[Ffmpeg-devel-irc] ffmpeg.log.20190716

[burek]

[00:17:29 CEST] <Thomas_J> No matter where II put "-filter "adelay=2s" at, I get a different error message. https://pastebin.com/fqC6sGfF
[00:19:59 CEST] <durandal_1707> Thomas_J: you need to move it after last " -i "
[00:21:16 CEST] <Thomas_J> I tried it there and it gave me "Filtergraph 'adelay=2s' was defined for video output stream 0:0 but codec copy was selected.
[00:21:16 CEST] <Thomas_J> Filtering and streamcopy cannot be used together."
[00:22:17 CEST] <durandal_1707> -filter:a
[00:22:22 CEST] <durandal_1707> or shorter -af
[00:22:25 CEST] <Thomas_J> OOHHHHHH
[00:29:08 CEST] <Thomas_J> Very interesting. I guess I need to specify 2 channels of audio delay. The right channel is ahead of the left by 2 seconds.
[00:29:35 CEST] <durandal_1707> 2|2
[00:29:56 CEST] <furq> !filter adelay @Thomas_J
[00:29:57 CEST] <nfobot> Thomas_J: http://ffmpeg.org/ffmpeg-filters.html#adelay
[00:30:10 CEST] <furq> almost forgot i had that thing
[00:35:11 CEST] <Thomas_J> Delaying audio has caused a 4 second A/V spread now. It seams that it moved the video instead of the audio.
[00:39:20 CEST] <Thomas_J> Audio was 2 seconds ahead of audio. I used -filter:a "adelay=2s|2s" and now the audio is ahead of the video by 4 seconds.
[00:39:46 CEST] <Thomas_J> That doesn't make sense.
[01:11:48 CEST] <Thomas_J> I can't find any information on why adelay advances the audio instead of delaying it. All I've Googled, just copies what's in the ffmpeg docs and adds no insight.
[01:29:08 CEST] <KittyCat> hmm, the two videos I could provide as a test for my issue are 60+MB each. I tried using ffmpeg to create a small snippet (using -vcodec copy -acodec copy), but the resulting file doesn't cause the issue
[01:36:57 CEST] <Thomas_J> I got it working with -itsoffset but I ahd to increade the input audio buffer to 2048 with "thread_queue_size".
[01:38:53 CEST] <Thomas_J> Forgive my spelling. My fingers are old and they don't type straight anymore.
[04:51:50 CEST] <nine_milli> blb
[05:46:02 CEST] <ossifrage> I think my brain is leaking out my ear, I used to have a clever way to generate cr/cb continuous color patterns and I now I can't remember how I did it...
[05:52:16 CEST] <noahmg123> I'm trying to get this video file into a good format for my friend on Windows. But no matter what codec or other options I use with ffmpeg, it doesn't appear to do anything. I'm using VLC as a baseline to see if it can play the file, and it always gives a codec error talking about h264.
[05:52:41 CEST] <noahmg123> I don't really care about the codec error, I just want it in a form Windows Media Player can play
[05:53:02 CEST] <noahmg123> (but I don't have access to Windows, so I'm using VLC as my sort-of proxy)
[05:54:26 CEST] <noahmg123> Converting to WMV with wmv2 and wmva2 codecs still has VLC complaining.
[06:00:31 CEST] <andrew_46> What is vlc's complaint?
[08:22:07 CEST] <MoziM> is concatonating 2 video files typically a fast operation?
[08:35:23 CEST] <sector_surfer> Hi there
[08:35:46 CEST] <sector_surfer> will ffmpeg ever support sample AES encryption?
[08:36:37 CEST] <JEEB> if someone implements it?
[08:37:14 CEST] <sector_surfer> I'm trying to download a TV show and I get the manifest url
[08:37:17 CEST] <sector_surfer> from here
[08:37:18 CEST] <sector_surfer> https://www.9now.com.au/murder-lies-and-alibis/season-1/episode-4
[08:37:51 CEST] <sector_surfer> do you have any idea when it will be implemented?
[08:38:17 CEST] <JEEB> when someone cares? it's open source, after all
[08:38:19 CEST] <sector_surfer> is there anything else I can do to obtain the video in future?
[08:38:52 CEST] <JEEB> if it's drm you need the license/key handling anyways
[08:40:37 CEST] <JEEB> and I think if you stick yourself to the eme module, that thing will do the decryption for you
[08:41:30 CEST] <sector_surfer> EME Module? I'm sorry, I have no knowledge of programming. what do I need to do to get this eme module?
