- Cin - Cinelerra-GG Mailing Lists

Checked into GIT some of Andrew's latest changes
by Phyllis Smith 11 Feb '26

11 Feb '26

Minor changes to the following files was checked into GIT (these were the accumulated patches BEFORE Andrew re-sent them all to make it easier for us/Andrea to test, so the numbers may vary). ffmpeg.C - for mediacodec / Termux 0002-Mediacidec-decode-to-use-with-remap_video_codec.patch window.C - for yuav16 formats with env. variable 0002-Allow-16bpc-only-with-ALLOW_OLD_16BPC-env.-var.patch Makefile.am - for potential speed up "make install" 0002-Make-parallel-make-install-work.patch 0001-Do-not-use-smart-filename-expansion-in-Makefile.am.patch ffv2_vulcan.mov - came from email message so users have at least one render format if vulkan works msg/txt - just consolidated more important changes done in 2025 I did NOT add in 0001-Experimental-vulkan-encode.patch because I have questions about the use of "type", although it did compile ffmpeg.C without errors. Also, did not have time to look at Ladspa change.

3 10

LADSPA plugin errors from onevpl appimage on T2
by Andrew Randrianasulu 10 Feb '26

10 Feb '26

root@Standard_PC:~# DISPLAY="10.0.2.2:1" ./squashfs-root/usr/bin/cin Cinelerra Infinity - built: Feb 4 2026 16:44:02 git://git.cinelerra-gg.org/goodguy/cinelerra.git (c) 2006-2019 Heroine Virtual Ltd. by Adam Williams 2007-2020 mods for Cinelerra-GG by W.P.Morrow aka goodguy 2003-2017 mods for Cinelerra-CV by CinelerraCV team 2015-2025 mods for Cinelerra-GG by Cinelerra-GG team Libav version: Lavc62.11.100 Cinelerra is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. There is absolutely no warranty for Cinelerra. lang changed from '' to 'C.UTF-8' build plugin index for: /root/squashfs-root/usr/bin/plugins build lv2 index for: $CIN_PATH/lv2 build ladspa plugin index for: /root/squashfs-root/usr/bin/ladspa int PluginServer::open_plugin(int, Preferences*, EDL*, Plugin*): PluginServer::open_plugin: load_obj /root/squashfs-root/usr/bin/ladspa/imp_1199.so = libfftw3f.so.3: cannot open shared objy int PluginServer::open_plugin(int, Preferences*, EDL*, Plugin*): PluginServer::open_plugin: load_obj /root/squashfs-root/usr/bin/ladspa/mbeq_1197.so = libfftw3f.so.3: cannot open shared oby int PluginServer::open_plugin(int, Preferences*, EDL*, Plugin*): PluginServer::open_plugin: load_obj /root/squashfs-root/usr/bin/ladspa/pitch_scale_1193.so = libfftw3f.so.3: cannot open shy int PluginServer::open_plugin(int, Preferences*, EDL*, Plugin*): PluginServer::open_plugin: load_obj /root/squashfs-root/usr/bin/ladspa/pitch_scale_1194.so = libfftw3f.so.3: cannot open shy MWindow::init_ladspa_plugins: can't create ladspa plugin index: /root/.bcast5/ladspa_plugins._root_squashfs-root_usr_bin_laa so, list is bin/ladspa/imp_1199.so bin/ladspa/mbeq_1197.so bin/ladspa/pitch_scale_1193.so bin/ladspa/pitch_scale_1194.so I think those should be added to "-e" directive in bld_appimage.sh script ?

2 2

Makefile for libdav1d 1.5.3
by Andrew Randrianasulu 10 Feb '26

10 Feb '26

I tried to modify our thirdparty/src/dav1d-0.5.1.patch so it will build static version of new libdav1d library (I think ffmpeg.git dropped support for pre-1.0 versions of dav1d). need nasm 2.14+ I build it on Slackware15.0 i586 with gcc11.2.0 and nasm 2.15.05 (there was some problem with multiple definitions in shared library, but we do not use it, so I just commented out linking it) orig file: https://ftp.debian.org/debian/pool/main/d/dav1d/ Notably I can't see 0.5.1 there (probably still exist on videoLAN site somewhere?)

