), AV_RN32(QP_store +
> (i<<2)) & 0x3F3F3F3F);
> - }
> - for(i<<=2; i<count; i++){
> - c->nonBQPTable[i] = QP_store[i] & 0x3F;
> - }
> - } else {
> - int i,j;
> - for(i=0; i<mbHeight; i++) {
> - for(j=0; j<absQPStride; j++) {
> - c->nonBQPTable[i*absQPStride+j] =
> QP_store[i*QPStride+j] & 0x3F;
> - }
> - }
> - }
> - }
> -
> - av_log(c, AV_LOG_DEBUG, "using npp filters 0x%X/0x%X\n",
> - mode->lumMode, mode->chromMode);
> -
> - postProcess(src[0], srcStride[0], dst[0], dstStride[0],
> - width, height, QP_store, QPStride, 0, mode, c);
> -
> - if (!(src[1] && src[2] && dst[1] && dst[2]))
> - return;
> -
> - width = (width )>>c->hChromaSubSample;
> - height = (height)>>c->vChromaSubSample;
> -
> - if(mode->chromMode){
> - postProcess(src[1], srcStride[1], dst[1], dstStride[1],
> - width, height, QP_store, QPStride, 1, mode, c);
> - postProcess(src[2], srcStride[2], dst[2], dstStride[2],
> - width, height, QP_store, QPStride, 2, mode, c);
> - }
> - else if(srcStride[1] == dstStride[1] && srcStride[2] == dstStride[2]){
> - linecpy(dst[1], src[1], height, srcStride[1]);
> - linecpy(dst[2], src[2], height, srcStride[2]);
> - }else{
> - int y;
> - for(y=0; y<height; y++){
> - memcpy(&(dst[1][y*dstStride[1]]), &(src[1][y*srcStride[1]]),
> width);
> - memcpy(&(dst[2][y*dstStride[2]]), &(src[2][y*srcStride[2]]),
> width);
> - }
> - }
> -}
> diff --git a/libpostproc/postprocess.h b/libpostproc/postprocess.h
> deleted file mode 100644
> index d2adb6ccad7..00000000000
> --- a/libpostproc/postprocess.h
> +++ /dev/null
> @@ -1,109 +0,0 @@
> -/*
> - * Copyright (C) 2001-2003 Michael Niedermayer (michaelni(a)gmx.at)
> - *
> - * This file is part of FFmpeg.
> - *
> - * FFmpeg is free software; you can redistribute it and/or modify
> - * it under the terms of the GNU General Public License as published by
> - * the Free Software Foundation; either version 2 of the License, or
> - * (at your option) any later version.
> - *
> - * FFmpeg is distributed in the hope that it will be useful,
> - * but WITHOUT ANY WARRANTY; without even the implied warranty of
> - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> - * GNU General Public License for more details.
> - *
> - * You should have received a copy of the GNU General Public License
> - * along with FFmpeg; if not, write to the Free Software
> - * Foundation, Inc., 51 Franklin Street
> <https://www.google.com/maps/search/51+Franklin+Street?entry=gmail&source=g>,
> Fifth Floor, Boston, MA 02110-1301 USA
> - */
> -
> -#ifndef POSTPROC_POSTPROCESS_H
> -#define POSTPROC_POSTPROCESS_H
> -
> -/**
> - * @file
> - * @ingroup lpp
> - * external API header
> - */
> -
> -/**
> - * @defgroup lpp libpostproc
> - * Video postprocessing library.
> - *
> - * @{
> - */
> -
> -#include "libpostproc/version_major.h"
> -#ifndef HAVE_AV_CONFIG_H
> -/* When included as part of the ffmpeg build, only include the major
> version
> - * to avoid unnecessary rebuilds. When included externally, keep including
> - * the full version information. */
> -#include "libpostproc/version.h"
> -#endif
> -
> -/**
> - * Return the LIBPOSTPROC_VERSION_INT constant.
> - */
> -unsigned postproc_version(void);
> -
> -/**
> - * Return the libpostproc build-time configuration.
> - */
> -const char *postproc_configuration(void);
> -
> -/**
> - * Return the libpostproc license.
