[Ffmpeg-devel-irc] ffmpeg.log.20150312

burek burek021 at gmail.com
Fri Mar 13 02:05:01 CET 2015


[04:51:07 CET] <pinPoint> has anyone messed around with prores yet?
[05:53:32 CET] <jerware> moin.
[05:53:48 CET] <jerware> If I stream my webcam to an other computer, will there be a delay like with vlc?
[05:54:05 CET] <jerware> I'm not sure if the delay is vlc's fault or if that's how streaming works.
[05:54:35 CET] <pzich> there will always be some amount of delay, with camera <- encode -> network <- decode -> display
[05:58:13 CET] <jerware> does any of you folks have anyoneliners to stream so I can compare vlc with ffmpeg?
[06:01:58 CET] <sam__> Is anyone around?
[06:05:42 CET] <pinPoint> some
[06:24:21 CET] <sam__> Pinpoint you still there?
[06:25:14 CET] <pinPoint> yes but not really an ffmpeg prodigy....
[06:51:31 CET] <sam__> I hear ya
[06:51:38 CET] <sam__> I figured out my issue though
[06:53:19 CET] <sam__> if anyone ever comes in asking to convert an unkown series of images that are not stored on the files system to h264: -f image2pipe is a lifesaver
[06:54:20 CET] <sam__> So long as they have access to ffmpeg's stdin,  cat streams/videos.png | ffmpeg -loop 1 -f image2pipe -vcodec png -i - -c:v libx264 -crf 28 -r 25 -pix_fmt yuv420p ./out.mp4
[06:54:35 CET] <sam__> I'm sure thats a known one, but it took me a while to figure it out :)
[06:54:38 CET] <sam__> take care
[09:57:01 CET] <ffmpeguserss> Hello
[09:57:15 CET] <ffmpeguserss> I get error when encoding
[09:57:37 CET] <ffmpeguserss> Error message is [aac @ 0601e020] decode_band_types: Input buffer exhausted before END element found. Error while decoding stream #0:1: Invalid data found when processing input
[09:57:52 CET] <ffmpeguserss> What's wrong?
[10:02:42 CET] <ffmpeguserss> Log is here - http://pastebin.com/6eqf6f0A
[10:03:16 CET] <ffmpeguserss> Error message reapeat but encoding continue without stop
[10:03:33 CET] <ffmpeguserss> Actually no problem on video
[10:04:18 CET] <ffmpeguserss> I don't understand why appear this error message
[10:07:39 CET] <relaxed> can you test again with the latest git build from http://ffmpeg.zeranoe.com/builds/ ?
[10:08:50 CET] <relaxed> remove -preset fast
[10:09:55 CET] <ffmpeguserss> It's same
[10:10:01 CET] <ffmpeguserss> removed preset
[10:11:25 CET] <relaxed> what happens when you play the input?
[10:11:43 CET] <ffmpeguserss> no problem
[10:11:53 CET] <ffmpeguserss> and output also no problem
[10:12:02 CET] <ffmpeguserss> just error message appeared
[10:13:44 CET] <ffmpeguserss> That error message is serious or negligible?
[10:14:32 CET] <ffmpeguserss> If negligible, I want remove that error message from my ffmpeg build source
[10:15:29 CET] <ffmpeguserss> what do you think about it?
[11:06:54 CET] <termos> I'm trying to use the overlay filter through C++ API, but I can't figure out how to get my overlay image as input to the filter graph. From the command line it's specified by -i arguments
[11:24:47 CET] <Andy2> Hi guys, any idea where the mp filter did go? http://ffmpeg.org/ffmpeg.html#mp
[11:25:05 CET] <Andy2> I was using it before but suddenly it is gone
[11:31:45 CET] <relaxed> Andy2: it was removed. which filter were you really using?
[11:34:09 CET] <relaxed> mp was a proxy to use mplayer filters, but since they were all ported to ffmpeg as native filters it was removed.
