March 2020 - Cin - Cinelerra-GG Mailing Lists

OpenEXR encoding is not multithreaded?
by Andrew Randrianasulu 14 Mar '20

14 Mar '20

Hello again! I had this qestion about why I can't see OpenEXR as choise in 'background render' format chooser. Now I have answer.... I tried to composite 2 x 1920x1080 images (project format RGBA-float) Speed while just previewing was up to 5.5 fps (on full speed.) Now, rendering 10 seconds of such video took 7 min 17 sec when I choose OpenEXR, include alpha, compression pxr24. (into tmpfs, full speed usually more than 500Mb/s) Each file took around 2.4 Mb. Definitely not realtime. Of course, even perfect multithreading on 4-way machine like I have will not make it even closer to realtime - but I think it will make such output format still more useful. If frame-level multithreading in OpenEXR encoder as done by CinGG can be implemented relatively trivially - it will be great for such speed addict as me :} I mean, like creating pool of image buffers in encoder itself, with n = num_of_cpus, but not more than, say total_buf_size than 1/2 ram. Then call OpenEXR compressor on each buffer. Then wait until one buffer become ready, write it, load buffer with new image. Repeat untill all images are compressed. Of course there is question about ordering ..or not? If OpenEXR actually 'sequence' of images, like I choose - each file can be written independently, even? If image order somewhat important - equip each buffer with serial number, and then reuse timestamp info from core? Also, on question of GPU precision ... Extended-Precision Floating-Point Numbers for GPU Computation from Andrew Thall, Alma College Abstract addition " [Addendum (July 2009): the presence of IEEE compliant double-precision hardware in modern GPUs from NVidia and other manufacturers has reduced the need for these techniques. The double-precision capabilities can be accessed using CUDA or other GPGPU software, but are not (as of this writing) exposed in the graphics pipeline for use in Cg-based shader code. Shader writers or those still using a graphics API for their numerical computing may still ﬁnd the methods described herein to be of interest.]" " Testing accuracy and precision of hardware systems is problem- atic; for a survey see Cuyt et al [23]. The Paranoia system [24], developed by Kahan in the early 1980s, was the inspiration for GPU Paranoia by Hillesland and Lastra [25], a software system for characterizing the ﬂoating point behavior of GPU hardware. GPU Paranoia provides a test-suite estimating ﬂoating point arithmetic error in the basic function, as well as char- acterizing apparent number of guard digits and correctness of rounding modes. Recent GPU hardware has shown considerable improvement over earlier. For the purposes of extended-precision computation, 32-bit ﬂoats are required. For Nvidia GPUs, the 6800 series and above were the ﬁrst to have sufﬁcient precision and IEEE compliance for extended-precision numerical work; the 7800/7900/7950 series has better IEEE compliance, allowing all but the transcendental functions to be computed; for computa- tions involving Taylor series, rather than self-correcting Newton- Raphson iterations, the superior round-off behavior and guard- digits of the 8800 series are necessary." ---quote end--- paper: http://andrewthall.org/papers/df64_qf128.pdf But I guess if Natron is happy with OpenGL - then Cinelerra also can be ....

1 0

New libdav1d, ICC profile support in ffm[eg git
by Andrew Randrianasulu 13 Mar '20