> - */
> -const char *postproc_license(void);
> -
> -#define PP_QUALITY_MAX 6
> -
> -#include <inttypes.h>
> -
> -typedef void pp_context;
> -typedef void pp_mode;
> -
> -extern const char pp_help[]; ///< a simple help text
> -
> -void pp_postprocess(const uint8_t * src[3], const int srcStride[3],
> - uint8_t * dst[3], const int dstStride[3],
> - int horizontalSize, int verticalSize,
> - const int8_t *QP_store, int QP_stride,
> - pp_mode *mode, pp_context *ppContext, int pict_type);
> -
> -
> -/**
> - * Return a pp_mode or NULL if an error occurred.
> - *
> - * @param name the string after "-pp" on the command line
> - * @param quality a number from 0 to PP_QUALITY_MAX
> - */
> -pp_mode *pp_get_mode_by_name_and_quality(const char *name, int quality);
> -void pp_free_mode(pp_mode *mode);
> -
> -pp_context *pp_get_context(int width, int height, int flags);
> -void pp_free_context(pp_context *ppContext);
> -
> -#define PP_CPU_CAPS_MMX 0x80000000
> -#define PP_CPU_CAPS_MMX2 0x20000000
> -#if FF_API_PP_AMD_3DNOW
> -#define PP_CPU_CAPS_3DNOW 0x40000000
> -#endif
> -#define PP_CPU_CAPS_ALTIVEC 0x10000000
> -#define PP_CPU_CAPS_AUTO 0x00080000
> -
> -#define PP_FORMAT 0x00000008
> -#define PP_FORMAT_420 (0x00000011|PP_FORMAT)
> -#define PP_FORMAT_422 (0x00000001|PP_FORMAT)
> -#define PP_FORMAT_411 (0x00000002|PP_FORMAT)
> -#define PP_FORMAT_444 (0x00000000|PP_FORMAT)
> -#define PP_FORMAT_440 (0x00000010|PP_FORMAT)
> -
> -#define PP_PICT_TYPE_QP2 0x00000010 ///< MPEG2 style QScale
> -
> -/**
> - * @}
> - */
> -
> -#endif /* POSTPROC_POSTPROCESS_H */
> diff --git a/libpostproc/postprocess_altivec_template.c
> b/libpostproc/postprocess_altivec_template.c
> deleted file mode 100644
> index feddab50356..00000000000
> --- a/libpostproc/postprocess_altivec_template.c
> +++ /dev/null
> @@ -1,1214 +0,0 @@
> -/*
> - * AltiVec optimizations (C) 2004 Romain Dolbeau <romain(a)dolbeau.org>
> - *
> - * based on code by Copyright (C) 2001-2003 Michael Niedermayer (
> michaelni(a)gmx.at)
> - *
> - * This file is part of FFmpeg.
> - *
> - * FFmpeg is free software; you can redistribute it and/or modify
> - * it under the terms of the GNU General Public License as published by
> - * the Free Software Foundation; either version 2 of the License, or
> - * (at your option) any later version.
> - *
> - * FFmpeg is distributed in the hope that it will be useful,
> - * but WITHOUT ANY WARRANTY; without even the implied warranty of
> - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> - * GNU General Public License for more details.
