[Ffmpeg-devel-irc] ffmpeg.log.20180706

[burek]

[01:27:09 CEST] <Trel> To set metadata for the first subtitle, is it -metadata:s:0 or -metadata:s:s:0 ?  I'm having a hard time reading the docs on that.
[01:31:22 CEST] <Trel> Aside from the fact I can't get it to work
[01:34:47 CEST] <CoreX> what type of metadata you trying to set
[01:36:18 CEST] <Trel> Title and language.
[01:39:19 CEST] <Trel> Nevermind, my fault, container didn't support subtitles, and video player was picking up the srt file.
[01:44:03 CEST] <lvlscape> jkqxz: thanks for the tip.
[06:00:47 CEST] <Hackerpcs> ffmpeg -loop 1 -i image.png -i audio.mp3 -acodec aac -vcodec libx264 -t $(ffprobe -v error -show_entries format=duration -of default=noprint_wrappers=1:nokey=1 audio.mp3) -tune stillimage -pix_fmt yuv420p final.mp4
[06:01:04 CEST] <Hackerpcs> why does it fail with "Invalid duration specification for t: 17.867750"?
[10:06:43 CEST] <Hackerpcs> above the problem was that it had a carriage return on the end, storing it on a variable and then removing it and similar with   duration="${duration//[$'\t\r\n ']}"
[11:52:06 CEST] <keglevich> ffmpeg -re -i 1.mkv -c:v mpeg2video -threads 0 -r 25 -g 50 -b:v 6000k -c:a mp2 -ac 2 -b:a 192k -ar 48000 -minrate 6000k -maxrate 6000k -bufsize 645k -pcr_period 30 -f mpegts -muxrate 6450k "udp://239.1.1.1:10000?pkt_size=1316"
[11:52:48 CEST] <keglevich> this command should produce perfect CBR, but it's not... there are lots of bit-bursts, spikes... is there even an option to generate prefect straight line CBR with ffmpeg?
[11:54:23 CEST] <keglevich> I have actually tried all possible combinations for mpeg2, mpeg4, different ffmpeg versions, the generated udp output is never actually a CBR, just an improved VBR at best with jumps +-2mbits up/down
[11:56:20 CEST] <ChocolateArmpits> keglevich, isn't it that mpegts simply pads data to match a particular overall bitrate rather than strictly making sure the encoded video is uniform in bitrate ?
[11:56:51 CEST] <ChocolateArmpits> not sure if that's possible though
[11:57:33 CEST] <keglevich> I'm not sure, but it looks like it... muxrate sets correct PCR accuracy and delay, it also keeps close to muxrate settings...but it jumps in bursts up and down
[11:58:13 CEST] <keglevich> I compared it to cinegy udp output...cinegy's udp output is actually a 100% perfect CBR straight line...
[11:58:31 CEST] <keglevich> I need that type of stream for sending it out to hardware TS muxers...
[12:01:37 CEST] <keglevich> are there any other parameters maybe worth mentioning (something buffer related) that would help to improve the CBR output?
[12:05:07 CEST] <ChocolateArmpits> keglevich, not sure what cinegy is, but did you try to demux the video stream and check it's bitrate ?
[12:05:59 CEST] <ChocolateArmpits> I suppose you checked the overall mpegts bitrate of that thing
[12:06:00 CEST] <keglevich> I tried to play the stream in VLC... it's almost CBR, but it isn't
[12:06:16 CEST] <keglevich> the bitrate I'm checking with mpeg analyzer
[12:06:26 CEST] <keglevich> the complete bitrate I mean
[12:08:03 CEST] <keglevich> cinegy is just a pro playout app... and it has also UDP multicast streaming output option where you can set the CBR bitrate...and the rate is 100% perfect all the time, also PCR's are fine
[12:08:24 CEST] <keglevich> but with ffmpeg's udp output this seems quite impossible to achieve
[12:10:46 CEST] <keglevich> also, I'm suspecting the mpegts udp out is an issue...if I just record the same stream to .ts file and play it inside VLC, the bitrate is way more accurate then if I watch the same stream directly over UDP
[12:16:03 CEST] <ChocolateArmpits> keglevich, udp has bitrate and burst_bits options, maybe that's the key ?
