[Ffmpeg-devel-irc] ffmpeg.log.20150126
burek
burek021 at gmail.com
Tue Jan 27 02:05:01 CET 2015
[00:21] <c_14> framerate is specific to a couple of muxers/demuxers. -r is more general
[00:21] <c_14> And the actual numbers differ
[03:29] <lol123> Can ffmpeg extract video as delimited by title markers?
[03:29] <lol123> Like if I just want to extract the theme song from something
[03:30] <zumba_ad_> thanks c_14
[03:30] <zumba_ad_> i got disconnected, didn't notice
[03:32] <zumba_ad_> need help on this, my sources are full 1920x1080 clip without side or top bars. When I execute this command to merge the 2 mp4 together, the resulting mp4 has huge black bars on side. My video is not squeeze to the middle - http://pastebin.com/k6gy6JDS
[03:39] <lol123> -vf 'select=between(n\,211\,730)'
[03:39] <lol123> That's the correct syntax, yes?
[03:40] <lol123> For discarding everything except the video from frames 211-730?
[03:41] <c_14> If the escaping works, ye.
[03:42] <lol123> Oh no that recorded everything up to frame 211
[03:48] <lol123> The ffmpeg manual doesn't seem to document the select filter very thoroughly
[03:55] <c_14> The expressions are documented in ffmpeg-utils
[08:06] <k_sze[work]> I'm running into a weird seek bug.
[08:07] <k_sze[work]> I try to exact the first three frames of a video, but ffmpeg returns exactly the same frame.
[08:08] <k_sze[work]> This is the command I use: `ffmpeg -ss 0.000 -i my_video.avi -ss 0 -frames 1 -f rawvideo -pix_fmt yuvj444p frame0.yuvj`
[08:08] <k_sze[work]> And then I repeat for -ss 0.034 and -ss 0.067
[08:10] <k_sze[work]> (It's a 30 fps H.264 video)
[08:46] <relaxed> k_sze[work]: did you try demuxing the h264 first t0 see if that makes any difference
[08:46] <relaxed> to, not t0 :)
[08:53] <k_sze[work]> relaxed: you mean demuxing into a raw h264 file?
[09:00] <relaxed> yes
[10:13] <cousin_luigi> Greetings.
[12:38] <hEx_Mv> hey guys, ive been using ffmpeg and tried to upgrade. After installing "deb-multimedia-keyring" my ffmpef throws out the following error: ffmpeg: Symbol `av_pix_fmt_descriptors' has different size in shared object, consider re-linking
[12:38] <hEx_Mv> Illegal instruction
[12:38] <hEx_Mv> ffmpeg: Symbol `av_pix_fmt_descriptors' has different size in shared object, consider re-linking
[12:38] <hEx_Mv> Illegal instruction
[12:38] <hEx_Mv> ffmpeg: Symbol `av_pix_fmt_descriptors' has different size in shared object, consider re-linking
[12:38] <hEx_Mv> Illegal instruction
[12:38] <hEx_Mv> ffmpeg: Symbol `av_pix_fmt_descriptors' has different size in shared object, consider re-linking
[12:38] <hEx_Mv> Illegal instruction
[12:38] <hEx_Mv> ffmpeg: Symbol `av_pix_fmt_descriptors' has different size in shared object, consider re-linking
[12:38] <hEx_Mv> Illegal instruction
[12:38] <hEx_Mv> ffmpeg: Symbol `av_pix_fmt_descriptors' has different size in shared object, consider re-linking
[12:38] <hEx_Mv> Illegal instruction
[12:38] <hEx_Mv> ffmpeg: Symbol `av_pix_fmt_descriptors' has different size in shared object, consider re-linking
[12:38] <hEx_Mv> Illegal instruction
[12:38] <hEx_Mv> uhm well... only once for sure sorry for spam
[12:39] <hEx_Mv> i have a raspberry pi with raspian
[12:39] <hEx_Mv> already removed the dev multimeia keyring and also ffmpeg with apt-get --purge autoremove ffmpeg
[12:39] <hEx_Mv> error still comes up
[12:57] <BtbN> looks like you installed ffmpeg for the wrong architecture.
[13:47] <opmrcrab> hi guys, hopeing some one can help - im using ffmpeg to capture a HLS stream (m3i8+ts files), my segment format ends in -%03d.ts, and always starts *-000.ts, is there an argument im missing to start it at a number of my choosing?
[16:45] <c_14> opmrcrab: -segment_start_number
[19:01] <Guest20637> what does [0:v] and [1:v] mean in filter complex?
[19:02] <Guest20637> Is there any tutorial for this online? Or any place where I can learn about it in depth>
[19:02] <c_14> https://ffmpeg.org/ffmpeg.html#Stream-specifiers-1
[19:02] <c_14> It's the same thing.
[19:02] <c_14> Just looks a bit different.
[19:03] <Guest20637> Basically 0 is the first stream?
[19:04] <c_14> The first file
[19:04] <c_14> [0:v] means all the video streams of the first file
[19:04] <c_14> [0:0] would be the first stream of the first file
[19:04] <c_14> [0:a] is all the audio streams of the first file
[19:04] <c_14> [0:a:0] is the first audio stream of the first file
[19:04] <c_14> etc
[19:11] <Guest20637> If we are talking about an image, do we say [o:i] or [o:v]
[19:12] <c_14> An image as in a picture is a video stream
[19:12] <pzich> also, '0', not 'o'
[19:20] <Guest20637> Okay, That clears it up. Thanks a ton!
[19:34] <Wegg> I would like to use ffmpeg to record my screen as a sequence of images that I can compile together at a later date into a timelapse video. Right now I have this. . . http://www.pasteall.org/56324
[19:34] <Wegg> But the resolution is horrible.