> - *
> - * You should have received a copy of the GNU General Public License
> - * along with FFmpeg; if not, write to the Free Software
> - * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
> 02110-1301 USA
> - */
> -
> -#include "libavutil/avutil.h"
> -#include "libavutil/mem_internal.h"
> -
> -#define
> ALTIVEC_TRANSPOSE_8x8_SHORT(src_a,src_b,src_c,src_d,src_e,src_f,src_g,src_h)
> \
> - do { \
> - __typeof__(src_a) tempA1, tempB1, tempC1, tempD1; \
> - __typeof__(src_a) tempE1, tempF1, tempG1, tempH1; \
> - __typeof__(src_a) tempA2, tempB2, tempC2, tempD2; \
> - __typeof__(src_a) tempE2, tempF2, tempG2, tempH2; \
> - tempA1 = vec_mergeh (src_a, src_e); \
> - tempB1 = vec_mergel (src_a, src_e); \
> - tempC1 = vec_mergeh (src_b, src_f); \
> - tempD1 = vec_mergel (src_b, src_f); \
> - tempE1 = vec_mergeh (src_c, src_g); \
> - tempF1 = vec_mergel (src_c, src_g); \
> - tempG1 = vec_mergeh (src_d, src_h); \
> - tempH1 = vec_mergel (src_d, src_h); \
> - tempA2 = vec_mergeh (tempA1, tempE1); \
> - tempB2 = vec_mergel (tempA1, tempE1); \
> - tempC2 = vec_mergeh (tempB1, tempF1); \
> - tempD2 = vec_mergel (tempB1, tempF1); \
> - tempE2 = vec_mergeh (tempC1, tempG1); \
> - tempF2 = vec_mergel (tempC1, tempG1); \
> - tempG2 = vec_mergeh (tempD1, tempH1); \
> - tempH2 = vec_mergel (tempD1, tempH1); \
> - src_a = vec_mergeh (tempA2, tempE2); \
> - src_b = vec_mergel (tempA2, tempE2); \
> - src_c = vec_mergeh (tempB2, tempF2); \
> - src_d = vec_mergel (tempB2, tempF2); \
> - src_e = vec_mergeh (tempC2, tempG2); \
> - src_f = vec_mergel (tempC2, tempG2); \
> - src_g = vec_mergeh (tempD2, tempH2); \
> - src_h = vec_mergel (tempD2, tempH2); \
> - } while (0)
> -
> -
> -static inline int vertClassify_altivec(uint8_t src[], int stride,
> PPContext *c) {
> - /*
> - this code makes no assumption on src or stride.
> - One could remove the recomputation of the perm
> - vector by assuming (stride % 16) == 0, unfortunately
> - this is not always true.
> - */
> - short data_0 = ((c->nonBQP*c->ppMode.baseDcDiff)>>8) + 1;
> - DECLARE_ALIGNED(16, short, data)[8] =
> - {
> - data_0,
> - data_0 * 2 + 1,
> - c->QP * 2,
> - c->QP * 4
> - };
> - int numEq;
> - uint8_t *src2 = src;
> - vector signed short v_dcOffset;
> - vector signed short v2QP;
> - vector unsigned short v4QP;
> - vector unsigned short v_dcThreshold;
> - const int properStride = (stride % 16);
> - const int srcAlign = ((unsigned long)src2 % 16);
> - const int two_vectors = ((srcAlign > 8) || properStride) ? 1 : 0;
> - const vector signed int zero = vec_splat_s32(0);
> - const vector signed short mask = vec_splat_s16(1);
> - vector signed int v_numEq = vec_splat_s32(0);
> - vector signed short v_data = vec_ld(0, data);
> - vector signed short v_srcAss0, v_srcAss1, v_srcAss2, v_srcAss3,
> - v_srcAss4, v_srcAss5, v_srcAss6, v_srcAss7;
> -//FIXME avoid this mess if possible
> - register int j0 = 0,
> - j1 = stride,
> - j2 = 2 * stride,
> - j3 = 3 * stride,
> - j4 = 4 * stride,
> - j5 = 5 * stride,
> - j6 = 6 * stride,
> - j7 = 7 * stride;
> - vector unsigned char v_srcA0, v_srcA1, v_srcA2, v_srcA3,
> - v_srcA4, v_srcA5, v_srcA6, v_srcA7;
> -
> - v_dcOffset = vec_splat(v_data, 0);
> - v_dcThreshold = (vector unsigned short)vec_splat(v_data, 1);
> - v2QP = vec_splat(v_data, 2);
> - v4QP = (vector unsigned short)vec_splat(v_data, 3);
> -
> - src2 += stride * 4;
> -
> -#define LOAD_LINE(i) \
> - { \
> - vector unsigned char perm##i = vec_lvsl(j##i, src2); \
> - vector unsigned char v_srcA2##i; \