1 0

Making another round around HDR ....
by Andrew Randrianasulu 09 Feb '26

09 Feb '26

Still no monitor so it all theory .. err, internet search! I wondered how DaVinchi Resolve resolved that " HDR preview on non-dedicated monitor" problem .. Turned out .. they did not on Windows until like version 19 ? https://forum.blackmagicdesign.com/viewtopic.php?f=33&t=116599 ==== [image: Post] <https://forum.blackmagicdesign.com/viewtopic.php?p=998264#p998264>Mon Nov 13, 2023 3:06 pm Currently Resolve renders image to the screen over OpenGL surface. Because OpenGL can output 30-bit images, we can also get automatically 10-bit on Windows. But HDR on Windows cannot get output via OpenGL. DirectX code and conversion to Rec2020 curve is required. This feature requires a lot of Resolve code changes with many potential bugs and workarounds. So, I think the developers will not create HDR in a near future, before Windows HDR OpenGL output in GPU drivers. ==== few months later ===== [image: Post] <https://forum.blackmagicdesign.com/viewtopic.php?p=1062398#p1062398>Wed Jul 10, 2024 12:45 am Resolve supports HDR in Windows now (yay). However currently you need to launch the app on the HDR monitor (you can move it to another monitor afterwards), or it will not enable HDR in any viewer. To do that, first move it to this monitor, exit and reload (it will now open on that monitor and HDR will be enabled) And unfortunately again, even with this workaround, it now enables HDR on all viewers, even those that are on non-HDR monitors. On the plus side, the Clean Feed will work correctly on the HDR monitor at least (I use it with an HDR projector, without any HDR capable PC monitors). ===== I wonder if it works via proton/wine? :p Pekka, sorry for bothering you, but do you have some Vulkan-encode capable hardware? I'm trying to add some support for this in cingg, but it sort of hard without hardware?

2 2

test images (8/16bpc)
by Andrew Randrianasulu 09 Feb '26

09 Feb '26

https://testimages.org/color/ Firefox complained about https but opened the site ==== from description ==== For every resolution listed above and for both 8bpp and 16bpp, the following 7 images were generated: - color_bars_CMYKWRGB : Eight vertical bars at maximum intensity (255 for 8bpp and 65535 for 16bpp) representing different colors in the following order: Cyan (Green + Blue), Magenta (Red + Blue), Yellow (Red + Green), Black, White (Red + Green + Blue), Red, Green, Blue. - color_bars_CMYKWRGB_100IRE : Variation of the color_bars_CMYKWRGB where the minimum intensity level is set to 7.5IRE (16 for 8 bit, 4096 for 16 bit) and the maximum intensity level is set to 100IRE (235 for 8 bit, 60160 for 16 bit). - color_bars_CMYKWRGB_75IRE : Variation of the color_bars_CMYKWRGB where the minimum intensity level is set to 7.5IRE (16 for 8 bit, 4096 for 16 bit) and the maximum intensity level is set to 75IRE (180 for 8 bit, 46144 for 16 bit). - color_bars_CMYKWRGB_gradient : Variation of the color_bars_CMYKWRGB where intensity levels varies linearly from top 0% to bottom 100%. - color_rainbow : Six vertical color sections with the following intensity variations to obtain a "rainbow" effect: 1. red=MAX, green=0 to MAX, blue=0; 2. red=MAX to 0, green=MAX, blue=0; 3. red=0, green=MAX, blue=0 to MAX; 4. red=0, green=MAX to 0, blue=MAX; 5. red=0 to MAX, green=0, blue=MAX; 6. red=MAX, green=0, blue=MAX to 0. - color_rainbow_gradient : Variation of the color_rainbow where intensity levels varies linearly to obtain a black to white variation from top to bottom. - color_SMPTE_RP_219_2002 : This image is generated as described in SMPTE RP 219:2002 and resized to perfectly match different resolution. === download link lead to sourceforge https://sourceforge.net/projects/testimages/files/COLOR/16BIT/ If you click on small "i" you get filesize, usually not very big (500 kb or so)