[12:17:52 CEST] <keglevich> bitrate as udp output options I also tried...burst_bits don't know which value to set...I didn't find any real time use examples for burst_bits anywhere
[14:34:54 CEST] <greysondn> So I'm wanting to export 120 frames/second from a video to PNGS, do some magical editing that shouldn't affect FFMPEG with the frames, and then later reassemble those into a video at 120 FPS. This means the original time span of the source video and the time span of the final video should be the same, such that I could mux the audio from the original into the final and still have it synchronized. Can someone give me a hand with that
[14:36:17 CEST] <greysondn> NB: the original is actually running at a target of 60FPS with variable frame rates due to encoder lagging.
[14:36:58 CEST] <atomnuker> yeah, no, you really should use the API and filter the frames yourself
[14:37:15 CEST] <atomnuker> there's no way to dump to files and keep the timestamps variable
[14:37:26 CEST] <greysondn> I don't need the timestamps to be variable.
[14:37:34 CEST] <greysondn> I need a constant output rate of 120/second
[14:37:39 CEST] <greysondn> So the final would be CFR.
[14:38:37 CEST] <atomnuker> you shouldn't use png either as if your input is yuv you have to convert to rgb and back, which isn't lossless
[14:38:53 CEST] <greysondn> The input is RGB.
[14:40:05 CEST] <greysondn> I can get the "best interests" thing, but you seem determined to convince me that FFMPEG is incapable of basically exporting 120 equally spaced screenshots a second from a video.
[14:41:14 CEST] <Mavrik> It's perfectly capable of doing that.
[14:41:21 CEST] <Mavrik> It's just kinda silly I guess.
[14:41:31 CEST] <greysondn> In what sense is it silly?
[14:41:50 CEST] <Mavrik> You'll be wasting a lot of CPU time and I/O compressing, storing and decompressing those frames.
[14:41:53 CEST] <Mavrik> It might take awhile.
[14:41:56 CEST] <atomnuker> yep, that's why its silly
[14:42:01 CEST] <atomnuker> you ought to use the API if you can
[14:42:05 CEST] <greysondn> Computational time is non-issue.
[14:42:10 CEST] <Mavrik> greysondn: but if you just output to PNGs it should "just work"
[14:42:19 CEST] <Mavrik> and then set the same frame rate on the input when reassembling it
[14:43:04 CEST] <greysondn> Okay. So, how do I do that? Assume everything's ideal except for the fact that I want 120 snapshots a second out of the video, preferably sequentially numbered.
[14:43:18 CEST] <TheAMM> How are you going to edit the frames?
[14:43:22 CEST] <TheAMM> As in, what will you do to them?
[14:43:33 CEST] <Mavrik> greysondn: is your video already 120fps CFR?
[14:44:18 CEST] <TheAMM> Since CFR is okay, you can do -vf fps=120 and dump rgb24 or rgba to stdout, read that with something (eg. python) and have the something to dump edited frames to another ffmpeg that's encoding
[14:44:35 CEST] <TheAMM> No temp files or absurd amounts of RAM necessary
[14:44:40 CEST] <greysondn> Mavrik: No. It's approximately 60FPS, VFR. I'm trying to sample high enough that I get the majority of the frames if not all. Nyquist suggests doubling frequency which is where 120 came from.
[14:45:02 CEST] <TheAMM> You get all the frames, though?
[14:45:33 CEST] <Mavrik> ffmpeg -i <your video> -vf fps=120 image-%03d.png
[14:45:38 CEST] <Mavrik> Of course you might get judder
[14:45:49 CEST] <Mavrik> If the frames don't really duplicate cleanly
[14:45:54 CEST] <greysondn> Does it try to interframe blend beyond what the codec itself does?