[11:42:35 CET] <Andy2> relaxed: ah ok i was using: mp=eq=0:0 to bypass a bug when converting color spaces
[11:43:57 CET] <Andy2> relaxed: here is the bug i am refering to: https://trac.ffmpeg.org/ticket/2267
[12:10:03 CET] <Zevv> Hi; is there a way to access RTP stream statistics from a AVFormatContext? Bitrate, packet loss, etc?
[12:29:57 CET] <Andy2> relaxed: ok you were right, compiled 2.5.4 and it's working again
[12:30:29 CET] <Andy2> Any idea how to have the same effect as mp=eq=0:0 with the native filters?
[12:44:59 CET] <relaxed> Andy2: it would be -vf eq=0:0
[12:48:33 CET] <relaxed> You can see a list of all supported filters with "ffmpeg -filters", and to see a filter's option run "ffmpeg -h filter=eq"
[12:50:26 CET] <Andy2> ok
[12:50:28 CET] <Andy2> thx
[12:50:49 CET] <Andy2> i tried it but the filter seems to be different, at least it gives me a completely different result
[12:53:29 CET] <relaxed> Andy2: can you upload a sample of the input somewhere?
[13:00:43 CET] <Andy2> relaxed: what exactly? I mean, the bug i am getting happens when the color correcting kicks in, and the mplayer filter fixed that.
[13:01:01 CET] <Andy2> You are interested in the eq fix or the original bug?
[13:02:39 CET] <relaxed> I want a sample of the input video so I can see the issue first hand.
[13:05:52 CET] <Andy2> https://trac.ffmpeg.org/ticket/2267
[13:05:55 CET] <Andy2> here for example
[13:06:05 CET] <Andy2> get the file  example_input.MOV
[13:07:36 CET] <Andy2> to reproduce the bug you have to use a filter that transforms the color space
[13:07:43 CET] <Andy2> i described it here: http://ffmpeg.org/pipermail/ffmpeg-user/2014-August/023110.html
[13:08:01 CET] <Andy2> i got some feedback there as well but then the involved people stopped replying
[13:09:17 CET] <Andy2> and i was able to avoid the bug when using: mp=eq=0:0 in addition to my delogo call
[13:09:58 CET] <Andy2> but on ffmpeg 2.6 eq=0:0 seems to have a different effect
[13:10:09 CET] <Andy2> (hope that helps to reproduce it) :)
[13:11:48 CET] <relaxed> I don't see any difference. Look at
[13:11:55 CET] <relaxed> http://johnvansickle.com/misc/
[13:12:43 CET] <relaxed> input.png is from example_input.MOV and output.png is from mpeg4 encoding from example_input.MOV
[13:12:59 CET] <relaxed> which does yuvj -> yuv conversion
[13:13:39 CET] <Andy2> there is a difference
[13:13:50 CET] <Andy2> try compare  -metric PSNR input.png output.png test.png
[13:14:08 CET] <Andy2> but it's not that visible in your examples i agree
[13:14:24 CET] <Andy2> there are other imagas here it is much more visible
[13:14:40 CET] <relaxed> you're going to lose some quality through the conversion
[13:15:15 CET] <Andy2> well, try to grep a frame with and without the delogo filter
[13:15:28 CET] <Andy2> as described in my mailing list mail
[13:16:05 CET] <Andy2> let me download my original video sample
[13:16:10 CET] <Andy2> i have to recreate it
[13:16:19 CET] <Andy2> so it takes some time cause i don't have the video anymore
[13:16:25 CET] <Andy2> it was much much more visible there
[13:16:41 CET] <relaxed> you can avoid the conversion by using h264
[13:16:47 CET] <relaxed> is that not an option?
[13:16:49 CET] <Andy2> sadly the stupid file hoster deleted my sample
[13:17:30 CET] <Andy2> no, cause i generate images from it
[13:18:03 CET] <relaxed> what are you doing?