1 2

Vulkan/oneVPL testing AppImage
by Phyllis Smith 09 Feb '26

09 Feb '26

There is now a vulkan and onevpl included AppImage for testing purposes at the website Download area: https://download.cinelerra-gg.org/?path=testing&sort=date_desc file name is vulkan_onevpl-x86_64.AppImage. It includes the latest vulkan-wip.diff patch tne the ffv1_vulkan.mov render format. (And THIS TIME, I verified with "nm -D" to make absolutely sure it contained qsv, vulkan, and VPL symbols!)

3 4

ONEVPL_SEARCH_PATH
by Andrew Randrianasulu 08 Feb '26

08 Feb '26

https://github.com/obsproject/obs-studio/pull/10047 May be setting this to some location where other parts of onevpl live on suse will help with appimage's use of onevpl? I'll inspect appimage, at least some traces of qsv should be there .....

2 16

Build from git
by Andrea paz 08 Feb '26

08 Feb '26

I tried to build with the 5 patches that Phyllis used. Everything is OK, no errors. Does the “0002-Make-Paralell...” patch perhaps serve to speed up compilation? It seemed faster to me, but I would need to do more testing. In ./configure, I tried adding “--with-vulkan” and, for the first time, it worked, and now CinGG sees my Vulkan decoder (even though it doesn't work with h264). Decoder h264 does not support device type vulkan. HW device init failed, using SW decode. NB: I have to use X11-OpenGL, otherwise Vulkan is not recognized. For the encoder, I have: /home/paz/cinelerra/cinelerra-5.1/thirdparty/ffmpeg-8.0/./ffmpeg -encoders | grep vulkan V....D av1_vulkan AV1 (Vulkan) (codec av1) V....D h264_vulkan H.264/AVC (Vulkan) (codec h264) V....D hevc_vulkan H.265/HEVC (Vulkan) (codec hevc) When I try to render, I don't see any presets with Vulkan. Do I have to create them myself? Finally, what tests should I run for “yuva16”?