[08:42:09 CEST] <JEEB> eme is the encrypted media extensions stuff :p
[08:42:21 CEST] <sector_surfer> and where do I get it?
[08:42:25 CEST] <JEEB> your browser of choice probably has at least one :p
[08:42:37 CEST] <JEEB> either google's or ms's
[08:43:15 CEST] <JEEB> i know people rip stuff so in that sense you "just" need to hook into its interfaces :p
[08:43:28 CEST] <JEEB> and then you should get decrypted content
[08:44:11 CEST] <sector_surfer> so once the eme in the browser starts working, how do I then obtain the unencrypted stream? I have no idea what you mean by hooking into the interface, I wouldn't even know where to start
[08:44:17 CEST] <sector_surfer> all this seems a little beyond me
[08:45:02 CEST] <JEEB> someone probably has written stuff for this already. after all, the kodi modules for services utilize it :p
[08:45:15 CEST] <sector_surfer> you're talking to a newby here so I might need to speel things out a bit more
[08:45:35 CEST] <sector_surfer> I mean you might need to spell thing out a bit
[08:46:55 CEST] <JEEB> but yea, even if someone from FFmpeg implements sample aes, you would still need to figure out how to gwt the key since in most cases you're dealing with drm and not like with hls where you have a link to the encryption key in the manifest
[08:47:05 CEST] <JEEB> *get the key
[08:47:42 CEST] <sector_surfer> erm. how to I obtain the key?
[08:48:34 CEST] <JEEB> most likely not at all, or through one of the implementations of the drm system
[08:48:56 CEST] <sector_surfer> it uses a m3u8 manifest. is the key in there?
[08:49:15 CEST] <JEEB> if it's non-drm, maybe?
[08:49:43 CEST] <JEEB> although if it has references to fairplay etc then it's apple drm :p
[08:50:34 CEST] <sector_surfer> Yes, it's fairplay
[08:50:36 CEST] <sector_surfer> ugh
[08:51:03 CEST] <sector_surfer> unfairplay
[08:51:18 CEST] <JEEB> then you probably want to check if there's a dash manifest (usually .mpd) available as well
[08:51:28 CEST] <sector_surfer> OK
[08:51:37 CEST] <sector_surfer> one moment
[08:51:49 CEST] <JEEB> that usually has pr/wv references but at least for those you have the browser module
[08:52:13 CEST] <JEEB> that said, unless someone has coded you a thing using thta you are not getting decrypted stuff
[08:53:07 CEST] <JEEB> so yea, with that I cannot help since I have not looked into it otherwise than since the interfaces are standardized it should be possible
[08:53:42 CEST] <JEEB> and you can find people decrypting them on the interwebs since they end up shared :p
[08:54:27 CEST] <sector_surfer> I've no idea where I should look
[08:54:32 CEST] <JEEB> in other words, sample aes is a pita to implement, and it still doesn't help you with getting the key :p
[08:55:29 CEST] <sector_surfer> hmmm, I've just checked, and there are no references to an .mpd file anywhere, only an .m3u8
[08:56:31 CEST] <JEEB> so the stream would only work on apple devices in that case? which is unlikely.
[08:56:51 CEST] <sector_surfer> https://www.9now.com.au/murder-lies-and-alibis/season-1/episode-4
[08:57:03 CEST] <sector_surfer> will it let you play it?
[09:03:15 CEST] <sector_surfer> Ah. I got the key?
[09:03:22 CEST] <sector_surfer> er.. I mean. I got the key!
[09:03:36 CEST] <sector_surfer> It's using an unsecured http server
[09:04:24 CEST] <JEEB> if it's fairplay that's not the key most likely
[09:05:29 CEST] <JEEB> if you can play it in a browser you should pay attention at what traffic it does, but with drm the key if in manifest is encrypted with another that you get from the fairplay or whatever server if your license is valid :p
[09:07:14 CEST] <sector_surfer> #EXT-X-KEY:METHOD=SAMPLE-AES,URI="skd://assetId=df82c09b-3d62-44fc-94b0-ee9754cc4128
[09:08:26 CEST] <sector_surfer> variantId=fairplay",IV=0xca9cbfbf9b99af3a0962dab4d9282f68,KEYFORMAT="com.apple.streamingkeydelivery",KEYFORMATVERSIONS="1"
[09:15:09 CEST] <sector_surfer> I give up........ for now
[10:20:06 CEST] <MoziM> how to merge flv files with ffmpeg?