[13:18:18 CET] <Andy2> well, it's a big process all in all
[13:18:33 CET] <Andy2> i am writing a webapp to generate gifs from clips
[13:19:28 CET] <Andy2> works pretty good but the delogo call and now the vidstab call distorts the result
[13:23:21 CET] <relaxed> and those are seperate issues?
[13:23:59 CET] <Andy2> the contrest loss while color space conversion is one issue
[13:24:17 CET] <Andy2> and the eq filtre not working anymore as before is, i don't know
[13:24:22 CET] <Andy2> maybe not an issue
[13:24:38 CET] <Andy2> but i cannot use it anymore to fix the color space problem
[13:24:41 CET] <Andy2> (as it seems)
[13:25:12 CET] <relaxed> if the eq filter doesn't work like mp=eq then you should file a bug report
[13:26:08 CET] <Andy2> yeah it does, teh call sets the contrast to 0 i guess
[13:26:18 CET] <Andy2> but it was a workaround anyway
[13:26:49 CET] <Andy2> do you still need a video to see it yourself?
[13:29:07 CET] <relaxed> I used example_input.MOV
[13:29:34 CET] <Andy2> yeah it's reproducable with that one as well, just not as visible
[13:29:49 CET] <Andy2> but the compare command shows it
[13:31:05 CET] <relaxed> look at "ffmpeg -h filter=eq" and play with the settings
[13:31:34 CET] <Andy2> yeah i am doing that currently
[13:31:43 CET] <relaxed> you need to accept the fact the color conversion incures quality loss
[13:32:48 CET] <Andy2> relaxed: yes sure but it not just "loss" it ruins teh image
[13:33:00 CET] <Andy2> teh contrast is completely different
[13:33:03 CET] <Andy2> much brighter
[13:33:50 CET] <relaxed> then I suggest you upload a sample to the mentioned bug report because I can't tell a difference with example_input.MOV
[13:34:48 CET] <relaxed> Andy2: did you try the colormatrix filter?
[13:34:57 CET] <Andy2> no
[13:39:34 CET] <Andy2> relaxed: http://ge.tt/api/1/files/37Ibw5C2/0/blob?download
[13:39:38 CET] <Andy2> i uploaded a file
[13:39:49 CET] <Andy2> you can better see it there
[13:39:54 CET] <Andy2> the dark colors are brighter
[13:42:31 CET] <relaxed> is that the output?
[13:42:43 CET] <Andy2> no it's the test video
[13:42:59 CET] <relaxed> look at http://johnvansickle.com/misc/test.avi
[13:43:13 CET] <relaxed> again, I can't see a difference
[13:44:18 CET] <Andy2> i can see it
[13:44:23 CET] <Andy2> your video is much brighter
[13:44:28 CET] <Andy2> well, maybe not much
[13:44:30 CET] <Andy2> but it is :)
[13:46:21 CET] <Andy2> relaxed: http://i.imgur.com/BHGX1SK.png
[13:46:27 CET] <Andy2> can you see it now?
[13:47:16 CET] <Andy2> your monitor could be the reason you cannot see it as well
[13:47:21 CET] <relaxed> the mp4 looks brighter
[13:47:51 CET] <Andy2> tmaybe i am not using the correct terms
[13:48:24 CET] <Andy2> to me the right video looks like it has a grey fog
[13:48:32 CET] <relaxed> I see what you're saying
[13:48:51 CET] <Andy2> great :)
[13:49:19 CET] <Andy2> the bug report is not from me so i am not the only one noticing it
[13:49:50 CET] <Andy2> maybe no one cared cause there was a fix
[13:57:52 CET] <Andy2> btw, thx for the help so far, really appreciate it.