13 Mar '20

Fresh news! https://www.phoronix.com/scan.php?page=news_item&px=Dav1d-0.6-AV1-Benchmarks ---quote--- But where things really got interesting for the Intel/AMD processors was the much faster 10-bit decoding of AV1 content with dav1d 0.6.0 for both Intel and AMD. Very significant gains in this case. ---quote end--- In ffmpeg we can see https://git.ffmpeg.org/gitweb/ffmpeg.git/commit/05d27f342be28cf92f3c9470e70… API: add AV_PKT_DATA_ICC_PROFILE to AVPacketSideDataType and next few patches .... comment in avcodec.h says: ---quote-- + * ICC profile data consisting of an opaque octet buffer following the + * format described by ISO 15076-1. + */ + AV_PKT_DATA_ICC_PROFILE, --end of quote---- pdf itself is huge, but hopefully other libs already have something? http://www.color.org/icc_specs2.xalter v4 ICC Specification This version is technically identical to ISO 15076-1:2010, which is available from ISO. This is the most widely used version, first published in 2001. Specification ICC.1:2010-12 (Profile version 4.3.0.0) Image technology colour management - Architecture, profile format, and data structure http://www.color.org/specification/ICC1v43_2010-12.pdf Strange enough, but website littlecms.com doesn't work for me .... Anyway, hopefully no-one need to *print* films right now, so it all will be about monitors and files.. I may misunderstand something, but whole point of color management is to get same color on slightly different devices? So, if you work with Monitor A and you have color profile for your monitor, you can store this profile somewhere in image and now video file, and then on load supported player/editor will auto-correct things, relative to another (monitor B) output device, so you can work on colors across few monitors in different locations, say. I may misunderstand this, never feel special need for such fine-tuning of colors.

2 1

OpenMP
by Andrew Randrianasulu 12 Mar '20

12 Mar '20

Hi, all! Currently I'm experimenting with OpenMP https://bisqwit.iki.fi/story/howto/openmp/ --quote--- Support in different compilers GCC (GNU Compiler Collection) supports OpenMP 4.5 since version 6.1, OpenMP 4.0 since version 4.9, OpenMP 3.1 since version 4.7, OpenMP 3.0 since version 4.4, and OpenMP 2.5 since version 4.2. Add the commandline option -fopenmp to enable it. OpenMP offloading is supported for Intel MIC targets only (Intel Xeon Phi KNL + emulation) since version 5.1, and to NVidia (NVPTX) targets since version 7 or so. [...] The syntax All OpenMP constructs in C and C++ are indicated with a #pragma omp followed by parameters, ending in a newline. The pragma usually applies only into the statement immediately following it, except for the barrier and flush commands, which do not have associated statements. The parallel construct The parallel construct starts a parallel block. It creates a team of N threads (where N is determined at runtime, usually from the number of CPU cores, but may be affected by a few things), all of which execute the next statement (or the next block, if the statement is a {…} -enclosure). After the statement, the threads join back into one. #pragma omp parallel { // Code inside this region runs in parallel. printf("Hello!\n"); } This code creates a team of threads, and each thread executes the same code. It prints the text "Hello!" followed by a newline, as many times as there are threads in the team created. For a dual-core system, it will output the text twice. (Note: It may also output something like "HeHlellolo", depending on system, because the printing happens in parallel.) At the }, the threads are joined back into one, as if in non-threaded program. Internally, GCC implements this by creating a magic function and moving the associated code into that function, so that all the variables declared within that block become local variables of that function (and thus, locals to each thread). ICC, on the other hand, uses a mechanism resembling fork(), and does not create a magic function. Both implementations are, of course, valid, and semantically identical. Variables shared from the context are handled transparently, sometimes by passing a reference and sometimes by using register variables which are flushed at the end of the parallel block (or whenever a flush is executed). --quote end--- http://gregslabaugh.net/publications/OpenMP_SPM.pdf Multicore Image Processing with OpenMP Greg Slabaugh, Richard Boyes, Xiaoyun Yang https://nccastaff.bournemouth.ac.uk/jmacey/OpenMP/ -quote- OpenMP by Rob Bateman Introduction OpenMP is an open standard that lets you easily make use of multi-threaded processors. It's currently supported by the following compilers: Visual C++, gcc (though not the Win32 version that comes with cygwin), XCode, and the Intel compiler; and It's supported on the following platforms: Win32, Linux, MacOS, XBox360*, and PS3*. * Not amazingly well on those platforms --quote end-- I used bcast2000 example , namely bcast/overlayframe.C and those CFLAGS: CFLAGS = -O3 -fpermissive -fomit-frame-pointer -march=pentium3 -ffast-math -mfpmath=both -fopenmp -I/usr/local/include + enabled linking with libgomp (gcc 5.5.0) by adding -lgomp to bcast-2000c/bcast/Makefile it makes code slower, so far :} but it eats all processors :} unlike original code