[10:20:51 CEST] <MoziM> ffmpeg -f concat -safe 0 -i input.txt -c copy output gives me the error 'Line 1: unknown keyword 'flv''
[10:25:23 CEST] <MoziM> flv
[10:34:28 CEST] <durandal_1707> MoziM: pastebin input.txt
[10:44:19 CEST] <MoziM> durandal_1707: https://termbin.com/3rkoz
[12:12:39 CEST] <MoziM> durandal_1707: any ideas? i can get images to work fine but trying to concat video clips is always failing...
[12:16:33 CEST] <durandal_1707> MoziM: dunno, check that you have valid ascii chars there
[12:17:08 CEST] <durandal_1707> the error given does not make sense
[12:47:08 CEST] <k-man> a while ago i used the libfdk_aac codec to encode some files - but in debian buster, I don't seem to have that encoder available in ffmpeg
[12:47:17 CEST] <k-man> any ideas on how to work around this?
[12:47:23 CEST] <pink_mist> compile ffmpeg yoursefl
[12:47:25 CEST] <pink_mist> *self
[12:48:00 CEST] <k-man> pink_mist: yeah, I have a vague memory of doing that once before
[12:48:03 CEST] <k-man> must have been what I did
[12:49:12 CEST] <pink_mist> some of the things ffmpeg can be built with will not allow debian to actually distribute the binary legally, so they don't build with everything
[12:49:35 CEST] <pink_mist> dunno if that's the case for libfdk_aac or not though
[12:49:53 CEST] <k-man> righto
[12:49:55 CEST] <k-man> thanks pink_mist
[13:43:19 CEST] <Kadigan_KSB> pink_mist, k-man - it is. libfdk-aac has a non-GPL-compatible license (https://trac.ffmpeg.org/wiki/Encode/AAC)
[13:43:47 CEST] <JEEB> yes, so the combined work cannot be distributed
[13:44:25 CEST] <Kadigan_KSB> It's also absolutely fabulous, and usually the reason I go through the pain of setting up a build env. :)
[13:44:55 CEST] <JEEB> thankfully the internal AAC encoder is no longer audible pop bad
[13:44:59 CEST] <JEEB> so I generally just use it
[13:45:22 CEST] <JEEB> I do have some use cases where I need fdk-aac with, and in those cases I build myself on a box
[13:45:30 CEST] <JEEB> or just make it not need any GPL components
[13:45:33 CEST] <JEEB> like libx264
[13:45:49 CEST] <JEEB> since LGPL + fdk-aac license seems to be kind of OK?
[13:45:57 CEST] <Kadigan_KSB> Eh, we've gotten used to it, and I figure - I'm not all that Linux-savvy, might as well take the exercise when it shows up. Also, first tentative steps in cross-compiling for Win-x64 and all that.
[13:46:06 CEST] <JEEB> sure
[13:46:11 CEST] <JEEB> learning to cross-compile is generally worth it
[13:46:22 CEST] <JEEB> I moved to that when FFmpeg's configure script took 10+ min to finish :P
[13:46:25 CEST] <JEEB> (on native windows)
[13:46:46 CEST] <JEEB> thankfully a nice soul then improved that, but at that point I already had my cross-compilation setup done
[13:47:06 CEST] <JEEB> also I do publish some of my binaries so I keep fdk-aac 100% separate from any GPL sysroots
[13:47:21 CEST] <Kadigan_KSB> Yeah, I never got through it, <native> Windows or not... thankfully my Linux (and Linux-like) boxes now outnumber my Windows boxes, so I have plenty of hardware to use for it.
[13:48:00 CEST] <Kadigan_KSB> It's kind of funny. I'm mainly a Windows guy, and that's my primary desktop. And yet... :D
[13:48:04 CEST] <Kadigan_KSB> But anyway.
[13:54:29 CEST] <k-man> i rememeberd i used a build script from github before
[13:54:37 CEST] <k-man> just used that
[13:54:43 CEST] <k-man> had to fix a bug in it first though
[13:55:30 CEST] <k-man> so i'm trying to convert mp3 to m4b files. is there some way, in the ffmpeg command to encode the new file at the same bitrate as the source file?