> - vector unsigned char v_srcA1##i = vec_ld(j##i, src2); \
> - if (two_vectors) \
> - v_srcA2##i = vec_ld(j##i + 16, src2); \
> - v_srcA##i = \
> - vec_perm(v_srcA1##i, v_srcA2##i, perm##i); \
> - v_srcAss##i = \
> - (vector signed short)vec_mergeh((vector signed char)zero, \
> - (vector signed char)v_srcA##i); }
> -
> -#define LOAD_LINE_ALIGNED(i) \
> - v_srcA##i = vec_ld(j##i, src2); \
> - v_srcAss##i = \
> - (vector signed short)vec_mergeh((vector signed char)zero, \
> - (vector signed char)v_srcA##i)
> -
> - /* Special-casing the aligned case is worthwhile, as all calls from
> - * the (transposed) horizontable deblocks will be aligned, in addition
> - * to the naturally aligned vertical deblocks. */
> - if (properStride && srcAlign) {
> - LOAD_LINE_ALIGNED(0);
> - LOAD_LINE_ALIGNED(1);
> - LOAD_LINE_ALIGNED(2);
> - LOAD_LINE_ALIGNED(3);
> - LOAD_LINE_ALIGNED(4);
> - LOAD_LINE_ALIGNED(5);
> - LOAD_LINE_ALIGNED(6);
> - LOAD_LINE_ALIGNED(7);
> - } else {
> - LOAD_LINE(0);
> - LOAD_LINE(1);
> - LOAD_LINE(2);
> - LOAD_LINE(3);
> - LOAD_LINE(4);
> - LOAD_LINE(5);
> - LOAD_LINE(6);
> - LOAD_LINE(7);
> - }
> -#undef LOAD_LINE
> -#undef LOAD_LINE_ALIGNED
> -
> -#define ITER(i, j) \
> - const vector signed short v_diff##i = \
> - vec_sub(v_srcAss##i, v_srcAss##j); \
> - const vector signed short v_sum##i = \
> - vec_add(v_diff##i, v_dcOffset); \
> - const vector signed short v_comp##i = \
> - (vector signed short)vec_cmplt((vector unsigned short)v_sum##i, \
> - v_dcThreshold); \
> - const vector signed short v_part##i = vec_and(mask, v_comp##i);
> -
> - {
> - ITER(0, 1)
> - ITER(1, 2)
> - ITER(2, 3)
> - ITER(3, 4)
> - ITER(4, 5)
> - ITER(5, 6)
> - ITER(6, 7)
> -
> - v_numEq = vec_sum4s(v_part0, v_numEq);
> - v_numEq = vec_sum4s(v_part1, v_numEq);
> - v_numEq = vec_sum4s(v_part2, v_numEq);
> - v_numEq = vec_sum4s(v_part3, v_numEq);
> - v_numEq = vec_sum4s(v_part4, v_numEq);
> - v_numEq = vec_sum4s(v_part5, v_numEq);
> - v_numEq = vec_sum4s(v_part6, v_numEq);
> - }
> -
> -#undef ITER
> -
> - v_numEq = vec_sums(v_numEq, zero);
> -
> - v_numEq = vec_splat(v_numEq, 3);
> - vec_ste(v_numEq, 0, &numEq);
> -
> - if (numEq > c->ppMode.flatnessThreshold){
> - const vector unsigned char mmoP1 = (const vector unsigned char)
> - {0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f,
> - 0x00, 0x01, 0x12, 0x13, 0x08, 0x09, 0x1A, 0x1B};
> - const vector unsigned char mmoP2 = (const vector unsigned char)
> - {0x04, 0x05, 0x16, 0x17, 0x0C, 0x0D, 0x1E, 0x1F,
> - 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f, 0x1f};
> - const vector unsigned char mmoP = (const vector unsigned char)
> - vec_lvsl(8, (unsigned char*)0);
> -
> - vector signed short mmoL1 = vec_perm(v_srcAss0, v_srcAss2, mmoP1);
> - vector signed short mmoL2 = vec_perm(v_srcAss4, v_srcAss6, mmoP2);
> - vector signed short mmoL = vec_perm(mmoL1, mmoL2, mmoP);
> - vector signed short mmoR1 = vec_perm(v_srcAss5, v_srcAss7, mmoP1);
> - vector signed short mmoR2 = vec_perm(v_srcAss1, v_srcAss3, mmoP2);
> - vector signed short mmoR = vec_perm(mmoR1, mmoR2, mmoP);
> - vector signed short mmoDiff = vec_sub(mmoL, mmoR);
> - vector unsigned short mmoSum = (vector unsigned
> short)vec_add(mmoDiff, v2QP);
> -
> - if (vec_any_gt(mmoSum, v4QP))
> - return 0;
> - else
> - return 1;
> - }
> - else return 2;
> -}
> -
> -static inline void doVertLowPass_altivec(uint8_t *src, int stride,
> PPContext *c) {
> - /*
> - this code makes no assumption on src or stride.