[19:34] <Wegg> Can someone help me figure this out a bit better?
[19:35] <c_14> What do you mean, the "resolution is horrible"
[19:35] <Wegg> quality of the jpeg it outputs.
[19:35] <Wegg> clarity of the image.
[19:35] <c_14> use -q:v 2
[19:35] <c_14> or 1
[19:35] <c_14> Just set it to something low.
[19:37] <Wegg> Unrecognized option 'q:v'
[19:38] <Wegg> I'd much rather use .png but the .png files it produces appear to have their alpha set and the images are transparent.
[19:39] <c_14> Where did you put the -q:v ? before or after the input
[19:39] <c_14> Also, I don't think the png should have any alpha set...
[19:39] <c_14> Never tried it though
[19:40] <Wegg> before the input.
[19:41] <c_14> Has to be after.
[19:41] <c_14> Right before the output filename.
[19:41] <Wegg> -video_size 1920x1080 -framerate 1/10 -f x11grab -i :0.0+0,0 -q:v 2 ~weggingt/images/Screenshots/test%05d.jpg
[19:42] <Wegg> still get the same error
[19:43] <c_14> Works for me. What version are you running?
[19:44] <Wegg> 6:0.816-1
[19:44] <Wegg> 6:0.8.16-1
[19:47] <JEEB> first of all, that is most probably a Libav binary (it probably says "Libav developers" and not "FFmpeg developers" in the copyright), and second of all in Libav the ffmpeg binary is something that didn't get updated ,so you will want to use the avconv binary
[19:47] <JEEB> although I must say that in any case that is an old binary so I recommend you get a newer one, either Libav or FFmpeg.
[19:50] <Wegg> ahh ok avconv works.
[19:51] <JEEB> libav 0.8 is also the last version to contain the ffmpeg binary, so after that it will only have avconv, and FFmpeg only has ffmpeg
[19:52] <Wegg> png output still is totally transparent when I open it in krita
[19:56] <Wegg> avconv -video_size 1920x1080 -framerate 1/10 -f x11grab -i :0.0+0,0 -q:v 2 ~weggingt/images/Screenshots/test%05d.png produces blank images.
[19:57] <howzah> why is -map used in ffmpeg?
[19:57] <llogan> Wegg: avconv is not from the FFmpeg project.
[19:57] <JEEB> also I'm not sure if -q:v does anything with the png encoder :P
[19:58] <JEEB> and yes, it's a Libav binary, and it's old (0.8)
[19:58] <Wegg> so avconv forked from FFmpeg?
[19:58] <JEEB> as I said, I recommend you update your toolset. be it from FFmpeg or Libav
[19:58] <JEEB> Libav is a fork of FFmpeg, yes
[19:58] <JEEB> it's the reason why FFmpeg is better now than it was before the fork
[19:59] <JEEB> howzah, because you need a way to select tracks?
[20:00] <howzah> Can you explain it with an example, please?
[20:00] <llogan> http://ffmpeg.org/ffmpeg.html#Advanced-options
[20:00] <JEEB> ffmpeg -i hurr_durr.nut -map 0:a -c:a shitcodec loers.lara
[20:00] <pzich> or https://trac.ffmpeg.org/wiki/How%20to%20use%20-map%20option
[20:00] <JEEB> selects all audio tracks from the first input file :P
[20:00] <JEEB> and yes, there's plenty of documentation
[20:01] <howzah> OOh. Thank you kind people. :)
[20:01] <llogan> also see http://ffmpeg.org/ffmpeg.html#Stream-selection
[20:03] <Wegg> thank you for your help JEEB
[20:03] <JEEB> np
[21:04] <nodejs> is there a ffmpeg plugin video player for firefox ? anywhere
[21:07] <QKO> hi
[21:08] <QKO> how do I record from x11grab, my audio output and my microphone at the same time?
[21:08] <nodejs> look into ffmpeg desktop command record on google
[21:08] <QKO> I tried adding the microphone and -filter_complex amix=inputs=2 but it gives me [alsa @ 0x7f8d61a78c40] cannot set channel count to 2 (Invalid argument)
[21:11] <llogan> https://trac.ffmpeg.org/wiki/Capture/ALSA
[21:12] <llogan> http://ffmpeg.org/ffmpeg-devices.html#alsa
[21:12] <QKO> I was a fool
[21:12] <QKO> I was trying to use the hardware device rather than the dsnoop device
[21:59] <hungrybird> my system hangs up whenever I run a complex ffmpeg command.
[22:00] <hungrybird> what could be the problem?
[22:01] <hungrybird> I read somewhere it is related to pipes.
[22:01] <hungrybird> But I couldn't find a fix for it.
[22:03] <hungrybird> please
[22:21] <hungrybird> Here
[22:21] <hungrybird> http://pastebin.com/6k4SPx2h
[22:22] <llogan> hungrybird: where is the rest?
[22:23] Action: llogan goes on lunch break
[22:23] <hungrybird> just that.
[22:23] <hungrybird> I am using it through CMD
[22:23] <hungrybird> I am using Windows 8
[22:26] <hungrybird> My system's disk usage touches a 100% when I see in task manager
[22:28] <hungrybird> fflogger, anything?
[22:32] <hungrybird> guys, anything?
[22:58] <itsme_> if I have a file like this http://paste.debian.net/plain/142426 , is there an automagic way to split by chapters?
[23:21] <itsme_> if I have a file like this http://paste.debian.net/plain/142426 , is there an automagic way to split by chapters? < cough, anyone? (cant be done < is also an answer(
[00:00] --- Tue Jan 27 2015
More information about the Ffmpeg-devel-irc
mailing list