[Ffmpeg-devel-irc] ffmpeg.log.20190311

burek burek021 at gmail.com
Tue Mar 12 03:05:02 EET 2019


[00:07:09 CET] <ongo> I'm not sure what I'm doing wrong here, but I'm getting option pixel_format not found: "( r=$(slop) ; ffmpeg -pix_fmt yuv420p -f x11grab -video_size $(cut -d + -f 1 <<<$r) -i :0.0+$(sed 's/.*+\(.*\)+\(.*\)/\1,\2/' <<<$r) tmp_vid.mp4 )"
[00:08:06 CET] <CyberShadow> ongo: Move the "-pix_fmt yuv420p" to before the output file name
[00:09:05 CET] <ongo> CyberShadow: though it is before
[00:09:17 CET] <ongo> tmp_vid.mp4 is the output, which is at the end
[00:11:20 CET] <CyberShadow> Strange, it works for me.
[00:11:20 CET] <ongo> ah u mean right before it :D
[00:11:29 CET] <ongo>     ( r=$(slop) ; ffmpeg -f x11grab -video_size $(cut -d + -f 1 <<<$r) -i :0.0+$(sed 's/.*+\(.*\)+\(.*\)/\1,\2/' <<<$r) -pix_fmt yuv420p tmp_vid.mp4 ) this worked
[00:15:00 CET] <ongo> hmm, this is getting weird: https://0x0.st/zHyq.txt
[00:15:04 CET] <ongo> any ideas CyberShadow?
[00:16:31 CET] <CyberShadow> Oh yeah, that
[00:16:53 CET] <CyberShadow> chroma subsampling needs even coordinates. Needs a bit of arithmetics with the slop coords
[00:17:34 CET] <ongo> 0.0
[00:18:09 CET] <ongo> I'm not too sure what to do here
[00:18:12 CET] <ongo> [x11grab @ 0x55a0e4bf8f00] Stream #0: not enough frames to estimate rate; consider increasing probesize
[00:18:14 CET] <ongo> Input #0, x11grab, from ':0.0+247,217':
[00:18:16 CET] <ongo>   Duration: N/A, start: 1552259680.371950, bitrate: N/A
[00:18:18 CET] <ongo>     Stream #0:0: Video: rawvideo (BGR[0] / 0x524742), bgr0, 1029x812, 29.97 fps, 1000k tbr, 1000k tbn, 1000k tbc
[00:18:20 CET] <ongo> File '/tmp/tmp_vid.mp4' already exists. Overwrite ? [y/N] y
[00:18:22 CET] <ongo> Stream mapping:
[00:18:24 CET] <ongo>   Stream #0:0 -> #0:0 (rawvideo (native) -> h264 (libx264))
[00:18:26 CET] <ongo> Press [q] to stop, [?] for help
[00:18:28 CET] <ongo> [swscaler @ 0x55a0e4c112c0] Warning: data is not aligned! This can lead to a speed loss
[00:18:30 CET] <ongo> [libx264 @ 0x55a0e4c06840] width not divisible by 2 (1029x812)
[00:20:34 CET] <CyberShadow> ( . <(slop -f 'x=%x;y=%y;w=%w;h=%h') ; w=$((w/2*2)) ; h=$((h/2*2)) ; ffmpeg -y -f x11grab -video_size ${w}x${h} -i :0.0+$x,$y -pix_fmt yuv420p out.mp4 )
[00:21:17 CET] <ongo> sweet, thanks!
[00:26:33 CET] <ongo> is it possible to record to a new file?
[00:26:37 CET] <ongo> for instance with mktemp?
[00:29:05 CET] <CyberShadow> Sure, why not. What I like to do is use $(date +%s).mp4 or such for the file name
[00:56:44 CET] <damdai> what is the command to exact subtitle file .mkv file that has subtitle file
[01:21:23 CET] <CyberShadow> damdai: I don't know how to do it with ffmpeg but usually I'd use mkvextract for that
[01:21:47 CET] <damdai> how do i do it with mkvextract
[01:22:55 CET] <CyberShadow> `mkvextract tracks file.mkv`, find the track number, then `mkvextract tracks file.mkv 2:subtitles.ext`, replacing 2 with the track number for mkvextract
[01:24:08 CET] <damdai> i see
[01:24:40 CET] <damdai> ffmpeg cannot accomplish this?