[14:01:47 CET] <relaxed> Andy2: Does this look better? ffmpeg -i test.mp4 -vf scale=w=iw:h=ih:sws_flags=full_chroma_inp -c:v libx264 -pix_fmt yuv420p out.mkv
[14:03:16 CET] <Andy2> no
[14:03:33 CET] <Andy2> looks the same to me
[14:03:42 CET] <Andy2> same grey fog
[14:05:24 CET] <Andy2> http://i.imgur.com/AjGsLeC.jpg
[14:11:21 CET] <Jonas__> after much work, the most robust solution I have arrived to, to use ffmpeg to stream to twitch, appears to be ffmeg grabbing mic, audio loopback, screen footage and webcam footage, and encoding that to a file
[14:11:53 CET] <Jonas__> and then a second ffmpeg grabbing footage straight from that file and streaming it using copy-codecs to the twitch rtmp
[14:12:14 CET] <Jonas__> can't seem to find another solution that doesn't cause undue video lag
[14:12:19 CET] <Jonas__> or audio lag for that matter
[14:12:48 CET] <Jonas__> I assume because of output blocking, since the above solution works well
[14:14:03 CET] <Jonas__> ie to file, to udp works well, to rtmp directly or to a pipe streamed to another ffmpeg process responsible for the rtmp does not
[14:16:52 CET] <Andy2> Jonas__: have you tried simplescreenrecorder?
[14:17:51 CET] <Andy2> relaxed: i made two gifs: http://i.imgur.com/uvxXZme.gif http://i.imgur.com/ymUW36s.gif
[14:18:07 CET] <Andy2> if you can tell me which one is which i am not insane ;)
[14:19:30 CET] <Jonas__> Andy2, I don't know, is simplescreenrecorder able to merge multiple audio and video streams and apply arbitrary frei0r and other filters?
[14:20:01 CET] <Jonas__> because if it's more flexible and agile than ffmpeg, it might be worth a shot, but it isn't really
[14:20:14 CET] <relaxed> Andy2: -vf scale=w=iw:h=ih:sws_flags=full_chroma_int:sws_dither=a_dither looks a little better
[14:20:58 CET] <Andy2> Jonas__: i don't think so
[14:22:09 CET] <Andy2> relaxed: indeed it dows look better
[14:22:27 CET] <Andy2> a lot better to be honest
[14:22:54 CET] <relaxed> Andy2: look at "man ffmpeg-scaler"
[14:23:28 CET] <Andy2> ah well
[14:23:42 CET] <Andy2> did you forget to do the yuv420p conversion?
[14:24:10 CET] <relaxed> no, -pix_fmt yuv420p was in my command
[14:25:05 CET] <Andy2> ah not in the one i tried
[14:25:26 CET] <relaxed> gifs aren't yuv420p so you're doing another conversion there
[14:25:42 CET] <Andy2> yes but it shows teh difference
[14:26:18 CET] <Andy2> ah, now i tried you last commend, looks as bad as before now :/
[14:26:24 CET] <Andy2> sry for the confusion
[15:29:33 CET] <greenit> hi, i have a question: everytime i create a slideshow using a program which uses ffmpeg, the images are blurry and at every beat of the background-music, it gets worse for a short time. do i do something wrong? with windows movie maker (or how it is called), there is no such problem, the slideshow-video looks like the original images....
[16:14:42 CET] <prajj> hi
[16:15:09 CET] <prajj> say i wanted to record my screen on a mac using ffmpeg's avfoundation devicee
[16:15:27 CET] <prajj> i want to do it using code written in C
[16:16:03 CET] <prajj> I am currently stuck on dealing with the UYVY422 format used by the captured screen stream on OS X
[16:19:14 CET] <arufl> hi
[16:19:40 CET] <arufl> Could i build rtmpdump without libssl depencende?
[16:19:50 CET] <iive> prajj: libswscale is the library for converting colorspaces.
[16:20:52 CET] <iive> and color formats.
[16:36:01 CET] <prajj> thanks ill check it out
[16:49:49 CET] <jonascj> Hi all. If I do "ffmpeg -i input.mkv -vcodec copy -acodec copy output.mp4" will it leave the video and audio untouched, but change container? Does it recognize the "mp4" from the output file name?