> - One could remove the recomputation of the perm
> - vector by assuming (stride % 16) == 0, unfortunately
> - this is not always true. Quite a lot of load/stores
> - can be removed by assuming proper alignment of
> - src & stride :-(
> - */
> - uint8_t *src2 = src;
> - const vector signed int zero = vec_splat_s32(0);
> - const int properStride = (stride % 16);
> - const int srcAlign = ((unsigned long)src2 % 16);
> - DECLARE_ALIGNED(16, short, qp)[8] = {c->QP};
> - vector signed short vqp = vec_ld(0, qp);
> - vector signed short vb0, vb1, vb2, vb3, vb4, vb5, vb6, vb7, vb8, vb9;
> - vector unsigned char vbA0, av_uninit(vbA1), av_uninit(vbA2),
> av_uninit(vbA3), av_uninit(vbA4), av_uninit(vbA5), av_uninit(vbA6),
> av_uninit(vbA7), av_uninit(vbA8), vbA9;
> - vector unsigned char vbB0, av_uninit(vbB1), av_uninit(vbB2),
> av_uninit(vbB3), av_uninit(vbB4), av_uninit(vbB5), av_uninit(vbB6),
> av_uninit(vbB7), av_uninit(vbB8), vbB9;
> - vector unsigned char vbT0, vbT1, vbT2, vbT3, vbT4, vbT5, vbT6, vbT7,
> vbT8, vbT9;
> - vector unsigned char perml0, perml1, perml2, perml3, perml4,
> - perml5, perml6, perml7, perml8, perml9;
> - register int j0 = 0,
> - j1 = stride,
> - j2 = 2 * stride,
> - j3 = 3 * stride,
> - j4 = 4 * stride,
> - j5 = 5 * stride,
> - j6 = 6 * stride,
> - j7 = 7 * stride,
> - j8 = 8 * stride,
> - j9 = 9 * stride;
> -
> - vqp = vec_splat(vqp, 0);
> -
> - src2 += stride*3;
> -
> -#define LOAD_LINE(i) \
> - perml##i = vec_lvsl(i * stride, src2); \
> - vbA##i = vec_ld(i * stride, src2); \
> - vbB##i = vec_ld(i * stride + 16, src2); \
> - vbT##i = vec_perm(vbA##i, vbB##i, perml##i); \
> - vb##i = \
> - (vector signed short)vec_mergeh((vector unsigned char)zero, \
> - (vector unsigned char)vbT##i)
> -
> -#define LOAD_LINE_ALIGNED(i) \
> - vbT##i = vec_ld(j##i, src2); \
> - vb##i = \
> - (vector signed short)vec_mergeh((vector signed char)zero, \
> - (vector signed char)vbT##i)
> -
> - /* Special-casing the aligned case is worthwhile, as all calls from
> - * the (transposed) horizontable deblocks will be aligned, in
> addition
> - * to the naturally aligned vertical deblocks. */
> - if (properStride && srcAlign) {
> - LOAD_LINE_ALIGNED(0);
> - LOAD_LINE_ALIGNED(1);
> - LOAD_LINE_ALIGNED(2);
> - LOAD_LINE_ALIGNED(3);
> - LOAD_LINE_ALIGNED(4);
> - LOAD_LINE_ALIGNED(5);
> - LOAD_LINE_ALIGNED(6);
> - LOAD_LINE_ALIGNED(7);
> - LOAD_LINE_ALIGNED(8);
> - LOAD_LINE_ALIGNED(9);
> - } else {
> - LOAD_LINE(0);
> - LOAD_LINE(1);
> - LOAD_LINE(2);
> - LOAD_LINE(3);
> - LOAD_LINE(4);
> - LOAD_LINE(5);
> - LOAD_LINE(6);
> - LOAD_LINE(7);
> - LOAD_LINE(8);
> - LOAD_LINE(9);
> - }
> -#undef LOAD_LINE
> -#undef LOAD_LINE_ALIGNED
> - {
> - const vector unsigned short v_2 = vec_splat_u16(2);
> - const vector unsigned short v_4 = vec_splat_u16(4);
> -
> - const vector signed short v_diff01 = vec_sub(vb0, vb1);
> - const vector unsigned short v_cmp01 =
> - (const vector unsigned short) vec_cmplt(vec_abs(v_diff01),
> vqp);
> - const vector signed short v_first = vec_sel(vb1, vb0, v_cmp01);