[14:46:01 CEST] <greysondn> Or should it basically match the codec?
[14:46:13 CEST] <greysondn> Because, I mean, I'm not going to blame FFMPEG if I'm using an awful format XD
[14:46:14 CEST] <Mavrik> It doesn't do anything except drop or duplicate frames.
[14:46:22 CEST] <greysondn> Okay. That should be fine enough then.
[14:46:28 CEST] <greysondn> Only so much can be done.
[14:46:57 CEST] <TheAMM> Can you elaborate further on the 120fps thing?
[14:47:27 CEST] <greysondn> Nyquist's theorem says that to capture all samples at a given framerate you need to sample at twice that framerate.
[14:47:54 CEST] <greysondn> I know I'll drop a few frames here and there because it exceeds 60 at times, but I should catch *most the frames* in a format that I can use to create a CFR video from.
[14:48:02 CEST] <TheAMM> I think I'm misunderstanding something
[14:48:08 CEST] <TheAMM> Or you're misunderstanding video
[14:48:13 CEST] <greysondn> ** all samples at a given frequency
[14:48:19 CEST] <greysondn> ** which framerate is
[14:48:25 CEST] <TheAMM> If you want to record audio, then yes nyquist applies
[14:48:27 CEST] <atomnuker> yeah, you're misunderstanding how vfr works
[14:48:38 CEST] <TheAMM> But you can literally extract all the frames from a video
[14:48:51 CEST] <TheAMM> It's not... arbitary signals in the air
[14:49:02 CEST] <greysondn> Regardless, I'd still need a constant framerate for my own sanity.
[14:49:16 CEST] <greysondn> So extracting some fixed number per second still is the target
[14:49:23 CEST] <TheAMM> It is true though that you might "miss" "frames" if there's a frame with a longer duration
[14:49:35 CEST] <TheAMM> ffmpeg -i in.mkv -vf fps=60 out.mkv
[14:49:39 CEST] <TheAMM> CFR video, done
[14:49:52 CEST] <greysondn> The AMM: doesn't work, loses too many frames.
[14:50:00 CEST] <atomnuker> it won't lose _any_frames
[14:50:03 CEST] <greysondn> Also fails to give me intermediary PNGs.
[14:50:18 CEST] <TheAMM> Can you answer my initial question though
[14:50:19 CEST] <atomnuker> it will duplicate frames only at optimal intervals to make it 60fps
[14:50:25 CEST] <TheAMM> ^
[14:50:40 CEST] <greysondn> It will have to drop some because at times it bleeds over 60fps.
[14:51:07 CEST] <TheAMM> Even if you go 120fps and encode that down to 60fps, you'll lose those frames
[14:51:20 CEST] <TheAMM> You'd have to manually choose which frames you want to keep
[14:51:41 CEST] <greysondn> I could conceivably do that once I had 120/second out, but the eventual output video is planned to be 120fps.
[14:52:16 CEST] <TheAMM> I guess you won't need anything more then
[14:52:27 CEST] <TheAMM> Dump pngs at 120fps and do what you do
[14:52:59 CEST] <greysondn> *shakes head* Developer communities always puzzle me.
[14:53:11 CEST] <greysondn> You stated you had an initial question I hadn't answered. What was it?
[14:57:40 CEST] <greysondn> Oooooookay then. I'll check back in if that doesn't work out.
[14:58:12 CEST] <TheAMM> <TheAMM> How are you going to edit the frames?
[14:58:12 CEST] <TheAMM> <TheAMM> As in, what will you do to them?
[14:58:38 CEST] <greysondn> I fail to see how that's relevant. What seems relevant is that I will edit them, beyond the scope of simple filters.
[14:59:06 CEST] <TheAMM> XY problem
[14:59:13 CEST] <TheAMM> If you can handle NUT, you could keep VFR
[14:59:14 CEST] <greysondn> ?