[13:57:00 CEST] <JEEB> no, and it shouldn't matter
[13:57:15 CEST] <JEEB> having the same bit rate after already once lossily encoding a thing doesn't mean anything
[13:59:56 CEST] <k-man> oh yes, i agree with that. i just want to avoid the m4b files being much larger than the source mp3 files
[16:26:13 CEST] <MoziM> how to remove duplicate frames but keep the time scale?
[16:47:44 CEST] <kepstin> MoziM: https://ffmpeg.org/ffmpeg-filters.html#mpdecimate
[16:55:16 CEST] <nine_milli> blb
[17:01:10 CEST] <N4ppeL> hey, does anyone know about an option to "just use decoding" on ffmpeg CLI? some codecs support output format "rawvideo", which I suppose is exactly that, but somehow not all codecs do (on a potential target environment I have a ffmpeg with only h264 available and it doesnt list rawvideo as supported format) Internally it is however always handled
[17:01:11 CEST] <N4ppeL> with raw data and I was wondering if there is a way to grab that data and send it to a canvas for display
[17:12:21 CEST] <BtbN> So... you want a video player?
[17:14:21 CEST] <N4ppeL> well a bit more than that, but yes, I need the display-ready frames from the encoded mp4 file
[17:16:01 CEST] <BtbN> Well, all decoders output some raw pixel data, in various formats, but most commonly yuv420p or nv12
[17:16:13 CEST] <BtbN> You can do with it whatever you please
[17:16:55 CEST] <N4ppeL> but how can i use just the decoder using ffmpeg CLI?
[17:17:10 CEST] <N4ppeL> ffmpeg always wants an output format and the ones supported with that build are always encoded again
[17:17:25 CEST] <BtbN> Sure, but it's relatively pointless, cause it'll just throw the decoded frames away if you don't send them anywhere
[17:17:55 CEST] <BtbN> What are you actually trying to do? On-Disc raw pixel data isn't usually very useful, due to its sheer size
[17:18:07 CEST] <N4ppeL> well I would send it to a output file / framebuffer
[17:19:18 CEST] <BtbN> Use the rawvideo "encoder"
[17:19:30 CEST] <BtbN> But like I said, raw pixel data for any video is _huge_
[17:19:40 CEST] <BtbN> and likely can't be read from or written to disk in real time
[17:20:11 CEST] <N4ppeL> yeah i know, videos are only 2s 512x512 and there should be a maximum of 6-9 videos displayed simultaneously, so I suppose its managable
[17:20:39 CEST] <BtbN> What are you _actually_ trying to do?
[17:20:48 CEST] <N4ppeL> problem with players is they wont synchronize the videos
[17:21:07 CEST] <N4ppeL> so i need to handle framebuffers manually
[17:21:20 CEST] <BtbN> So your actual problem is you want to play a bunch of videos in a mosaic in sync?
[17:21:39 CEST] <N4ppeL> yeah I guess
[17:21:46 CEST] <BtbN> Cause there's a filter for that
[17:22:07 CEST] <N4ppeL> mh
[17:22:08 CEST] <BtbN> That turns multiple input videos into one output video with the inputs tiled/overlaid/whatever
[17:22:19 CEST] <N4ppeL> well the thing is sources change
[17:22:40 CEST] <BtbN> so?
[17:23:01 CEST] <N4ppeL> I have one huge video 4096x4096 which isnt streamable obviously so the full resolution is only displayed when you zoom in
[17:23:12 CEST] <N4ppeL> i cant just encode one video in advance ;)
[17:23:28 CEST] <BtbN> You'll have to write custom software for that then
[17:23:42 CEST] <N4ppeL> yep
[17:24:09 CEST] <N4ppeL> but as said, the ffmpeg built with only h264 codec cant "encode" rawvideo (doesnt support format rawvideo)
[17:24:13 CEST] <N4ppeL> but then there is no other way ?
[17:24:33 CEST] <BtbN> That's really nothing you can hack together with a bunch of shell scripting with the ffmpeg cli, you need a proper software
[17:25:43 CEST] <N4ppeL> yeah but at somepoint i need decoding functionality from somewhere
[17:26:11 CEST] <BtbN> Use libavcodec directly
[17:27:55 CEST] <N4ppeL> hm yeah. If I want to do that I have to build it myself
[17:30:59 CEST] <BtbN> ?