[01:25:21 CET] <damdai> C:\Program Files\MKVToolNix>mkvextract tracks R:\1.mkv
[01:25:21 CET] <damdai> Error: Nothing to do.
[01:25:33 CET] <CyberShadow> damdai: Checking Google finds this: https://superuser.com/questions/583393/how-to-extract-subtitle-from-video-using-ffmpeg
[01:25:49 CET] <damdai> CyberShadow i already tried that
[01:27:05 CET] <CyberShadow> Sorry the first command is not mkvextract, it's `mkvinfo file.mkv`
[01:27:34 CET] <fling> JEEB: How should it be set from the commandline flags point of view?
[01:28:05 CET] <JEEB> fling: you added an AVOption?
[01:28:07 CET] <fling> JEEB: should I just hook the options to input flags like input bitrate?
[01:28:31 CET] <damdai> C:\Program Files\MKVToolNix>mkvextract tracks R:\1.mkv 3:sub.srt
[01:28:31 CET] <damdai> Error: No track with the ID 3 was found in the source file.
[01:28:32 CET] <fling> Or should I use a separate flag for all the avoptions?
[01:28:41 CET] <JEEB> I'm trying to figure out what you mean
[01:28:48 CET] <fling> ok
[01:29:06 CET] <CyberShadow> damdai: You need to use the "track ID for mkvmerge & mkvextract:"
[01:29:15 CET] <fling> Should I make this work? -> ffmpeg -f v4l2 -b:v 300 -i /dev/video0 &
[01:29:23 CET] <JEEB> the command line apps just take AVOptions like any other options and let you set them with the - in the front ("my_option PARAM" -> -my_option param)
[01:29:40 CET] <JEEB> fling: uhh I thought trying to set the bit rate for inputs was some sort of global check :P
[01:29:59 CET] <JEEB> which is why I recommended the AVOption style of having an AVOption called bit_rate or something
[01:30:15 CET] <fling> JEEB: do you have a commandline example?
[01:30:20 CET] <JEEB> ?
[01:30:32 CET] <damdai> CyberShadow  okay it worked using  mkvextract tracks R:\1.mkv 2:R:\sub.srt    eventhough  mkvinfo says it's in track 3
[01:30:36 CET] <JEEB> if you added an AVOption to an input module, you set it before the input
[01:30:51 CET] <JEEB> AVOptions literally become set'able from cli by default
[01:30:58 CET] <JEEB> they should pop up under -h demuxer=v4l2
[01:31:03 CET] <CyberShadow> mkvtoolnix has two kinds of track IDs confusingly
[01:31:03 CET] <fling> ohh
[01:31:26 CET] <fling> so just ffmpeg -f v4l2 -bit_rate 300 -i /dev/video0 & ?
[01:31:31 CET] <JEEB> yup
[01:31:46 CET] <fling> thanks! this is what I wanted to know.
[01:32:02 CET] <JEEB> it will attempt to set the bit_rate AVOption to the avformat context or the avcodec context
[01:32:05 CET] <JEEB> basically
[01:32:24 CET] <fling> Will try to patch things today.
[01:32:25 CET] <JEEB> those that the lavf context ignores, it tries setting to the avcodec side
[01:32:42 CET] <JEEB> (at least if I remember the logic correctly)
[01:33:16 CET] <fling> It does not look proper for me maybe because I like the flags to be unified but it is just a different approach.
[01:33:48 CET] <JEEB> fling: also I'm not sure if the bit rate AVCodec parameter ends up in the input module
[01:34:04 CET] <JEEB> since you're dealing with a protocol/demuxer (whatever avdevice is)
[01:34:38 CET] <JEEB> while setting a bit rate affects the bit rate in the relevant codec context(s)
[01:34:57 CET] <fling> ok makes sense
[02:03:38 CET] <damdai> i have this super weird situation/video:   i have this video file that has a subtitle.  but  mediainfo/ffprobe  doesn't show that it has a subtitle track:   and only VLC can play this video with subtitles
[02:04:17 CET] <damdai> how is this possible
[02:12:56 CET] <Hello71> do you have an external sub file
[02:13:43 CET] <damdai> no
[02:16:59 CET] <damdai> only VLC shows the subtitle track : it shows as "EIA-608 subtitles (c608)
[02:34:01 CET] <Hello71> do you have some ass-balls old ffmpeg
[02:34:17 CET] <Hello71> try git master, there are binaries available
[02:40:40 CET] <damdai> Hello71 huh? i am using latest version
[02:40:55 CET] <Hello71> what is "latest"
[02:41:00 CET] <damdai> 4.1.1
[02:41:02 CET] <Hello71> everybody always says they have the latest version
[02:41:06 CET] <damdai> lol
[02:41:15 CET] <damdai> and they don't? lol
[02:41:22 CET] <Hello71> half the time it turns out they're on fucking debian oldstable or some shit
[02:41:26 CET] <dongs> damdai: EIA608 is line21 VBI subtitles, what freaking video is this
[02:41:40 CET] <damdai> dongs what do you mean?