[14:59:21 CEST] <greysondn> Oh, I see.
[14:59:34 CEST] <TheAMM> More information is better
[14:59:48 CEST] <TheAMM> Please don't come asking for help and then respond with literally "I fail to see how that's relevant"
[14:59:50 CEST] <greysondn> Well, presumably I have the codecs handy for anything I'd like to toss at this machine, but as far as me personally? I'm not interested in working with VFR because I'm hand programming animations by frame count.
[15:01:22 CEST] <greysondn> The "real" frame count is a part of the frame. I can just extract it there.
[15:01:51 CEST] <greysondn> It would match the image source.
[15:02:20 CEST] <greysondn> There is no guarantee that it will match the video, however, which would lead to an audio desync if I didn't just work with it as CFR and animate in that context.
[15:02:49 CEST] <greysondn> I've no doubt there's an easier way, but after about a year mulling it over and driving myself crazy, this really seems like the simplest way.
[15:03:40 CEST] <greysondn> While FFMpeg and friends aren't my strong suit at all, I do have a degree that covers game design and computer animation. In other words, I at least sat in on lectures on such materials, regardless of whether or not I learned anything.
[15:04:07 CEST] <greysondn> Point being, I have some credibility to make that decision for myself.
[15:06:01 CEST] <greysondn> Hopefully my last question of the morning: Why try to give every piece of data but the one requested by my question when I came in? That seems to needlessly complicate things. The question was basically straightforward and giftwrapped.
[15:07:52 CEST] <Mavrik> Because people commonly come here trying to do dumb stuff in a wrong way which will result in broken systems.
[15:07:52 CEST] <TheAMM> Because -vf fps=120 is such a simple thing, and there was no explanation on *why* you want 120fps
[15:08:00 CEST] <TheAMM> The initial assumption is that you don't know what you're doing
[15:08:09 CEST] <Mavrik> This is why we always ask what's the goal.
[15:08:14 CEST] <Mavrik> No answer the question
[15:08:19 CEST] <Mavrik> Especially when people try to do something insane.
[15:08:28 CEST] <Mavrik> Like trying to process 120fps video on PNGs :P
[15:08:38 CEST] <greysondn> I don't find that insane.
[15:09:04 CEST] <greysondn> Nor did I find the video filter documentation easy to read, to the point it made more sense to ask.
[15:09:16 CEST] <Mavrik> Sure.
[15:09:44 CEST] <greysondn> I knew it was in there - rather, found it absurd that a fairly standard encoder/decoder wouldn't have such a thing - but failed to find it.
[15:10:06 CEST] <Mavrik> Hence why we always try to figure out the intent instead of giving the minimal possible answer.
[15:10:21 CEST] <Mavrik> (Especially since probably a lot of us end up maintaning such software after a few years.)
[15:10:49 CEST] <greysondn> I don't know. I would assume this means you've had to deal with a lot of people who weren't clear on what they were at least trying to do in that case. (Regardless of whether what they were trying to do was right or wrong.)
[15:11:23 CEST] <greysondn> That is, I would assume that you deal with people who do well to copy-paste reliably.
[15:12:26 CEST] <Shibe> does av_hwframe_ctx_create_derived have to have different devices?
