- Cin - Cinelerra-GG Mailing Lists

HDR test on youtube (vp9 HLG 10bit)
by Andrew Randrianasulu 06 May '25

06 May '25

https://m.youtube.com/watch?v=j0j-40NrGGU&pp=ygUMMTAwMCBuaXQgaGRy 110 mbytes,4 min, 4k ;)

1 1

BlendAlgebra plugin and cc="ccache gcc"
by Andrew Randrianasulu 06 May '25

06 May '25

I tried to rebuild cingg Slackware package with my custom Slackbuild and noticed that BlendAlgebra breaks if I left out initial line from year 2016 cc="ccache gcc" it seems that ccache breaks with unknown (for it) arguments like -noabi. Commenting this line from Slackbuild fixed it for me. At some point mass package building was done with ccache, but now probably only two Appimages build manually, so this error was unnoticed.

2 3

10bit VAAPI encoding test thread
by Andrew Randrianasulu 06 May '25

06 May '25

We branching! I looked at gpu-screen-recorder https://git.dec05eba.com/gpu-screen-recorder/tree/src/encoder/video/vaapi.c it sets AVHWFramesContext *hw_frame_context = (AVHWFramesContext*)frame_context->data; hw_frame_context->width = video_codec_context->width; hw_frame_context->height = video_codec_context->height; hw_frame_context->sw_format = self->params.color_depth == GSR_COLOR_DEPTH_10_BITS ? AV_PIX_FMT_P010LE : AV_PIX_FMT_NV12; hw_frame_context->format = video_codec_context->pix_fmt; hw_frame_context->device_ctx = (AVHWDeviceContext*)self->device_ctx->data; so may be we need to set AV_PIX_FMT_P010LE and not just P010?

2 21

Build crash in Arch
by Andrea paz 06 May '25

06 May '25

With guilty delay I tried to make a compilation to try Andrew's patch for Blend Algebra/Program. I got a crash; the fault is not with the patch because I keep getting crashes even without patching. I attach the cin5.log. (Won't it still be the fault of cmake, which has had several updates in Arch)?

3 7

blogpost about FCP's HDR support
by Andrew Randrianasulu 06 May '25

06 May '25

but it also contain link to big shot ARRI camera samples https://www.toodee.de/panicpost/final-cut-pro-x/1-2-fcpx-hdr-dolby-pq/ https://www.arri.com/en/learn-help/learn-help-camera-system/camera-sample-f… Also interesting to note that there seems to be usb-c to hdmi adapter! Not sure if such thing supported under Linux ...

2 1

So, I poked FCP 10.4 ....
by Andrew Randrianasulu 05 May '25

05 May '25

And it created such h264 test file (from 320p , lol). Interestingly, only mediainfo show transfer characteristics, not ffprobe from 7.1.1 (and this not cingg) https://cloud.mail.ru/public/chpM/VfPpDTnsv I wonder if I should put this as issue in ffmpeg's trac ...

2 1

CinGG Analog S-Video to Digital HDMI converter + ms2130 UHD HDMI-USB3 Capture card
by Terje J. Hanssen 05 May '25

05 May '25

https://www.mail-archive.com/[email protected]/msg05748.html I contine this new thread with an short extract from the long growing [Cin] ms2130 HDMI-USB3 UHD capture card - not quite as expected? But shouldn't a 422 and not a 420 output format be created? >> it seems in some cases ffmpeg not geared towards preserving most information, just picks up most commonly defined/first pixel format. >> Try mpeg2video as output codec, and/or specify pixel format (pix_fmt) more explicitly before output's filename? >>> Could you possibly examplify the latter? >>>> I think I just theorized ffmpeg may try to get some info from alsa device file directly, as it tries with /dev/video. But I never >>>> tried this way, so I can be totally wrong! ------------------------------- Instead of 50i (pal interlaced fields), I try 50 fps (progressive) that at least is a pal frequence. And yes, it seems from the output that ffmpeg uses mpeg1video as default output codec. Furter I try to combine the current line ffmpeg -hide_banner -f v4l2 -input_format yuyv422 -video_size 720x576 -framerate 25 -t 20 -i /dev/video2 out-720x576_ms2130.mpeg with and utilize the syntax developed 2021 for "Blu-ray compliant MPEG-2 SD video with LPCM audio" ffmpeg -i dv28.dv -c:v mpeg2video -refs 1 -bf 2 -b:v 25M -maxrate 25M -minrate 25M -bufsize 45M -muxrate 45M -dc 10 -c:a pcm_s16be -f vob dv28.mpg and for the current ongoing "DVD compliant MPEG-2 Video and PCM Audio" ffmpeg -hide_banner -i dv01.dv -c:v mpeg2video -refs 1 -bf 2 -b:v 8M -maxrate 8M -minrate 8M -bufsize 20M -muxrate 11M -dc 10 -c:a pcm_dvd -f vob -flags +ilme+ildct dv01.mpg to first ffmpeg -hide_banner -f v4l2 -input_format yuyv422 -video_size 720x576 -framerate 50 -c:v mpeg2video -refs 1 -bf 2 -b:v 8M -maxrate 8M -minrate 8M -bufsize 20M -muxrate 11M -dc 10 -c:a pcm_dvd -f vob -t 20 -i /dev/video2 720x576_50p_ms2130.mpg Option b:v (video bitrate (please use -b:v)) cannot be applied to input url /dev/video2 -- you are trying to apply an input option to an output file or vice versa. Move this option before the file it belongs to. Error parsing options for input file /dev/video2. Error opening input files: Invalid argument and next revised with no output. That is I need help to debug the syntax error: ffmpeg -hide_banner -f v4l2 -input_format yuyv422 -video_size 720x576 -framerate 50 -i /dev/video2 -c:v mpeg2video -refs 1 -bf 2 -b:v 8M -maxrate 8M -minrate 8M -bufsize 20M -muxrate 11M -dc 10 -c:a pcm_dvd -f vob -t 20 720x576_50p_ms2130.mpg [video4linux2,v4l2 @ 0x55db770ef9c0] Dequeued v4l2 buffer contains corrupted data (0 bytes). Input #0, video4linux2,v4l2, from '/dev/video2': Duration: N/A, start: 0.000000, bitrate: 331776 kb/s Stream #0:0: Video: rawvideo (YUY2 / 0x32595559), yuyv422, 720x576, 331776 kb/s, 50 fps, 50 tbr, 1000k tbn Stream mapping: Stream #0:0 -> #0:0 (rawvideo (native) -> mpeg2video (native)) Press [q] to stop, [?] for help [video4linux2,v4l2 @ 0x55db770ef9c0] Dequeued v4l2 buffer contains corrupted data (0 bytes). Last message repeated 30 times [mpeg2video @ 0x55db77114f00] Warning vbv_delay will be set to 0xFFFF (=VBR) as the specified vbv buffer is too large for the given bitrate! Output #0, vob, to '720x576_50p_ms2130.mpg': Metadata: encoder : Lavf59.27.100 Stream #0:0: Video: mpeg2video (4:2:2), yuv422p(tv, progressive), 720x576, q=2-31, 8000 kb/s, 50 fps, 90k tbn Metadata: encoder : Lavc59.37.100 mpeg2video Side data: cpb: bitrate max/min/avg: 8000000/8000000/8000000 buffer size: 20000000 vbv_delay: N/A frame= 0 fps=0.0 q=0.0 Lsize= 0kB time=00:00:00.00 bitrate=N/A speed= 0x video:0kB audio:0kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown Output file is empty, nothing was encoded (check -ss / -t / -frames parameters if used)

