- Cin - Cinelerra-GG Mailing Lists

libvpx via ffmpeg on high-cpu machines
by Andrew Randrianasulu 26 Jan '23

26 Jan '23

I think this one can be safe to test, if you have >16 cpu/thread machine ... https://git.ffmpeg.org/gitweb/ffmpeg.git/commit/c13d95934327ddad4db30f6aee9…

3 7

Complete system hangs (likely out of memory)
by Stefan de Konink 25 Jan '23

25 Jan '23

Hi, When editing with Cinelerra (current git) I experience frequently that my entire system hangs after scratching through either the timeline or the viewer. As experienced before I have noticed that Cinerella eats a lot of memory. The total amount of assets I have currently in my project is about 70GB, my system memory is 16GB, and yesterday I have added 8GB of swap (normally no swap). It seems that with every video that I preview my memory usage significantly increases, but is never released. I think we have discussed this behavior on the mailinglist before being that the indices never get released. But still the amount does not make sense to me. As user my expectation would be that when I would close the *viewer*, the memory would get released, especially when the content is not placed on the timeline. Are there any statistics available on the memory usage what goes where? Practical question: is it possible to have a different path for the emergency backups than /tmp? -- Stefan

5 26

Re: [Cin] Cin Digest, Vol 51, Issue 40
by Richard Nolde 24 Jan '23

24 Jan '23

Regarding memory leaks: > 1. Re: Complete system hangs (likely out of memory) > (Stefan de Konink) > 2. Re: Complete system hangs (likely out of memory) > (Stefan de Konink) > 3. Re: Complete system hangs (likely out of memory) > (Andrew Randrianasulu) The following tool has been instrumental in finding memory leaks and/or peaks of usage in a large graphically intensive program that I have helped another developer with from time to time. https://apps.kde.org/heaptrack/ https://github.com/KDE/heaptrack It comes with text and GUI analyzer tools to allow you to see which functions allocate/release/leak memory during the run of a program and it does not slow down the execution nearly as much as valgrind. It uses a replacement version of malloc and the LD_PRELOAD option to redirect system calls into its own instrumented versions of memory related system calls. Originally written by Millan Wolff in Berlin, it has now become a part of the KDE development environment but it does not require you to run a full KDE desktop. I run in with XFCE on Fedora 36 without any problems. Richard Nolde

2 1

Re: [Cin] Complete system hangs (likely out of memory)
by Andrew Randrianasulu 24 Jan '23

24 Jan '23

пн, 23 янв. 2023 г., 18:45 Stefan de Konink <stefan(a)konink.de>: > Hi Phyllis, > > With the suggestion of Richard Nolde, I decided to do a day of work in > Cinerella with heaptrack enabled in the background. It was a pleasant > surprise how well it performed. (And no crashes ;) > > And I am also surprised how functional Heaptrack itself has been designed, > I'll forward this application to my peers. To screenshots attached. > Thanks for being curious, Stephan! I looked at attached screenshots and I found it strange how Peaked and Leaked basically stays very near each other for each function line? > If anyone is intested in my heaptrack file, I can obviously make it > available. But I would really recommend adding this tool to your toolbelt! > > -- > Stefan

3 24

Re: [Cin] Complete system hangs (likely out of memory)
by Stefan de Konink 23 Jan '23

23 Jan '23

Hi Phyllis, With the suggestion of Richard Nolde, I decided to do a day of work in Cinerella with heaptrack enabled in the background. It was a pleasant surprise how well it performed. (And no crashes ;) And I am also surprised how functional Heaptrack itself has been designed, I'll forward this application to my peers. To screenshot attached. If anyone is intested in my heaptrack file, I can obviously make it available. But I would really recommend adding this tool to your toolbelt! -- Stefan

1 0

Multiple viewers (like mixer, but without it)
by Stefan de Konink 18 Jan '23

18 Jan '23

I have multiple video tracks. I would like to play them simultaniously. Is it possible to assign a viewer to a track, without going through the mixer step initially? Preferably I am looking for an output that does take in account the applied plugin, but does not do downward compositing (or the latter should be configurable). This is the documentation for the multi cam mixer; <https://cinelerra-gg.org/download/CinelerraGG_Manual/Multi_Camera_Mixer.html> It is mentioned "Using the track patchbay, mark the new tracks as mixer source to be added to the viewer." I fail to see how I could mark it (or if this is the correct approach). -- Stefan