[15:12:34 CEST] <Shibe> it says "Create and initialise an AVHWFramesContext as a mapping of another existing AVHWFramesContext on a different device. "
[15:12:46 CEST] <greysondn> I'll idle. Someone else clearly needs your help :D
[15:15:41 CEST] <ChocolateArmpits> greysondn, try -vsync 1 -r 60
[15:16:17 CEST] <atomnuker> Shibe: they should, yeah
[15:16:34 CEST] <atomnuker> if you want to duplicate a device context you can just ref it
[15:16:47 CEST] <Shibe> ah, that makes sense
[15:16:48 CEST] <Shibe> thanks
[16:25:00 CEST] <lvlscape_2> hey guys, i've been working on this one ffmpeg encoding puzzle for about two weeks now. not sure what i should be studying/reading about to solve it anymore, have kinda hit a wall. I have a single-boad-computer, the OdroidXU4 .. and i've set up a headless ubuntu on it (started with debian but ended up here) -- and on the headless uuntu, i've installed retropie for gaming..   this seems to be one of the most
[16:25:03 CEST] <lvlscape_2> accepted ways to set up the system to play retro games.  I spent a week tryign to modify the config fiels to record on ti the same way i did on a raspberry pi.. though ended up messing with trying to have ffmpeg capture the emulation.. which has been the challenge.. cant capture with x11grab, theres no xenviroment ont he headless server in use.. tried installign one... the emulationstation hates being on a
[16:25:05 CEST] <lvlscape_2> virtualverminal with xserver and wont start.... thougth i could record the virtual terminal then... recorded the framebuffer, and it appeared to only record the terminal, not whatever it was launching.. began to understand what was launching was 'egl / sdl?' began learning that.. saw sdl / sdl2 were valid ffmpeg... outputs -- not inputs.. didnt see a way to hook onto those to grab them...  spent a few days
[16:25:07 CEST] <lvlscape_2> tryign out kmsgrab.. recompiled ffmpeg a few times with --enable-libdrm to try out kmsgrab.. only to find out that my hardware 'may' not suppor tit and that wasnt working.. and now im unsure if i keep trying to understand kmsgrab, or if theres another way i should be tryign to captue this, or if its even possible...  it must be possible because in the retroarch binary they use ffmpeg to record.. so i spent a
[16:25:09 CEST] <lvlscape_2> day digging through its source's c code trying to udnerstand it.. but it was so abstracted with classes i couldnt find exactly how they hooked onto record the screen
[16:25:31 CEST] <lvlscape_2> (sorry for text wall... mornign coffee + utterly defeated)
[16:27:40 CEST] <lvlscape_2>  / since i've been devoloing with ffmpeg, maybe i should have posted in other channel? sorry for any misteps
[16:28:32 CEST] <furq> does retroarch not just support capturing input
[16:29:13 CEST] <furq> any emulator that actually relies on screen/window capturing for this is pretty dumb
[16:29:31 CEST] <furq> unless you're just trying to livestream
[16:30:23 CEST] <lvlscape_2> yeah, i've been tryign to stream to my rtmp server.
[16:30:56 CEST] <atomnuker> lvlscape_2: give up now
[16:31:19 CEST] <furq> yeah kms support on those sbcs is notoriously spotty
[16:31:52 CEST] <atomnuker> the odroid has crappy closed source drivers so its doubtful you'll be able to capture it via kmsgrab
[16:31:56 CEST] <furq> if you just want to record footage then there are other ways to do it
[16:31:57 CEST] <furq> but not live
[16:32:10 CEST] <atomnuker> plus, even if you do, it doesn't support v4l2 so you can't encode anything you capture
[16:32:31 CEST] <atomnuker> and you can't really do anything even if you were able to capture them as its CPU is far too weak to encode
[16:32:35 CEST] <furq> it's still pretty lazy of retroarch to just have ffmpeg capture the window though
[16:33:54 CEST] <lvlscape_2> atomnuker: i think i have v4l2 working on odroid .. through tryign my ubuntu install
[16:35:38 CEST] <Hello71> actually it only says it uses ffmpeg
[16:35:42 CEST] <lvlscape_2> i was abel to have the rpi capture and stream with no lag and high quality, so i knew odroid would be better.. i didnt want to encode with high compression, was just encoding for speed, and sending over ethernet to a local server to do ehavier encoding
[16:35:51 CEST] <Hello71> doesn't say it captures with ffmpeg
[16:37:20 CEST] <lvlscape_2> Hello71: the retroarch record driver seemed to be all ffmpeg related: https://github.com/libretro/RetroArch/blob/399011a4fa8e880b48dc049e856515131245be26/record/record_driver.h
[16:37:33 CEST] <lvlscape_2> oops thats the .h file, meant to link the .c file
[16:38:25 CEST] <lvlscape_2> https://github.com/libretro/RetroArch/blob/11f26ed549b03f58b8a8f1ea2eaabf1950bc45d3/record/record_driver.c
[16:41:54 CEST] <jkqxz> I've had kmsgrab working on an Odroid XU4, recording stuff like a terminal or kmscube (I was only testing).  The lack of any interop makes it pretty painfully slow, though, because the only thing you can do directly is download to CPU memory.