> - const vector signed short v_diff89 = vec_sub(vb8, vb9);
> - const vector unsigned short v_cmp89 =
> - (const vector unsigned short) vec_cmplt(vec_abs(v_diff89),
> vqp);
> - const vector signed short v_last = vec_sel(vb8, vb9, v_cmp89);
> -
> - const vector signed short temp01 = vec_mladd(v_first, (vector
> signed short)v_4, vb1);
> - const vector signed short temp02 = vec_add(vb2, vb3);
> - const vector signed short temp03 = vec_add(temp01, (vector signed
> short)v_4);
> - const vector signed short v_sumsB0 = vec_add(temp02, temp03);
> -
> - const vector signed short temp11 = vec_sub(v_sumsB0, v_first);
> - const vector signed short v_sumsB1 = vec_add(temp11, vb4);
> -
> - const vector signed short temp21 = vec_sub(v_sumsB1, v_first);
> - const vector signed short v_sumsB2 = vec_add(temp21, vb5);
> -
> - const vector signed short temp31 = vec_sub(v_sumsB2, v_first);
> - const vector signed short v_sumsB3 = vec_add(temp31, vb6);
> -
> - const vector signed short temp41 = vec_sub(v_sumsB3, v_first);
> - const vector signed short v_sumsB4 = vec_add(temp41, vb7);
> -
> - const vector signed short temp51 = vec_sub(v_sumsB4, vb1);
> - const vector signed short v_sumsB5 = vec_add(temp51, vb8);
> -
> - const vector signed short temp61 = vec_sub(v_sumsB5, vb2);
> - const vector signed short v_sumsB6 = vec_add(temp61, v_last);
> -
> - const vector signed short temp71 = vec_sub(v_sumsB6, vb3);
> - const vector signed short v_sumsB7 = vec_add(temp71, v_last);
> -
> - const vector signed short temp81 = vec_sub(v_sumsB7, vb4);
> - const vector signed short v_sumsB8 = vec_add(temp81, v_last);
> -
> - const vector signed short temp91 = vec_sub(v_sumsB8, vb5);
> - const vector signed short v_sumsB9 = vec_add(temp91, v_last);
> -
> - #define COMPUTE_VR(i, j, k)
> \
> - const vector signed short temps1##i =
> \
> - vec_add(v_sumsB##i, v_sumsB##k);
> \
> - const vector signed short temps2##i =
> \
> - vec_mladd(vb##j, (vector signed short)v_2, temps1##i);
> \
> - const vector signed short vr##j = vec_sra(temps2##i, v_4)
> -
> - COMPUTE_VR(0, 1, 2);
> - COMPUTE_VR(1, 2, 3);
> - COMPUTE_VR(2, 3, 4);
> - COMPUTE_VR(3, 4, 5);
> - COMPUTE_VR(4, 5, 6);
> - COMPUTE_VR(5, 6, 7);
> - COMPUTE_VR(6, 7, 8);
> - COMPUTE_VR(7, 8, 9);
> -
> - const vector signed char neg1 = vec_splat_s8(-1);
> - const vector unsigned char permHH = (const vector unsigned
> char){0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
> -
> 0x18, 0x19, 0x1A, 0x1B, 0x1C, 0x1D, 0x1E, 0x1F};
> -
> -#define PACK_AND_STORE(i) \
> -{ const vector unsigned char perms##i = \
> - vec_lvsr(i * stride, src2); \
> - const vector unsigned char vf##i = \
> - vec_packsu(vr##i, (vector signed short)zero); \
> - const vector unsigned char vg##i = \
> - vec_perm(vf##i, vbT##i, permHH); \
> - const vector unsigned char mask##i = \
> - vec_perm((vector unsigned char)zero, (vector unsigned char)neg1,
> perms##i); \
> - const vector unsigned char vg2##i = \
> - vec_perm(vg##i, vg##i, perms##i); \
> - const vector unsigned char svA##i = \
> - vec_sel(vbA##i, vg2##i, mask##i); \