[16:50:24 CET] <c_14> yes, yes
[16:50:45 CET] <jonascj> so I cannot make mp4 output with the filename "whatever.mp4.someting"?
[16:50:56 CET] <c_14> You can, but you need to use -f mp4
[16:51:13 CET] <jonascj> c_14: ah, that was what I was thinking about, specifying the container explicit.
[16:51:26 CET] <c_14> ffmpeg guesses the output format from the filename where possible, but you can always override it with -f
[16:51:34 CET] <klaxa|work> for example: ffmpeg -i input.mkv -c copy -f mp4 something.ext
[16:51:48 CET] <jonascj> if I want to scale the video at the same time shall I still use "-vcodec copy"? or is that in conflict with "-vf scale=640:-1"?
[16:51:59 CET] <klaxa|work> that will conflict
[16:51:59 CET] <c_14> That's in conflict
[16:53:01 CET] <jonascj> so I just drop the "-vcodec copy"? What happens to the video format then? is it just reencoded as h264 if that is the current codec?
[16:53:11 CET] <c_14> Depends on the output format.
[16:53:30 CET] <c_14> If your version is linked against libx264, mp4 defaults to h.264 output at crf 23 iirc
[16:54:07 CET] <jonascj> "ffmpeg -i someh264.mkv -vf scale=640:-1 -acodec copy -f mp4 output.mp4"
[16:54:33 CET] <c_14> use -2 instead of -1
[16:54:35 CET] <c_14> just in case
[16:54:44 CET] <jonascj> why -2?
[16:55:21 CET] <c_14> The default pixel format is yuv420p, yuv420p requires a height and width divisible by 2, -2 does the same thing as -1 except that it makes sure it's divisible by 2
[16:55:56 CET] <jonascj> while I have the assistance of you excelent fellows: will subtitles streams be dropped from the mkv container if I do the command I posted just before?
[16:56:10 CET] <jonascj> I want them to be dropped, I just need to figure out if I need to drop the explicitly
[16:56:11 CET] <klaxa|work> oh -2 is new is it not?
[16:56:17 CET] <c_14> ffmpeg by default maps 1 audio, 1 video, and 1 subtitle stream
[16:56:25 CET] <c_14> klaxa|work: not _that_ new, sometime last year iirc
[16:56:31 CET] <c_14> jonascj: use -sn
[16:56:40 CET] <jonascj> c_14: I want to drop the subtitles :)
[16:59:19 CET] <c_14> -sn
[17:00:23 CET] <jonascj> when I do this ffmpeg conversion from mkv h264 to mp4 with -acodec copy and -vf scale=640:266 I get quite few warnings/errors "[matroska @ 0xsomething] ... x<y ...invalid clipping". Do you guys know what that means?
[17:04:01 CET] <c_14> PTS/DTS ?
[17:04:08 CET] <jonascj> c_14: yes
[17:04:20 CET] <jonascj> like "[matroska @ 0x312dec0] st:0 PTS: 329355 DTS: 329355 < 329365 invalid, clipping"!
[17:04:26 CET] <jonascj> without !
[17:05:01 CET] <c_14> That means that the time at which the frame was supposed to be displayed is before the time at which it's supposed to be decoded.
[17:06:28 CET] <jonascj> Is that a problem? I do not know the inner workings of h264 and/or mkv...
[17:09:12 CET] <c_14> shouldn't be that big an issue
[17:11:43 CET] <jonascj> c_14: I could write log messages like that if some recoverable situation occured in a library of mine, so maybe this is just a message saying "hey, something could be more optimal, but we recovered"
[17:12:27 CET] <c_14> As long as it's only a warning and not an error, it isn't fatal.
[17:13:21 CET] <jonascj> the processing does not stop, and ffmpeg exits clean at the end, and the output file seems to be working
[17:16:05 CET] <jonascj> thank you for your input and assistance!