2 9

New Samsung codec "APV"
by Andrea paz 01 May '25

01 May '25

Ffmpeg has introduced (in decoding) a new “mezzanine codec” from Samsung: apv. There is also a proposal for an “OpenAPV” codec. The news is not particularly interesting: CinGG will read it because ffmpeg will read it; but the codec “draft” is useful and thorough in explaining how a codec is built and how it works. It can be found here: https://datatracker.ietf.org/doc/draft-lim-apv/ PS: I also attach an image of the “quality pentagon ” of a video. It is OT, but I particularly like it for its clarity and simplicity.

2 1

Updated AppImages for April 2025
by Phyllis Smith 30 Apr '25

30 Apr '25

There are now updated AppImages, Release Notes, and PDF/HTML manuals on the website and Andrey's packages are up to date at: https://github.com/einhander/cin-gg-packages/releases Release notes included below with some concern for the Speed mod as noted with no real concern for Blur and Titler. *GIT for Cinelerra-GG has the following changes from 04/01/2025-04/302025* *Blend Algebra and Blend Program plugins:* An additional SYSLIB Blend Program of tone_mapping.bp has been created by *Andrea* for tonemapping an HDR media and report its values within the range (0, 1.0). For usage, reference Example 6 in the file at: https://cinelerra-gg.org/download/testing/BlendPluginExamples/Examples.txt For netBSD, modifications were added by Andrew to fix shebang-perl and to not use ldl in Makefile. *Other small changes include:* 4.0 cmake requires 3 patches that *Andrew* has provided and Andrea had discovered/tested on Arch for x265 and libsvtav1. A speed bug has been found and was described reproducibly by *Lukas*. A one line change to auto.C seems to fix it, but it is unknown if reverting that line will lead to a previously diagnosed problem reoccuring. The old version of ffnvcodec (Video codec SDK) which was 10.0.26.0 was added back to the source as ffnvcodec-old.tar.xz to accommodate versions before the updated version of ffnvcodec for 12.2.72.0 was installed as the default. Blur and Titler plugins were modified so that these plugins can be used in combination with others (e.g. Blue Banana) without introducing the clipping into the process. Andrew provided the mod and Andrea tested it to ensure it worked without clipping and IgorB tested to make sure there was no impact to current usage of Blur and the Titler. Tiff was updated from 4.6.0t to 4.7.0.

1 0

Tried ffv1_vulkan ....
by Andrew Randrianasulu 29 Apr '25

29 Apr '25

It looks like tonemapping ate encoder's lunch ;) RADV_PERFTEST=video_decode,video_encode time -p ./ffmpeg -init_hw_device vulkan=vulkan -filter_hw_device vulkan -hwaccel vulkan -i ~/K38_sdcard1/Documents/iPhone11_4K-recorder_59.940HDR10.mov -vf libplacebo=format=yuv444p:w=1920:h=1080:force_original_aspect_ratio=decrease:normalize_sar=true:upscaler=ewa_lanczos:downscaler=ewa_lanczos:colorspace=bt709:color_primaries=bt709:color_trc=bt709:range=tv -c:a copy -c:v ffv1_vulkan -slicecrc 0 -level 4 -strict -2 -async_depth 5 -slices_v 32 -slices_h 32 -f matroska -benchmark /dev/null frame= 1148 fps=4.6 q=-0.0 Lsize= 2302672KiB time=00:00:19.16 bitrate=984419.6kbits/s speed=0.0769x bench: utime=28.586s stime=20.721s rtime=249.150s bench: maxrss=587076KiB real 251.00 user 28.67 sys 20.79 ===== But may be faster GPU/pcie bus like in Andrea's machine will speed it up enough to be practical

3 5