3 3

VERY WIP patch to test Vulkan encoding
by Andrew Randrianasulu 07 Feb '26

07 Feb '26

Might (even most likely!) break many other encoding devices! with such profile (wrongly named but still) cat /usr/share/cin/ffmpeg/video/ffv1_vulkan.mov mov hevc_vulkan cin_pix_fmt=nv12 cin_hw_device=vulkan #coder=0 #context=1 #g=1 level=4 strict=-2 and attached patch I finally got: root@slax:/dev/shm/tmp/cinelerra-goodguy-20260206/cinelerra-5.1# bin/cin Cinelerra Infinity - built: Feb 6 2026 06:54:53 git://git.cinelerra-gg.org/goodguy/cinelerra.git (c) 2006-2019 Heroine Virtual Ltd. by Adam Williams 2007-2020 mods for Cinelerra-GG by W.P.Morrow aka goodguy 2003-2017 mods for Cinelerra-CV by CinelerraCV team 2015-2026 mods for Cinelerra-GG by Cinelerra-GG team Libav version: Lavc62.11.100 Cinelerra is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. There is absolutely no warranty for Cinelerra. BC_DisplayInfo::gl_fb_config failed [Vulkan @ 0xc22e8c40] Supported layers: [Vulkan @ 0xc22e8c40] VK_LAYER_LUNARG_gfxreconstruct [Vulkan @ 0xc22e8c40] VK_LAYER_KHRONOS_validation [Vulkan @ 0xc22e8c40] VK_LAYER_KHRONOS_synchronization2 [Vulkan @ 0xc22e8c40] VK_LAYER_LUNARG_monitor [Vulkan @ 0xc22e8c40] VK_LAYER_KHRONOS_shader_object [Vulkan @ 0xc22e8c40] VK_LAYER_KHRONOS_timeline_semaphore [Vulkan @ 0xc22e8c40] VK_LAYER_LUNARG_api_dump [Vulkan @ 0xc22e8c40] VK_LAYER_KHRONOS_profiles [Vulkan @ 0xc22e8c40] VK_LAYER_LUNARG_screenshot [Vulkan @ 0xc22e8c40] GPU listing: [Vulkan @ 0xc22e8c40] 0: AMD Radeon RX 550 / 550 Series (RADV POLARIS12) (discrete) (0x699f) [Vulkan @ 0xc22e8c40] 1: llvmpipe (LLVM 20.1.3, 256 bits) (software) (0x0) [Vulkan @ 0xc22e8c40] Device 0 selected: AMD Radeon RX 550 / 550 Series (RADV POLARIS12) (discrete) (0x699f) [Vulkan @ 0xc22e8c40] Using device extension VK_KHR_push_descriptor [Vulkan @ 0xc22e8c40] Using device extension VK_EXT_descriptor_buffer [Vulkan @ 0xc22e8c40] Using device extension VK_EXT_physical_device_drm [Vulkan @ 0xc22e8c40] Using device extension VK_EXT_shader_atomic_float [Vulkan @ 0xc22e8c40] Using device extension VK_EXT_shader_object [Vulkan @ 0xc22e8c40] Using device extension VK_KHR_shader_subgroup_rotate [Vulkan @ 0xc22e8c40] Using device extension VK_KHR_shader_expect_assume [Vulkan @ 0xc22e8c40] Using device extension VK_KHR_external_memory_fd [Vulkan @ 0xc22e8c40] Using device extension VK_EXT_external_memory_dma_buf [Vulkan @ 0xc22e8c40] Using device extension VK_KHR_external_semaphore_fd [Vulkan @ 0xc22e8c40] Using device extension VK_EXT_external_memory_host [Vulkan @ 0xc22e8c40] Queue families: [Vulkan @ 0xc22e8c40] 0: graphics compute transfer sparse (queues: 1) [Vulkan @ 0xc22e8c40] 1: compute transfer sparse (queues: 4) [Vulkan @ 0xc22e8c40] 2: sparse (queues: 1) [Vulkan @ 0xc22e8c40] Using device: AMD Radeon RX 550 / 550 Series (RADV POLARIS12) [Vulkan @ 0xc22e8c40] Alignments: [Vulkan @ 0xc22e8c40] optimalBufferCopyRowPitchAlignment: 1 [Vulkan @ 0xc22e8c40] minMemoryMapAlignment: 4096 [Vulkan @ 0xc22e8c40] nonCoherentAtomSize: 64 [Vulkan @ 0xc22e8c40] minImportedHostPointerAlignment: 4096 [Vulkan @ 0xc22e8c40] Using queue family 0 (queues: 1) for graphics [Vulkan @ 0xc22e8c40] Using queue family 1 (queues: 4) for compute transfers type, inserting hwupload filter .. err: Error number 11 occurred detected 4 logical cores [hwupload @ 0xe493e480] A hardware device reference is required to upload frames to. Vulkan pixfmt! [hevc_vulkan @ 0xe4918340] Device does not support the VK_KHR_video_encode_queue extension! FFMPEG::open_encoder err: Функция не реализована int FFMPEG::open_encoder(const char*, const char*): open failed hevc_vulkan:/dev/shm/y4m_tut.mov Render::render_single: Session finished. Total excess of backups: -50 Session time: 0:00:24 Cpu time: user: 0:00:03.668 sys: 0:00:00.911 So, it goes deeper now?

1 0

Vulkan encoding re-search
by Andrew Randrianasulu 07 Feb '26

07 Feb '26

MLT for example set up hwupload filter https://github.com/mltframework/mlt/commit/e21b23b72f0b44a88f5a2fd04d697630… I'll look at how to set this up in our case.

1 0