2 2

add Frame on PiP on manual
by Andrea paz 17 Jan '23

17 Jan '23

I have added a new section to the Real-World appendix of the manual. It consists of adding a frame to a Picture in Picture and is based on a post in our forum. I enclose the latex file with image and also the text in odt for those who would like to review it. See if it is OK for you. Link to the files: https://ufile.io/zfshf1ct

5 8

small build fix for libbthread-master Makefile
by Andrew Randrianasulu 11 Jan '23

11 Jan '23

otherwise build fails on Freebsd 12.3 amd64 (used blds/bsd.bld)

2 2

Dual-screen support was working in cinCV at least
by Andrew Randrianasulu 10 Jan '23

10 Jan '23

Managed to get qemu (v7.2.0-334-g959b27f577-dirty) running with two virtual displays qemu-system-x86_64 -enable-kvm -m 2000 -cpu qemu64 -device virtio-vga,max_outputs=2 -device virtio-vga -cdrom /home/admin/CDROM-4.14.246-big-7.iso -smp 2 -device usb-mouse -M q35 -usb then in running system I run xconf (script configuring X from old Slax 6) and then startx bring me two displays. first i send xterm to second display by using DISPLAY=":0.1" from there I launchd cinelerra I typed :0.0 in CinCV's output config tab, and on next startup Compositor was on original first display..

2 3

Interesting confusion on X screens vs displays
by Andrew Randrianasulu 08 Jan '23

08 Jan '23

I tried to test multi-monitor support by firing another, nested X server Xephyr :1 -screen 1024x768 then Cin displays her windows on either DISPLAY :0 or :1 depend on preferences but I can\t set compositor to just one ...xserver. I looked into code :) in cwindowgui.C CWindowGUI::CWindowGUI(MWindow *mwindow, CWindow *cwindow) : BC_Window(_(PROGRAM_NAME ": Compositor"), mwindow->session->cwindow_x, mwindow->session->cwindow_y, mwindow->session->cwindow_w, mwindow->session->cwindow_h, xS(100), yS(100), 1, 1, 1, BC_WindowBase::get_resources()->bg_color, mwindow->get_cwindow_display()) in mwindow.C char *MWindow::get_cwindow_display() { char *x11_host = screens < 2 || session->window_config == 0 ? session->a_x11_host : session->b_x11_host; return *x11_host ? x11_host : 0; } and screens set in void MWindow::init_preferences() { preferences = new Preferences; preferences->load_defaults(defaults); File::setenv_path("LV2_PATH",preferences->lv2_path, 1); session = new MainSession(this); session->load_defaults(defaults); // set x11_host, screens, window_config screens = session->set_default_x11_host(); so into mainsession.C // set default x11 host, window_config, return screens int MainSession::set_default_x11_host(int win_config) { if( win_config < 0 ) win_config = window_config; const char *x11_host = win_config!=1 ? a_x11_host : b_x11_host; BC_DisplayInfo display_info(x11_host,0); int screen = display_info.get_screen(); if( screen < 0 && strcmp(a_x11_host, b_x11_host) ) { win_config = win_config==1 ? 0 : 1; x11_host = win_config!=1 ? a_x11_host : b_x11_host; display_info.init_window(x11_host,0); screen = display_info.get_screen(); } if( screen < 0 ) { x11_host = ""; display_info.init_window(x11_host,1); } int screens = 1; if( display_info.get_screen_count() > 1 ) screens = strcmp(a_x11_host, b_x11_host) != 0 ? 2 : 1; window_config = win_config; BC_Window::set_default_x11_host(x11_host); return screens; } and display_info lives in guicast/bcdisplayinfo.C and h in our case in h file int get_screen() { return scrnum; } and scrnum set to scrnum = DefaultScreen(display); in void BC_DisplayInfo::init_window(const char *display_name, int show_error) so, it seems having two 'x11 hosts" as separate X servers to send Compositor to actually not supported... ? we query number of screens on specific display/host ... so :0.0 and :0.1 works, while :1.0 and :0.0 not ? also, 3 or more displays will not work?

3 3