> - const vector unsigned char svB##i = \
> - vec_sel(vg2##i, vbB##i, mask##i); \
> - vec_st(svA##i, i * stride, src2); \
> - vec_st(svB##i, i * stride + 16, src2);}
> -
> -#define PACK_AND_STORE_ALIGNED(i) \
> -{ const vector unsigned char vf##i = \
> - vec_packsu(vr##i, (vector signed short)zero); \
> - const vector unsigned char vg##i = \
> - vec_perm(vf##i, vbT##i, permHH); \
> - vec_st(vg##i, i * stride, src2);}
> -
> - /* Special-casing the aligned case is worthwhile, as all calls
> from
> - * the (transposed) horizontable deblocks will be aligned, in
> addition
> - * to the naturally aligned vertical deblocks. */
> - if (properStride && srcAlign) {
> - PACK_AND_STORE_ALIGNED(1)
> - PACK_AND_STORE_ALIGNED(2)
> - PACK_AND_STORE_ALIGNED(3)
> - PACK_AND_STORE_ALIGNED(4)
> - PACK_AND_STORE_ALIGNED(5)
> - PACK_AND_STORE_ALIGNED(6)
> - PACK_AND_STORE_ALIGNED(7)
> - PACK_AND_STORE_ALIGNED(8)
> - } else {
> - PACK_AND_STORE(1)
> - PACK_AND_STORE(2)
> - PACK_AND_STORE(3)
> - PACK_AND_STORE(4)
> - PACK_AND_STORE(5)
> - PACK_AND_STORE(6)
> - PACK_AND_STORE(7)
> - PACK_AND_STORE(8)
> - }
> - #undef PACK_AND_STORE
> - #undef PACK_AND_STORE_ALIGNED
> - }
> -}
> -
> -
> -
> -static inline void doVertDefFilter_altivec(uint8_t src[], int stride,
> PPContext *c) {
> - /*
> - this code makes no assumption on src or stride.
> - One could remove the recomputation of the perm
> - vector by assuming (stride % 16) == 0, unfortunately
> - this is not always true. Quite a lot of load/stores
> - can be removed by assuming proper alignment of
> - src & stride :-(
> - */
> - uint8_t *src2 = src + stride*3;
> - const vector signed int zero = vec_splat_s32(0);
> - DECLARE_ALIGNED(16, short, qp)[8] = {8*c->QP};
> - vector signed short vqp = vec_splat(
> - (vector signed short)vec_ld(0, qp), 0);
> -
> -#define LOAD_LINE(i) \
> - const vector unsigned char perm##i = \
> - vec_lvsl(i * stride, src2); \
> - const vector unsigned char vbA##i = \
> - vec_ld(i * stride, src2); \
> - const vector unsigned char vbB##i = \
> - vec_ld(i * stride + 16, src2); \
> - const vector unsigned char vbT##i = \
> - vec_perm(vbA##i, vbB##i, perm##i); \
> - const vector signed short vb##i = \
> - (vector signed short)vec_mergeh((vector unsigned char)zero, \
> - (vector unsigned char)vbT##i)
> -
> - LOAD_LINE(1);
> - LOAD_LINE(2);
> - LOAD_LINE(3);
> - LOAD_LINE(4);
> - LOAD_LINE(5);
> - LOAD_LINE(6);
> - LOAD_LINE(7);
> - LOAD_LINE(8);
> -#undef LOAD_LINE
> -
> - const vector signed short v_1 = vec_splat_s16(1);
> - const vector signed short v_2 = vec_splat_s16(2);
> - const vector signed short v_5 = vec_splat_s16(5);
> - const vector signed short v_32 = vec_sl(v_1,
> - (vector unsigned short)v_5);
> - /* middle energy */
> - const vector signed short l3minusl6 = vec_sub(vb3, vb6);
> - const vector signed short l5minusl4 = vec_sub(vb5, vb4);
> - const vector signed short twotimes_l3minusl6 = vec_mladd(v_2,
> l3minusl6, (vector signed short)zero);
> - const vector signed short mE = vec_mladd(v_5, l5minusl4,
> twotimes_l3minusl6);
> - const vector signed short absmE = vec_abs(mE);