[02:41:46 CET] <dongs> i mean, its a thing from analog captures
[02:41:47 CET] <Hello71> not all the time but often enough that it's annoying
[02:41:57 CET] <dongs> like, it shouldn't be in a digital video?
[02:42:24 CET] <dongs> oh looks like some apple shit used them
[02:42:30 CET] <dongs> are you stealing stuff from itunes?
[02:43:03 CET] <damdai> https://pastebin.com/qK35k8zu
[02:44:15 CET] <dongs> what exactly does VLC call it
[02:44:24 CET] <dongs> 608 is old analog stuff. 708 is newer digital tv stuff
[02:44:27 CET] <damdai> "EIA-608 subtitles (c608)
[02:44:33 CET] <dongs> i still don't understand how either would survive conversion to mkv tho.
[02:45:46 CET] <damdai> https://i.imgur.com/29RZ845.jpg
[02:45:58 CET] <dongs> EIA-608 and similar closed captioning standards, the captions are embedded directly in the video bitstream as user data. H.264 bitstreams are stored as a sequence of NAL (network abstraction layer) units. Each unit has a type; user data is stored in a NAL unit of the supplemental enhancement information (SEI) type.
[02:46:13 CET] <dongs> apparently thats where they're hiding
[02:46:26 CET] <damdai> huh
[02:46:28 CET] <dongs> https://forum.videohelp.com/threads/390217-Deleting-CEA-608-%28EIA-608%29-captions-from-a-H-264-video-file
[02:46:32 CET] <dongs> this seems to be related.
[02:46:36 CET] <dongs> and even ffmpeg relevant
[02:47:19 CET] <DHE> when transcoding, several h264 encoders have a parameter, -a53cc, which can copy the captions while transcoding.
[02:47:39 CET] <damdai> dongs  doesn't VLC use ffmpeg ?
[02:47:50 CET] <dongs> no
[02:48:00 CET] <damdai> vlc doesn't use ffmpeg? then what does it use
[02:48:53 CET] <dongs> i honestly dont care
[02:49:39 CET] <dongs> i stopped usign vlc after version 2.something when they had some bug that made large .ts files almost unusable to seek in
[02:49:46 CET] <Hello71> vlc uses some forked ffmpeg components
[02:50:14 CET] <damdai> and ffmpeg no longer reports closed caption for the video stream
[02:50:28 CET] <Hello71> at least I think it did at some point in time
[02:50:48 CET] <damdai> dongs  according to that link you just gave me: it says  "and ffmpeg no longer reports closed caption for the video stream:
[02:50:57 CET] <damdai> is this true?
[02:51:01 CET] <dongs> shrugging
[02:51:26 CET] <dongs> you can use that filter_units thing to get to the bits that have teh subtitles in them
[02:52:03 CET] <damdai> and how do i do that
[02:52:07 CET] <DHE> ffprobe reports "Closed Captions" on the video stream for me... this is a reasonably new git version
[02:52:29 CET] <damdai> DHE on what video?
[02:53:54 CET] <DHE> damdai: over the air video captured to multicast
[03:11:00 CET] <damdai> how come AVI is rarely used now? it used to be most popular container
[03:11:15 CET] <dongs> hipsters took over
[03:11:23 CET] <damdai> huh
[03:50:53 CET] <damdai> does youtube ever create upsampled video?   for example  if i upload a 960x540 video,  would it create a 1280x720 video in the dash?
[03:51:46 CET] <furq> no
[03:52:58 CET] <damdai> furq are you sure?
[03:53:30 CET] <dongs> upload, wait a few hours then youtube-dl -F
[03:53:35 CET] <dongs> and see what formats it exposes.
