[Ffmpeg-devel-irc] ffmpeg.log.20180629

[burek]

[00:08:51 CEST] <Ars> teratorn, every time i have diferent exits
[00:09:31 CEST] <Ars> how can i check udp stream availability
[00:10:07 CEST] <Ars> am transcoding and multicasting stream like: fmpeg RTMP_Stream   udp://225.2.1.11:1300
[00:10:07 CEST] <Ars> Also i have stream check script (ffprobe udp://225.2.1.11:1300), but when executing that command ffmpeg stops.
[00:10:08 CEST] <Ars> I understand there is port binding problem.
[00:10:08 CEST] <Ars> But how i can check that UDP stream alive or no? How to chek stream on udp://225.2.1.11:1300 ?
[01:44:07 CEST] <teratorn> Ars: what is availability going to tell you that monitoring your failures isn't?
[02:41:40 CEST] <lemontea> I am having troubles getting a debug build of the ffmpeg with decklink sdk, while release is okay.  When --enable-debug --enable-static, I weirdly get this, value 'MT_StaticRelease' doesn't match value 'MTd_StaticDebug' in testfunc.obj	C:\dev\ffmpeglib\libavdevice.a(decklink_enc.o)
[02:43:20 CEST] <lemontea> The decklink stuff is in MT_StaticRelease code and I am trying to build it with /MTd in visual studio, bothers me that only Release works.  Before integrating the decklink SDK, it would work both debug/release.  Does anyone know how should I fix this? The MakeFile is not very obvious.
[02:43:28 CEST] <adjivas> Hi, please, what version of libquvi need ffmpeg?
[02:45:30 CEST] <adjivas> I try to FFmpeg-SIXEL but that not found by pkg-config
[12:32:16 CEST] <killown> ffmpeg -i "teste.mp4"  -vf "drawtext=text='test ':x=w-tw-20:y=h-th-100:fontsize=0.033*h:fontfile=/scripts/fonts/ttf/Hack-Bold.ttf:fontcolor=white at 0.6:shadowcolor=black at 0.6:shadowx=2:shadowy=2:box=1: boxcolor=black at 0.3:boxborderw=3" "out.mp4"
[12:32:27 CEST] <killown> how do I display the test in the top?
[12:35:27 CEST] <libertas> Hi, I need to use codec and profile of this line -codec:v h264 -qmin 0 -qmax 40 -crf 18 -profile:v high10 -pix_fmt yuv420p
[12:36:16 CEST] <libertas> in a lua script that has different settings:  format=yuv444p16,scale=in_color_matrix=$matrix,format=bgr24
[12:37:12 CEST] <libertas> in format, should it be format=yuv420pxx?  what should the xx be?  or is it not needed?
[12:41:05 CEST] <libertas> or put another way, would be nice if someone could build a similar line to the second I wrote using the setting of the first
[12:43:46 CEST] <ariyasu_> yes you want 4:2:0
[12:44:12 CEST] <ariyasu_> you probably want -vcodec libx264
[12:50:03 CEST] <libertas> ariyasu_: with the setting in the first line I can view videos in my panasonic tv, with newer profiles I can't, that's why
[12:50:35 CEST] <ariyasu_> your tv probably does not support 4:4:4
[13:23:22 CEST] <King_DuckZ> hi, I have this snippet of code https://paste.debian.net/1031274 which is asserting at line 15, the output I get from the commented snipped at line 9 is: https://paste.debian.net/1031273
[13:23:51 CEST] <King_DuckZ> anybody knows what's going on? shouldn't I get some memory allocated by the time the assert is reached?
[13:55:45 CEST] <King_DuckZ> picture->buf is also all null :/
[13:55:49 CEST] <King_DuckZ> wth?
[14:05:39 CEST] <c_14> picture->data[0] should be 0, no?
[14:05:54 CEST] <c_14> and it is so the assertion would obviously get triggered
[14:10:12 CEST] <King_DuckZ> shouldn't that function allocate some buffers at least?
[14:10:47 CEST] <c_14> it should, yeah
[14:16:46 CEST] <c_14> what are format, width and height set to?
[14:18:30 CEST] <King_DuckZ> it's printed by my debug code here https://paste.debian.net/1031273
[14:18:45 CEST] <King_DuckZ> I think 0 is YUVsomething
[14:25:12 CEST] <c_14> what language is that?
[14:30:24 CEST] <c_14> you might want to ask on the user mailing list or wait around here, I'm not really sure what's happening
[14:45:44 CEST] <King_DuckZ> <King_DuckZ> hi, I have this snippet of code https://paste.debian.net/1031274 which is asserting at line 15, the output I get from the commented snipped at line 9 is: https://paste.debian.net/1031273
[14:45:47 CEST] <King_DuckZ> c_14: ^^
[14:46:27 CEST] <King_DuckZ> c_14: the first link is c++, the second is just my program's debug output I copy-pasted from my console
[14:47:14 CEST] <King_DuckZ> I've changed it to also print buf[0] to 7, they're all null as well
[17:46:07 CEST] <dtatulea> Hi. Can ffmpeg show a raw nv12 frame (via ffplay)?
[17:59:17 CEST] <DHE> yes, but you'll have to build the command to tell it what the specs are (width, height, format, etc). and maybe set some super-slow framerate to keep it on-screen. like 0.1 fps maybe
[17:59:47 CEST] <dtatulea> DHE: thanks. And it doesn't need any container or anything?
[18:00:12 CEST] <DHE> you would use the rawvideo "container"
[18:00:32 CEST] <DHE> ffplay -f rawvideo -pixfmt nv12 -s 1024x768 -i FILENAME
[18:00:56 CEST] <DHE> actually no... let me rewrite that
[18:01:13 CEST] <DHE> ffplay -video_size 1024x768 -pixel_format nv12 -framerate 0.1 -i FILENAME
[18:07:29 CEST] <if_gaga9> hello guys, i wish implement this scheme https://i.imgur.com/YFKOaYm.png, it's a possible with ffserver? my question is there can ffserver change the source stream on the fly, according external command, which was sended through zmq?
[18:26:08 CEST] <dtatulea> DHE: thanks!
[19:58:54 CEST] <DHE> I'm using the API to write mpegts streams, but keep getting the error "frame size not set" from the mpegts muxer on startup. only seems to happen on an AAC audio stream. know what the fix for that is?
[20:39:44 CEST] <Hello71> did you set the frame size
[22:47:15 CEST] <temp> Currently I use "ffmpeg -i <file> -f null -" to read time of an audio file. It shows time=00:00:12.16.   How do I increase the precision of the output further? I want time=00:00:12.167
[23:12:30 CEST] <temp> How do i get better time precision when using this command? https://pastebin.com/iTvndR7Q
[23:14:10 CEST] <furq> temp: ffprobe -show_entries format=duration file
[23:14:50 CEST] <temp> furq: I can't use ffprobe. sigh
[23:15:07 CEST] <furq> why not
[23:16:18 CEST] <temp> we had some issues before when using ffprobe so we switched to using ffmpeg. Since we already have this in production it would be much easier to get this info from ffmpeg itself (if possible)
[00:00:00 CEST] --- Sat Jun 30 2018