3 4

cinelerra_scale.C
by Andrew Randrianasulu 12 Mar '20

12 Mar '20

I have stupid question again... in file https://git.cinelerra-gg.org/git/?p=goodguy/cinelerra.git;a=blob;f=cinelerr… I can see such lines: 60 61 void ScaleThread::run() 62 { 63 if(already_running) return; 64 already_running = 1; 65 constrain_ratio = mwindow->defaults->get("SCALECONSTRAIN", 0); 66 scale_data = mwindow->defaults->get("SCALEDATA", 0); 67 auto_aspect = mwindow->defaults->get("AUDIOASPECT", 0); 68 offsets[0] = offsets[1] = offsets[2] = offsets[3] = 0; 69 70 orig_dimension[0] = dimension[0] = mwindow->session->track_w; 71 orig_dimension[1] = dimension[1] = mwindow->session->track_h; 72 orig_dimension[2] = dimension[2] = mwindow->session->output_w; 73 orig_dimension[3] = dimension[3] = mwindow->session->output_h; 74 ratio[0] = ratio[1] = ratio[2] = ratio[3] = 1; 75 aspect_w = mwindow->session->aspect_w; 76 aspect_h = mwindow->session->aspect_h; 77 78 window = new ScaleWindow(this); 79 window->create_objects(); 80 int result = window->run_window(); 81 if(!result) 82 { 83 int dummy_offsets[4]; 84 dummy_offsets[0] = dummy_offsets[1] = dummy_offsets[2] = dummy_offsets[3] = 0; 85 // Fake the offsets if data is scaled. 86 87 // fix tracks 88 //mwindow->stop_playback(1); 89 // save the before undo 90 mwindow->undo->update_undo_edits(_("Resize"), 0); 91 mwindow->tracks->scale_video(dimension, scale_data ? dummy_offsets : offsets, scale_data); 92 mwindow->session->track_w = dimension[0]; 93 mwindow->session->track_h = dimension[1]; 94 mwindow->session->output_w = dimension[2]; 95 mwindow->session->output_h = dimension[3]; 96 mwindow->session->aspect_w = aspect_w; 97 mwindow->session->aspect_h = aspect_h; 98 mwindow->video_window->resize_window(); 99 mwindow->draw(); 100 mwindow->undo->update_undo_edits(); 101 mwindow->session->changes_made = 1; 102 mwindow->defaults->update("ASPECTW", aspect_w); 103 mwindow->defaults->update("ASPECTH", aspect_h); 104 mwindow->defaults->update("AUTOASPECT", auto_aspect); 105 } 106 delete window; 107 108 mwindow->defaults->update("SCALECONSTRAIN", constrain_ratio); 109 mwindow->defaults->update("SCALEDATA", scale_data); 110 already_running = 0; 111 } shouldn't line 67 reads as AUTOASPECT instead of AUDIOASPECT ? Found this while trying to debug bc2000 crash issue ...

2 2

website down Error 521 Ray ID: 572de7331a9ec447 • 2020-03-12 13:36:16 UTC Web server is down
by W P 12 Mar '20

12 Mar '20

Just wanted to point out that https://www.cinelerra-gg.org/ server's is down, don't know if Sam is doing maintenance so I just sent this message.

2 1

Creating subtitles .srt
by Andrea paz 11 Mar '20

11 Mar '20

@IgorBeg and all. I was watching your subtitle video. How do you create the .srt file? Manually with a plain text editor? Are there programs for this?