[03:53:41 CET] <dongs> why ask when you can just test yourself
[03:53:53 CET] <damdai> to save bandwidth
[03:55:10 CET] <damdai> dongs i was told that bandwidth is expensive; that's why youtube is willing sacrifice storage space to save bandwidth space
[03:55:28 CET] <dongs> im certain you can use ffmpeg to generate a 960x540 still frame video running for like 30 seconds with dickbutt on it and a sine tone for audio, which would be a few 100kbytes at most, and upload that.
[03:55:52 CET] <damdai> dickbutt ?
[03:56:23 CET] <dongs> you come off to me as a typical LARP'er . you spend all this time coming up with hypothetical what-if questions and then wasting time pondering about them, isntead of actually getting work done
[03:56:53 CET] <dongs> like, do you have a specific use case for 960x540 video on youtube? if you do, it takes just a few minutes to verify your theory
[03:57:01 CET] <dongs> all the other shit you asked about yesterday as well
[03:57:14 CET] <dongs> most of it you could test or verify by looking at source or wahtever
[03:57:41 CET] <damdai> dongs true, i could find the answer my testing
[03:57:47 CET] <dongs> get to it then
[03:57:51 CET] <damdai> by
[03:58:03 CET] <damdai> but bandwidth is expensive
[03:58:38 CET] <damdai> dongs furq did say "no"
[03:59:07 CET] <dongs> if I say "yes", will you actually get off your ass and do the test to find out?
[04:00:26 CET] <damdai> dongs okay how can i create minimum size vidoe files using ffmpeg
[04:00:41 CET] <damdai> so i can test
[04:00:43 CET] <dongs> hint: the less motion there is, the smaller size will be
[04:00:46 CET] <dongs> just make a static image.
[04:00:55 CET] <dongs> i.e. take a png, draw some shit on it in mspaint
[04:01:03 CET] <dongs> then feed it into ffmpeg as a single-frame image source.
[04:01:21 CET] <damdai> i see
[04:01:23 CET] <dongs> specify length + framerate, and either have no audio or use some filter to genrate a sinewave tone and use that
[04:01:41 CET] <damdai> does youtube allow video with no audio?
[04:02:04 CET] <dongs> https://superuser.com/questions/724391/how-to-generate-a-sine-wave-with-ffmpeg
[04:02:07 CET] <dongs> yes it certainly does
[04:02:16 CET] <dongs> i've uploaded several without audio
[04:02:24 CET] <damdai> i see
[04:02:36 CET] <damdai> i know that youtube doesn't allow "audio only" file though
[04:03:21 CET] <damdai> okay i got the .bmp file.  how do i create video with this using ffmpeg
[04:03:46 CET] <dongs> https://stackoverflow.com/questions/25891342/creating-a-video-from-a-single-image-for-a-specific-duration-in-ffmpeg
[04:04:25 CET] <dongs> you just need -loop and -t stuff, and iirc -r to set framerate or else it ends up being 25fps
[04:04:35 CET] <dongs> can ignore the scale/etc stuff
[04:05:26 CET] <dongs> ou want to make a video from a single pic and don't care about any details at all? ffmpeg -loop 1 -i image.png -t 5 out.mp4
[04:09:59 CET] <damdai> well i have to specify the resolution
[04:17:25 CET] <dongs> no you dont
[04:17:29 CET] <dongs> the image size is your resolution
[04:20:07 CET] <damdai> well i want to create 1120x630
[04:20:13 CET] <damdai> and it's not working
[04:22:27 CET] <dongs> http://worksonmymachine.pro
[04:25:06 CET] <damdai> okay i got it working
[04:25:15 CET] <damdai> lol the video file is smaller than the .png file
[04:25:26 CET] <damdai> how is that possible
[04:25:39 CET] <damdai> i made 30 second video
[04:26:49 CET] <another> lossy compression
[04:27:09 CET] <damdai> but it's 30 seconds
[04:27:17 CET] <dongs> its 30 seconds of static
[04:27:35 CET] <damdai> so what if i make 1 hour video of static ?
[04:27:44 CET] <dongs> remember what i said an hour about
[04:27:53 CET] <dongs> about 'asking' instead of "doing"?
[04:34:37 CET] <damdai> it created audio file dash  when the file had no audio in it?
[04:34:50 CET] <damdai> why would google this stupid?
[04:40:14 CET] <cards> player compatability?