[17:31:19 CEST] <BtbN> You have to build your own software yourself either way
[17:31:28 CEST] <BtbN> And apparently already have a custom built ffmpeg cli tool
[17:31:42 CEST] <N4ppeL> yeah that was already built
[17:31:47 CEST] <N4ppeL> not by me
[17:32:25 CEST] <N4ppeL> but yeah if there's no option for what I need in what I already have i need to build new stuff, sure :D
[17:32:57 CEST] <N4ppeL> thanks anyway :)
[18:02:13 CEST] <NoImNotNineVolt> so i'm using avconv to try to transcode a s16le audio stream from stdin to an rtp output...
[18:02:53 CEST] <NoImNotNineVolt> e.g. `rtl_fm -f 104300000 -M wbfm -s 200000 -r 48000 - | avconv -f s16le -i pipe:0 -acodec libmp3lame -ab 32k -ac 1 -f rtp rtp://192.168.1.42:1234`, and i'm getting "Value inf for parameter 'time_base' out of range" errors
[18:03:59 CEST] <NoImNotNineVolt> my understanding is that time_base is supposed to be 1/sample_rate, but every time i try to pass that in to avconv explicitly, i get "Codec AVOption time_base () specified for output file #0 (rtp://192.168.1.42:1234) is not an encoding option."
[18:04:54 CEST] <NoImNotNineVolt> (i can `| aplay -r 48000 -f S16_LE` the output of rtl_fm just fine)
[18:05:17 CEST] <NoImNotNineVolt> any ideas?
[18:15:58 CEST] <kepstin> NoImNotNineVolt: this is the ffmpeg channel, we don't support avconv here. If you retry with ffmpeg, please give us the *complete* log output of the command (use a pastebin site)
[18:19:10 CEST] <kepstin> NoImNotNineVolt: i dunno about avconv, but ffmpeg defaults to 1-channel, 44.1kHz sample rate on the input - you need to make sure you specify -sample_rate and -channels as input options.
[18:21:09 CEST] <nine_milli> blb
[19:25:56 CEST] <MoziM> i deleted all duplicate frames in my video using the following ffmpeg -i output.mp4 -vf mpdecimate,setpts=N/FRAME_RATE/TB compressed_output.mp4 and the output video had these stats https://termbin.com/da6g
[19:26:13 CEST] <MoziM> but when i try to play it in a video player i can't skip ahead very well... but it plays fine
[19:27:11 CEST] <MoziM> also when i try to skip to the end nothing is playing :\ i'm guess the actual video's length is smaller than the container it's in? sorry if that's not the correct way of describing it
[19:32:14 CEST] <NoImNotNineVolt> kepstin: -sample_rate, that did it, thanks!
[19:33:09 CEST] <kepstin> MoziM: i'm not sure what you're trying to do with that setpts filter...
[19:33:25 CEST] <kepstin> MoziM: if you want the video to play back at the same speed as originally, just delete that setpts.
[19:47:35 CEST] <NoImNotNineVolt> so, maybe `-f rtp rtp://somehost:1234` isn't what i want... just read https://trac.ffmpeg.org/wiki/StreamingGuide and apparently i should get one more piece.
[20:42:17 CEST] <NoImNotNineVolt> okay, so, i guess that's where i'm lost. the streaming guide specifically recommends nginx-rtmp-module but doesn't suggest what kind of output to feed it.
[20:43:05 CEST] <NoImNotNineVolt> e.g. do i just `-f rtp rtp://localhost:1935/live` my ffmpeg if i've done the basic nginx-rtmp-module config?
[20:47:37 CEST] <kepstin> the nginx rtmp module takes, as you might imagine, rtmp input
[20:49:15 CEST] <furq> NoImNotNineVolt: rtmp is h264/aac in flv
[20:49:22 CEST] <furq> but if you're only streaming audio then why not use icecast
[20:59:30 CEST] <NoImNotNineVolt> because i've heard of nginx but not icecast? :P
[21:00:27 CEST] <NoImNotNineVolt> icecast it is.
[21:09:06 CEST] <JEEB> &34
[21:17:31 CEST] <NoImNotNineVolt> so close! i'm getting audio through to a client running vlc player... but the pitch is all off :P
[21:22:20 CEST] <NoImNotNineVolt> there we go. sample rate / channel settings were off.
[21:22:36 CEST] <NoImNotNineVolt> thanks for the ffmpeg help and icecast recommendation.