[16:42:05 CEST] <lvlscape_2> furq: what other ways were you refering to, do you have any thoughts on how i could approach recording withint he same device? or were you thinking more along the lines of a second 'capture card' ?
[16:42:33 CEST] <jkqxz> If the V4L2 M2M stuff gained DRM PRIME interop then it should be usable.
[16:42:47 CEST] <furq> a lot of emulators will just spit out rawvideo directly
[16:43:07 CEST] <furq> although idk of any off the top of my head that will do it live
[16:43:24 CEST] <furq> normally you'd record an input file and then have it render that
[16:58:18 CEST] <lvlscape_2> jkqxz: even if this was a failed attempt, im enjoying reading about why it failed to understand it. Im semi-famielar with V4L2 as i've been using it to capture webcam, and i understand it coudl do a lot more .. so i was intrigued by your comment about M2M / DRM PRIME interop -- was wondering if you could add context here to assist in my study?
[17:39:50 CEST] <Hello71> yeah, it absolutely doesn't use window capture
[17:41:12 CEST] <Hello71> you can see in https://github.com/libretro/RetroArch/blob/master/record/drivers/record_ffmpeg.c it uses av_interleaved_write_frame
[17:41:17 CEST] <Hello71> ffmpeg doesn't mean the ffmpeg command
[17:42:11 CEST] <Hello71> and you can see in recording_dump_frame it has a function to read back the currently displayed frame
[17:42:14 CEST] <Hello71> lvlscape_2--
[18:51:25 CEST] <lvlscape_2> Hello71: thanks so much, i was way out of my depth, its great to understand what was happenign after spendign so many hours looking at it
[20:27:01 CEST] <dan9er> Ok so, i've been searching around for how to do this, but i've had no luck. Looking to get a ffmpeg pro here. I have a SBS Full Frame video that is 648x720p per eye. I want to use ffmpeg to add metadata to it so YouTube will accept it as a 3D video, with the 3D option in the cog menu and all. I also want to do this without re-encoding. I'm
[20:27:01 CEST] <dan9er> asking here because all of the tutorials I found ethier re-encode the video (bad!) or make YouTube treat it as a SBS Half Frame video (wrong!).
[20:27:42 CEST] <dan9er> Using orange for emphasis instead of bold because IRC
[20:29:56 CEST] <^Neo> Is there a way to multicast a MPEGTS file and preserve the TS header information?
[21:04:42 CEST] <dan9er> I just sent this to the user mailing list as well
[21:04:53 CEST] <dan9er> gonna part now, gtg
[22:04:24 CEST] <QuestionAboutMon> I have looked online for guides to compress mono, if I follow the advice `ffmpeg -i .\input.mp3 -ab 8k bad-output.mp4` the output is 16kbps not 8kbps and MediaInfo states that it is two channels. I have also tried `ffmpeg -i .\input.mp3 -ab 8k bad-output.mp4 -ac 1` (also does not work)
[22:05:59 CEST] <QuestionAboutMon> How do I compress a mono audio track without it changing to two channels?
[22:06:56 CEST] <klaxa> move the -ac 1 before output.mp4
[22:09:14 CEST] <QuestionAboutMon> klaxa: That does not change the issue
[22:55:09 CEST] <ChocolateArmpits> is -ab an alias for -b:a ??