[04:41:37 CET] <damdai> it's perfectly normal to have a video with no audio in it
[05:01:41 CET] <cards> must not fit in their model
[05:03:53 CET] <damdai> cards it's easy to make youtube detect if a video file has no audio in it
[05:44:33 CET] <fructose> When cutting video, is there a way to get ffmpeg to seek to the first pts time before -ss while keeping audio synced?
[05:46:47 CET] <Beam_Inn> hi, i'm recording in audacity and want to prevent clipping during the recording process
[05:50:35 CET] <cards> lower your microphone gain
[05:50:54 CET] <Beam_Inn> thanks.
[05:51:19 CET] <cards> or line-in gain
[07:10:04 CET] <boblamont> when I use this script https://paste.ee/p/AI1gn I get "Argument '/path/to/01-file.mp3' provided as input filename, but '/path/to/00-file.mp3 was already specified."
[07:18:44 CET] <CyberShadow> boblamont: Your problem is with the shell syntax, not your usage of ffmpeg. `for file in "$DIR"/*.mp3; do` / `  ffprobe "$file" | grep -vi "stereo" >mono.txt` (mind the quotes' placement).
[07:20:54 CET] <CyberShadow> boblamont: Also, shellcheck would have told you this. https://www.shellcheck.net/
[07:27:36 CET] <boblamont> ok, I've updated it and no longer get errors https://paste.ee/p/BEmFy but I also am not getting the expected results
[07:27:50 CET] <boblamont> mono.txt is blank, but I know there is a mono file in the directory
[08:03:28 CET] <CyberShadow> boblamont: ffprobe sends ALL its output to stderr
[08:05:13 CET] <boblamont> CyberShadow: do I need ">out 2>&1"?
[08:05:40 CET] <CyberShadow> That would be one way to do it probably, or you could 2>&1 | grep ...
[08:07:36 CET] <boblamont> ok, I sort of got it with that
[08:07:57 CET] <boblamont> but I've been foiled by it not counting each entry as a single line
[08:09:26 CET] <CyberShadow> The "correct" approach depends on what you want to do with e.g. files that don't have audio, or that have more than one audio track (out of which none, or some, or all can be mono). Also I guess 5.1 etc. are a thing too.
[08:12:06 CET] <boblamont> I think everything I'd be using this on has a single audio track, so it looks like I want the file names of anything that doesn't have stereo in Stream #0:0
[08:13:51 CET] <CyberShadow> for f in "$dir"/*.mkv ; do if ! ffprobe "$f" 2>&1 | grep -qF stereo ; then echo "$f is mono" ; fi ; done
[08:22:56 CET] <boblamont> CyberShadow: thanks! that did it
[08:54:26 CET] <mikemocha> hello, I'm looking for a static 64 bit linux build of ffmpeg with nvenc.
[08:54:40 CET] <mikemocha> I've seen this website:  https://johnvansickle.com/ffmpeg/git-readme.txt
[08:54:54 CET] <mikemocha> but it doesn't seem to have static builds compiled with nvenc.
[08:55:15 CET] <mikemocha> Thanks for any pointers.
[13:46:20 CET] <ossifrage> I only have PTS, but no DTS, what do I need to set to get avformat to compute the DTS?
[13:46:34 CET] <ossifrage> If I just set dts to AV_NOPTS_VALUE it complains
[13:47:39 CET] <JEEB> set the DTS, even if it's just a +1 on the time base
[13:48:06 CET] <ossifrage> pts = dts works fine for I and IP but not for IBBP
[13:49:01 CET] <JEEB> yes, generally with b-frame streams you need to add some difference between PTS and DTS so that you don't end up with PTS < DTS
[14:28:47 CET] <ossifrage> I looks like I do need to keep a small fifo PTS to generate DTS that will make avformat happy
[14:29:23 CET] <Mavrik>   Lack of DTS will make every single player unhappy too :P
[14:29:25 CET] <ossifrage> But I got something playing from chrome, not live but at least it is working
[14:29:56 CET] <ossifrage> Mavrik, it looks like libavformat use to be able to compute the DTS itself (at least in the past)
[14:30:28 CET] <ossifrage> Mavrik, I just dropped the b pictures to see if was working, poof magic bitrate savings
[14:30:41 CET] <Mavrik> ..