[21:32:33 CEST] <whitestone> hello
[21:33:21 CEST] <whitestone> does someone know when using the hls demuxer to not use the last segment of video?
[21:33:31 CEST] <whitestone> i know is witth the option live_start_index
[21:33:57 CEST] <DHE> if it's a live stream, the default is to start close to the end so as to give that "live" feeling
[21:34:02 CEST] <whitestone> and the negative values are from the end
[21:34:38 CEST] <whitestone> yes its a live stream
[21:35:05 CEST] <whitestone> i am downloading a HLS from youtube and i am creating a local repository in HLS too
[21:35:53 CEST] <DHE> if you're writing your own HLS, setting the video type to EVENT (-hls_playlist_type event) is probably what you want. it allows for a live stream but users will be encouraged to start playback from the beginning
[21:36:11 CEST] <whitestone> but, i am a little short of resources
[21:37:15 CEST] <whitestone> yes, the second process is taking the local repository and is putting it in a decklink card
[21:37:48 CEST] <whitestone> but sometimes the second process goes faster than the one is decoding the HLS from youtube
[21:38:04 CEST] <whitestone> so i want to give to the second process a little of delay
[21:39:11 CEST] <whitestone> i didnt find examples about what i want to do in the internet, thats because i am asking here
[21:43:33 CEST] <whitestone> does someone know how to use that option
[21:43:34 CEST] <whitestone> ?
[22:06:36 CEST] <nine_milli> blb
[23:05:01 CEST] <Trieste> Hi, if I do `ffmpeg foo bar.wav` and `ffmpeg foo -f wav - > baz.wav`, the bar.wav and baz.wav files are same size, but differ in a couple of bytes (seemingly just in the header), what gives?
[23:05:31 CEST] <Trieste> (wouldn't care if it didn't mean that the next tool fails on baz.wav but not on bar.wav for some reason)
[23:06:57 CEST] <durandal_1707> both commands are invalid iiuc
[23:07:36 CEST] <howaboutsynergy> hey uhm, does anyone wanna apply this patch? https://trac.ffmpeg.org/attachment/ticket/4749/handlebit_size0.patch  or a more up to date(with git) one: https://github.com/howaboutsynergy/q1q/blob/998b270964ea1ca79ecc2d94e7a43624c66ebe96/OSes/archlinux/home/user/build/1packages/4used/ffmpeg/30/ffmpeg-git/handlebit_size0_v2.patch#L1
[23:08:04 CEST] <howaboutsynergy> it's supposed to handle the case when bit_size == 0 (too)
[23:08:41 CEST] <howaboutsynergy> 'cause you know, that should (also) return AVERROR_INVALIDDATA;
[23:10:44 CEST] <JEEB> howaboutsynergy: that looks like a ticket that nobody went back to 4 years ago...
[23:11:04 CEST] <howaboutsynergy> the patch tho? :)
[23:11:16 CEST] <howaboutsynergy> doesn't it make sense?
[23:12:02 CEST] <JEEB> do you have a case which makes it bork. that makes checking things much simpler
[23:12:24 CEST] <howaboutsynergy> I don't
[23:12:43 CEST] <furq> Trieste: looks like the wav muxer writes some values to the header at the end
[23:12:51 CEST] <furq> which it obviously can't do if the output isn't seekable
[23:12:52 CEST] <JEEB> also the patch by itself isn't really self-documenting and at least the github patch has no author/commit message info
[23:13:51 CEST] <howaboutsynergy> JEEB, ah you mean it should've been made via `git format-patch` to be considered?
[23:13:56 CEST] <furq> https://github.com/FFmpeg/FFmpeg/blob/master/libavformat/wavenc.c#L408
[23:14:23 CEST] <Trieste> furq, oh right
[23:17:12 CEST] <JEEB> howaboutsynergy: or otherwise having that information
[23:17:24 CEST] <JEEB> howaboutsynergy: also fyi the development *of* FFmpeg channel is #ffmpeg-devel
[23:17:34 CEST] <howaboutsynergy> oh
[23:17:41 CEST] <JEEB> this is the user facing channel and a lot of developers aren't here, unfortunately
[23:17:48 CEST] <howaboutsynergy> oh right, it's in the title too
[23:17:58 CEST] <howaboutsynergy> thanks!
[00:00:00 CEST] --- Wed Jul 17 2019