[22:56:40 CEST] <JEEB> -ab is the old option, while -b:a is -b with the "audio-specific" suffix which can be used with other options, too
[22:56:46 CEST] <JEEB> or even -b:a:0
[22:56:52 CEST] <JEEB> for the first output audio track
[23:41:50 CEST] <ineedffmpegforwi> Hello, I'm looking for information on why https://ffmpeg.zeranoe.com is down.
[23:42:01 CEST] <ineedffmpegforwi> Could anyone help me out?
[23:42:18 CEST] <JEEB> it's handled by zeranoe, not related to FFmpeg itself at all
[23:42:48 CEST] <ineedffmpegforwi> I see, is there any other source to download ffmpeg builds for windows? As I'm not a developer.
[23:43:40 CEST] <ineedffmpegforwi> But I think it's atleast a bit related, because when you click on "Download" on the official website it redirects you to zeranoe.
[23:44:22 CEST] <JEEB> that's highly misleading in a way then :P
[23:44:29 CEST] <JEEB> there are no official binaries of FFmpeg
[23:44:38 CEST] <JEEB> FFmpeg only produces the source code officially
[23:45:31 CEST] <ineedffmpegforwi> Okay and do you know a website where I can download a windows build?
[23:45:32 CEST] <azarus> JEEB: go to https://ffmpeg.org -> click the big green "Download" button :/
[23:45:52 CEST] <ineedffmpegforwi> Yeah, but this takes me to zeranoe @azarus.
[23:46:10 CEST] <JEEB> azarus: that takes me to https://ffmpeg.org/download.html that has a big green thing linking to the tarball for 4.0.1
[23:46:17 CEST] <azarus> ineedffmpegforwi: yeah, i was talking to JEEB
[23:46:24 CEST] <ineedffmpegforwi> Sorry
[23:46:31 CEST] <JEEB> you have the blue windows logo tho, which then links to https://ffmpeg.org/download.html#build-windows
[23:46:41 CEST] <JEEB> which has a link to zeranoe, yes
[23:46:50 CEST] <JEEB> but at least it's not an automated redirect
[23:46:51 CEST] <JEEB> :V
[23:47:10 CEST] <azarus> well, sure, he could compile it himself...
[23:47:39 CEST] <JEEB> no, I'm just saying that it's not like he tried to note which was "redirects to zeranoe"
[23:47:51 CEST] <JEEB> but yea, sure - for windows end users binaries are useful
[23:54:54 CEST] <karlll> Hi. The website https://ffmpeg.zeranoe.com does not work for me, see a loop of HTTP redirects in webconsole. Anyone experience the same?
[23:55:00 CEST] <ineedffmpegforwi> HA
[23:55:07 CEST] <ineedffmpegforwi> I'm not alone!
[23:55:49 CEST] <karlll> Hehe
[23:56:20 CEST] <JEEB> karlll: unrelated to the project, it's some dude's (zeranoe's) builds
[23:56:24 CEST] <ineedffmpegforwi> I'm also looking for a windows build but I just can't find any. Can only find references to zeranoe.
[23:56:36 CEST] <JEEB> figuring out how to communicate with zeranoe would probably be most useful if you want his builds
[23:56:53 CEST] <ineedffmpegforwi> I don't really want "his" builds, I just want "a" windows build.
[23:57:49 CEST] <JEEB> can't really recommend any as legitimate to be honest, as I mostly just build my own when needed
[23:57:53 CEST] <karlll> I just kinda hoped he/she might be reading this
[23:58:04 CEST] <JEEB> he has been on IRC sometimes
[23:58:06 CEST] <JEEB> but he isn't now
[23:58:42 CEST] <karlll> But just found an email address on git hub, i'll try that. Thanks
[00:00:00 CEST] --- Sat Jul  7 2018