[14:31:01 CET] <Mavrik> Yeah, sounds something that would cause a lot of player authors to come kick you :P
[14:31:40 CET] <ossifrage> Well, my off-on-detector is fixating on this off by one in chrome: https://i.imgur.com/fPegwA8.png
[14:32:05 CET] <ossifrage> They put a transparent box around the ui overlay and they don't quite get it all the way to the bottom
[14:33:38 CET] <iive> what is that encoder that doesn't output the dts of the packet it has just released?
[14:34:05 CET] <ossifrage> iive, yeah the hardware just has a pts in the frame header
[14:35:19 CET] <iive> ossifrage, the thing is, when you feed it image frames with pts, the decoder is supposed to output it in dts order
[14:35:26 CET] <ossifrage> They also can't slew the PTS only jump it which is really annoying for keeping it synced to NTP
[14:39:36 CET] <ossifrage> iive, yeah I know what DTS is for, I just wonder why ffmpeg is deprecating the 'figure it out inside the muxer' part
[14:41:18 CET] <ossifrage> The code is actually there in libavcodec/nvenc.[ch]
[14:56:07 CET] <ossifrage> Is there a way to do what avformat_find_stream_info() does with the data in an AVPacket?
[14:56:36 CET] <ossifrage> I can make sure the the AVPacket contains all the header NALUs
[14:57:59 CET] <Mavrik> Because DTS shouldn't be just made up
[14:58:01 CET] <Mavrik> Neither PTS
[14:58:13 CET] <Mavrik> Since that causes sync issues and bunch of other headaches.
[15:02:14 CET] <ossifrage> It isn't made up, all the data is there, it looks like you need to at some delay to do it right
[15:06:42 CET] <ossifrage> I'm going to try and compute the DTS closer to the hardware, but I really don't want to add extra delay buffers to do it that would suck.
[17:49:41 CET] <belen_> When I try to run `ffmpeg -f concat -i outro.mp4 -i outro.mp4 -c copy test.mp4` I am told `outro.mp4: Invalid data found when processing input`
[17:50:06 CET] <belen_> is there a way to tell exactly what invalid data it's finding? `-loglevel verbose` didn't give me any additional output
[17:54:34 CET] <belen_> http://paste.debian.net/plain/1072700 is the output of `ffprobe -loglevel verbose outro.mp4`
[17:56:45 CET] <furq> belen_: -f concat expects a text file containing the files you want to concat
[17:57:14 CET] <furq> https://www.ffmpeg.org/ffmpeg-formats.html#concat-1
[17:57:38 CET] <belen_> oh. I thought that was a potentially convenient option, not a requirement.
[17:59:40 CET] <belen_> is there a way to force it to use "unsafe" file names like ones containing colons?
[18:00:49 CET] <belen_> -safe 0
[18:01:03 CET] <belen_> it's the absolute vs relative location actually
[18:02:29 CET] <belen_> thanks furq
[18:05:41 CET] <ksk> hola.
[18:06:39 CET] <ksk> I have some video files on server A), that I would like to view a transcoded version of at client B): I can use ffmpeg to convert the files to something else, and then use http to download the transcoded file.
[18:07:05 CET] <ksk> Is there some way I could already download the file to B) as it is transcoding on A)?
[18:07:30 CET] <ksk> Ive been looking at the wiki/StreamingGuide, but I am not sure if that is what I want..
[18:09:12 CET] <ksk> Iam looking for something like "put this ffmpeg string into nginx config and make nginx run ffmpeg and output it" if a client sent a request - Is that possible?
[18:13:38 CET] <ksk> reading the webz a little bit more it seems I will always need an rmpt service to archive that?
[19:56:37 CET] <JKSamir> Hi. I need some help. In the extract_mvs example, I was wondering if I could get the reference frame numbers from which the motion vectors were calculated instead of a +/- 1 direction field as mentioned in the motion_vector.h file. I think it's a TODO. Can anyone give me some pointers on where should I look to get the reference frames numbers not only the directions?
[23:25:45 CET] <AuroraAvenue> testing.
[23:26:42 CET] <AuroraAvenue> How do I splice 2 videos (back to back) that are webm, videos? What is the simple command in Ubuntu?
[23:59:45 CET] <systemd0wn> AuroraAvenue: Have you looked at https://trac.ffmpeg.org/wiki/Concatenate ?
[00:00:00 CET] --- Tue Mar 12 2019


More information about the Ffmpeg-devel-irc mailing list