[17:18:39 CET] <c_14> np
[17:40:01 CET] <JackWinter> just to verify my failure to find it in the man page.  is it possible to use ffmpeg to cut of a specific amount of samples in the beginning of a wave file?  messed up the compensation (for usb latency) when i recorded, and it's quite a few wave files to do.
[17:44:47 CET] <c_14> You can calculate how many seconds the number of samples is, and use -ss
[17:45:06 CET] <c_14> Or probably fractions of a second
[19:34:59 CET] <sagax> hi!
[19:35:52 CET] <sagax> how to push and pull stream audio with "PUSH: ffserver+ffmpeg" "PULL: ffplay or other player like as mplayer"
[19:35:54 CET] <sagax> ?
[19:36:21 CET] <sagax> hm
[19:36:36 CET] <sagax> push and pull base on UDP packet
[19:36:40 CET] <sagax> not tcp
[19:44:53 CET] <blazzer> I am trying to extract frames from a video file, so far so good : ffmpeg -ss 145 -i 1.mkv -ss 5 -t 10 -r 25 -f image2 a%03d.png
[19:45:17 CET] <blazzer> But I want subtitles hardcoded onto the picture generated. how to do that?
[19:46:31 CET] <c_14> https://trac.ffmpeg.org/wiki/HowToBurnSubtitlesIntoVideo
[19:47:15 CET] <santiago__> I used mencoder for this: mencoder tokyo_story.mp4 -sub tokyo_story.mp4.srt -o tokyo_story_hard.mp4 -oac pcm -ovc lavc -lavcopts vbitrate=1200
[19:47:27 CET] <santiago__> and it worked great
[19:47:39 CET] <blazzer> Tried that. It only prints the subtitles from beginning, where as the clip is extracted from the middle!
[19:47:50 CET] <blazzer> c_14: ^
[19:48:23 CET] <c_14> You have to move the -ss behind the -i 1.mkv
[19:48:38 CET] <blazzer> OK. will try
[19:48:41 CET] <santiago__> mmmh...it sounds to me you have edit the srt file and to substract the time you cut at the beginning....could work
[19:48:55 CET] <santiago__> yeag, c_14 advice is of course more logical
[19:49:56 CET] <blazzer> c_14b advice worked
[19:49:58 CET] <blazzer> !!!
[19:50:06 CET] <santiago__> cool!
[19:50:17 CET] <blazzer> B-)
[19:50:32 CET] <blazzer> but it now takes a longer time...
[19:51:48 CET] <blazzer> Anyway thank you to both of you. It will do for me!
[20:19:13 CET] <kaniu> so how is config.mak generated?
[20:19:52 CET] <kaniu> I'd like to set HAVE_PEEKNAMEDPIPE to 0 but I'm not sure where the origin of that variable is
[20:27:42 CET] <BtbN> The configure script sets it if it detected support for that.
[20:34:07 CET] <kaniu> I don't see HAVE_PEEKNAMEDPIPE anywhere in configure or in config.log, though
[20:39:54 CET] <c_14> line 4814 in configure
[21:00:10 CET] <arufl> Hi
[21:00:30 CET] <arufl> I cannot compile rtmpdump with polar ssl option
[21:01:22 CET] <arufl> http://paste.ubuntu.com/10587284/
[21:01:40 CET] <arufl> CRYPTO=POLARSSL
[21:23:40 CET] <Pit> my brain is about to explode
[21:23:46 CET] <Pit> [flv @ 0x21c9320] Failed to update header with correct duration. [flv @ 0x21c9320] Failed to update header with correct filesize.
[21:23:57 CET] <Pit> so many people complaining about this, but no answer
[21:25:04 CET] <Pit> does anybody knows what the hells wrong? i just wanted to stream a simple .flv video, and got this error, i've tried changing cmd's options but didnt work+
[21:30:53 CET] <sagax> howto minimize delay when use ffserver like as online-radio?