3 4

Unable to install latest cin package in debian testing
by jEsuSdA 8) 10 Mar '20

10 Mar '20

Hi! I was able to install cin package in my debian testing since long time ago, but the latest cin package causes dependency error. latest package depends on libva1 libva-x11-1 libva-drm1 But in debian stable and testing repositories there are the packages libva2 libva-x11-2 libva-drm2 Could you provide a package with these dependencies to allow the debian users to use latest Cinelerra versions ? Thanks a lot.

2 2

Offtopic, but I put up github with modified Broadcast2000
by Andrew Randrianasulu 10 Mar '20

10 Mar '20

Hugely off, but .. I like to have my own toy video editor :} https://github.com/Randrianasulu/broadcast2000/commits/master Video from yesterday: https://www.youtube.com/watch?v=j1xvaWXEXzA So far it works on IA32 :} (actually, kernel is 64-bit, but all userspace still 32-bit, this may contribute to some crash-on-pause behaviour, if playing audio via aoss) I think I'll try to make yuvpipe output, and may be try to fix alsa audio ... Actually, it seems 'floating point alpha' in this version of Broadcast kills perf most :} 2 fps on 1280x720 video if I use fader .... at 1.4Ghz x 4 But still nearly 12 fps for two-track video (1280x720, mjpeg - 75% and 50% faders) if I just use alpha channel and set CPU to max - 3.9 x 4 ghz :} Thing is, apparently 'bilinear interpolation' uses exactly FP alpha, so it will be SLOW on HDTV sized videos - time to dig into source to see if scaled interpolation can be used with just integer alpha channels Have nice days {nights}, and good health!

2 1

Re: [Cin] Shortcuts
by daniel-cinelerra-gg＠opayq.com 08 Mar '20

08 Mar '20

How hard would it be to create a menu for user-editable shortcuts? Ardour/mixbus has something like this, and I find it really, really convenient. Would it be possible to adapt their code, or would it require huge rewrites? http://www.harrisonconsoles.com/mixbus/mixbus32c-6-live-manual/1/en/topic/k… Dan On Sat, Mar 7, 2020 at 12:18 PM Igor BEGHETTO < > wrote: > > > > Hi! > > > *IMPORTANT: the shortcut for “Delete last track” has been changed* > > from the lower case letter “d” > > to Ctrl-Shift-D. Having just “d” was somewhat dangerous per Dan > > (Quinton?) because off screen. > > A “format” option button for ffmpeg in the Render menu shows up when > > you click on the Video > > wrench icon. This provides a few more tuning operators for the render > > for choices like mp4. > I think that the shorcuts are useful for the most/frequently used > operation/action. Cinelerra have some shortcuts where, I think, you can > do without them: for example in File Menu there are "BD Render..." > (Ctrl+D) and "DVD Render" (Alt+D). > I think it would be better to assign Ctrl+D shortcut for "Delete last > track" instead of Ctrl+Shif+D. > So, more intuitive, you have Shift+D to "Delete First track", and Ctrl+D > to "Delete Last track": Shift is the key up (then first), and Ctrl is > the key down (then last) in the keyboard. > > > About "shortcut for going trough *Autos* keyframes" thread. > > We are going to "attempt" to add code to use a shortcut of Alt-a for > > next fader/speed/autos and Ctrl-a for previous auto. This should come > > in handy. > I think, if possible, to use the same shortcuts of the Plugin's > keyframes because they are always keyframes, for me. > > Thanks > > IgorBeg > > > > > -- > Cin mailing list > Cin(a)lists.cinelerra-gg.org > https://lists.cinelerra-gg.org/mailman/listinfo/cin >

4 6

shortcut for going trough keyframes
by deim31 08 Mar '20

08 Mar '20

Hi, I sometimes want to walk through keyframes - in last situation I used video and audio faders so I managed to search shortcut in manual and I found there key [f] for that (go to next keyframe). But it seems it doesn't work the way I thought. It moves cursor to end (or begin) of track. I also noticed when I set fader keyframe I can't return to the same position again. I don't know if it's some feature or error. Thanks and have a nice day :-) deim

3 6