> - /* left & right energy */
> - const vector signed short l1minusl4 = vec_sub(vb1, vb4);
> - const vector signed short l3minusl2 = vec_sub(vb3, vb2);
> - const vector signed short l5minusl8 = vec_sub(vb5, vb8);
> - const vector signed short l7minusl6 = vec_sub(vb7, vb6);
> - const vector signed short twotimes_l1minusl4 = vec_mladd(v_2,
> l1minusl4, (vector signed short)zero);
> - const vector signed short twotimes_l5minusl8 = vec_mladd(v_2,
> l5minusl8, (vector signed short)zero);
> - const vector signed short lE = vec_mladd(v_5, l3minusl2,
> twotimes_l1minusl4);
> - const vector signed short rE = vec_mladd(v_5, l7minusl6,
> twotimes_l5minusl8);
> - /* d */
> - const vector signed short ddiff = vec_sub(absmE,
> - vec_min(vec_abs(lE),
> - vec_abs(rE)));
> - const vector signed short ddiffclamp = vec_max(ddiff, (vector signed
> short)zero);
> - const vector signed short dtimes64 = vec_mladd(v_5, ddiffclamp,
> v_32);
> - const vector signed short d = vec_sra(dtimes64, vec_splat_u16(6));
> - const vector signed short minusd = vec_sub((vector signed
> short)zero, d);
> - const vector signed short finald = vec_sel(minusd,
> - d,
> - vec_cmpgt(vec_sub((vector
> signed short)zero, mE),
> -
1
1
вт, 13 мая 2025 г., 12:11 Pekka Paalanen <pekka.paalanen(a)haloniitty.fi>:
> On Mon, 12 May 2025 15:52:04 +0300
> Andrew Randrianasulu <randrianasulu(a)gmail.com> wrote:
>
> > And modern reincarnation of this idea, using OBS's plugins!
> >
> > https://m.youtube.com/watch?v=qXrn4MqY1Wo
> > "DIY "infinity contrast" TV - with 100% recycled parts" from two months
> > ago, initially found by my friend Vladimir after I showed him all this.
> >
> > added two more recipients in cc.
> >
> > Pekka, you might like this too, as less secretive description of how HDR
> > part of HDR display system works.
>
> Hi Andrew,
>
> that's a neat project for sure.
>
> > >> PS: I played around with Colorspace property on my hdmi connector on
> > >> RX550 via xrandr:
> > >>
> > >> guest@slax:~/botva/src/src$ DISPLAY=":0" xrandr --output 0x54 --set
> > >> Colorspace Default
> > >> guest@slax:~/botva/src/src$ DISPLAY=":0" xrandr --output 0x54 --set
> > >> Colorspace BT2020_YCC
> > >> guest@slax:~/botva/src/src$ DISPLAY=":0" xrandr --output 0x54 --set
> > >> Colorspace Default guest@slax:~/botva/src/src$
> DISPLAY=":0"
> > >> xrandr --output 0x54 --set Colorspace opRGB guest@slax
> :~/botva/src/src$
> > >> DISPLAY=":0" xrandr --output 0x54 --set Colorspace Default
> > >>
> > >> Anything but Default resulted in VERY pink (nearly red!) image. Not
> sure
> > >> if this is limitation of 8bit per channel mode or hdmi to vga
> adapter I
> > >> use?
>
> Maybe https://gitlab.freedesktop.org/drm/amd/-/issues/3358 ?
>
Ah, thanks, looks similar, while I tested with fullscreen firefox window
(triggered xrandr from remote ssh session).
>
>
>
>
>
> Thanks,
> pq
>
1
0
Today you can find more refined ideas, papers and implementations, but it
all started somewhere ...
https://vccimaging.org/Publications/Rempel2007L2H/Rempel2007L2H.pdf
2
5