[21:31:20 CET] <sagax> i don't find the answer for this question
[22:54:53 CET] <hadees> So this sounds crazy but I swear its the only option, basically i need to create a super tiny video that runs for a very long time.  No one actually has to look at it, im just using to keep an html5 app open on an amazon fire tv
[22:55:15 CET] <hadees> I used this ffmpeg -framerate 1 -i 1x1.png -c:v libx264 -r 1 -pix_fmt yuv420p out.mp4 but the video is only a second, can I increase the length of the video with out increasing the size by too much?
[23:08:54 CET] <arufl> any idea?
[23:10:15 CET] <c_14> use the color filter as a source
[23:10:39 CET] <c_14> arufl: Why are you asking here and not somewhere geared to supporting rtmpdump.
[23:10:50 CET] <c_14> hadees: the color filter was aimed at you
[23:11:55 CET] <hadees> c_14: whats the color filter?  I ended up doing this ffmpeg -y -loop 1 -t 01:00:00 -i 1x1.png -r 1  -vcodec libx264 -f mp4 out.mp4 but im not sure if thats the optium way to do it, size and length are the biggest issues
[23:12:11 CET] <hadees> i need is to run as long as possible and be as small as possible
[23:12:18 CET] <hadees> in file size and image size
[23:13:06 CET] <c_14> https://ffmpeg.org/ffmpeg-filters.html#color_002c-haldclutsrc_002c-nullsrc_002c-rgbtestsrc_002c-smptebars_002c-smptehdbars_002c-testsrc
[23:13:26 CET] <c_14> with a size of 1x1 and without a duration it'll run forever
[23:13:51 CET] <hadees> c_14: so just replace my input source with that? is that the best way to do it?
[23:13:59 CET] <hadees> c_14: really? vlc didnt run it forever when i left that out
[23:14:17 CET] <hadees> this is going in a video tag though
[23:17:21 CET] <c_14> `ffmpeg -f lavfi -i color=c=black:s=2x2:d=6220800 out.mp4'
[23:17:25 CET] <c_14> for example
[23:17:31 CET] <hadees> thanks!
[23:17:40 CET] <c_14> That one will produce 72 hours of video
[23:17:46 CET] <arufl> I cannot compile rtmpdump with polar ssl option
[23:18:04 CET] <c_14> arufl: Why are you asking here and not somewhere geared to supporting rtmpdump.
[23:18:27 CET] <arufl> WHERE SHOULD I GO+
[23:18:30 CET] <hadees> c_14:  will that ever stop if I run it on my local machine? or is it going to take 72 hours?
[23:18:32 CET] <arufl> sorry caps
[23:18:54 CET] <hadees> c_14: it seems to be running for a while
[23:19:08 CET] <c_14> hadees: it will stop, on my system I'm encoding at 5 minutes of video /second
[23:19:24 CET] <c_14> arufl: According to https://rtmpdump.mplayerhq.hu/ there's a mailing list and discussion forums
[23:23:16 CET] <c_14> hadees: oh, eh. Just going to throw this out there, but that duration will actually give you 72 days of video and not 72 hours.
[23:23:33 CET] <c_14> Might want to adjust the duration
[23:24:24 CET] <hadees> c_14: yeah, so i ended up doing 3600 which is an hour but the file is actually bigger then what I did
[23:25:09 CET] <c_14> hadees: because the framerate is different, add -r 1. or actually even -r 1/3600
[23:27:27 CET] <hadees> -r 1 worked -r 1/3600 got some weird results, but i think that is enough
[23:27:28 CET] <hadees> thanks!
[23:34:36 CET] <MadTBone> is there a way to tell ffmpeg that some (raw) data is interlaced and in a particular field order?  I know about the fieldorder filter, but this relies on the interlaced flag already being set.
[00:00:00 CET] --- Fri Mar 13 2015


More information about the Ffmpeg-devel-irc mailing list