[Ffmpeg-devel-irc] ffmpeg.log.20151222

[burek]

[00:03:51 CET] <Guiri> I'm having some trouble grabbing a stream from ffserver.  I think I'm successfully streaming via ffmpeg to the ffm endpoint.  I can see it in both logs, the firewall is open.  But I don't have a feed at http://domain.com:8090/feed.flv
[00:35:44 CET] <scottchiefbaker> I've got a bunch of old avis I'm trying to convert to x264 and I'm doing ffmpeg -y -i in.avi -preset slower -crf 23 -strict -2 -c:a aac -q:a 0.8 /tmp/out.avi
[00:35:53 CET] <scottchiefbaker> But I'm getting: [libx264 @ 0x47cb920] height not divisible by 2 (320x237)
[00:35:55 CET] <scottchiefbaker> and it fails
[00:36:07 CET] <scottchiefbaker> What's the BEST way to fix that?
[00:55:00 CET] <c_14> -vf scale=iw:-2
[01:09:17 CET] <scottchiefbaker> c_14: Can you translate that to english for me?
[01:09:42 CET] <scottchiefbaker> c_14: It works... just not sure what it does :)
[01:10:04 CET] <c_14> scales the height to the nearest (larger) factor of 2
[01:10:30 CET] <scottchiefbaker> iw means height?
[01:11:16 CET] <scottchiefbaker> Also... what if my input is already a multiple of 2. In other words, is it safe to add this to my script to auto-scale all the incorrect ones, and leave the rest alone?
[01:12:27 CET] <Guest72961> Can someone tell me how to use iw/ow in scale when the video is being auto rotated (because of the rotate metadata)? Is it possible to scale after the auto rotate and have these values respect the rotated video?
[01:12:51 CET] <Guest72961> or do I have to set noautorotate and do it myself?
[01:13:24 CET] <c_14> scottchiefbaker: iw is the input width, that's to make sure it keeps the same input width. The -2 then means to scale the height to a factor of 2 preserving the aspect ratio.
[01:14:19 CET] <c_14> Guest72961: afaik the rotate should be the last filter applied, therefore every filter you manually set will execute before it
[01:18:35 CET] <Guest72961> c_14: Im using this "scale=trunc(min(ow\,1280)/32)*32:trunc(min((ow/dar)\,720)/32)*32" and what its doing is rotating then scaling so my video is squashed
[01:18:43 CET] <Guest72961> same with iw
[01:20:30 CET] <c_14> Y
[01:20:39 CET] <c_14> ou might just have to do it manually and set -noautorotate
[01:27:30 CET] <Guest72961> c_14: Do you know were the source code is for autorotate so I can emulate it?
[01:34:13 CET] <c_14> ffmpeg_filter.c lines 752-770
[01:41:16 CET] <Guest72961> ta
[02:09:40 CET] <E-TARD> I need help real bad! I'm trying to do a hls out from ffserver but I do not get how to pass the hls settings "like hls_flags single_file" and "hls_segment_filename" in the ffserver.conf I was thinking AVPresetVideo?
[02:11:31 CET] <E-TARD> how do i tell ffserver hls settings "like hls_flags single_file" and "hls_segment_filename" in the ffserver.conf
[02:44:12 CET] <E-TARD> can somone help me :(
[02:44:51 CET] <E-TARD> i have been working on ffserver for man many days now
[02:45:04 CET] <E-TARD> there is not much help when it comes t this
[02:46:09 CET] <E-TARD> anyone alive in here?
[02:52:32 CET] <E-TARD> ok amybe someone could help me with this
[02:52:33 CET] <E-TARD> [hls @ 0x808e39c]failed to rename file .tmp to
[02:54:08 CET] <E-TARD> .tmp when you cat it you see the data of a m3u8 file now why does it failed to rename file. i don't know
[03:08:42 CET] <waressearcher2> E-TARD: hallo, wie geht's es dir ?
[03:08:51 CET] <E-TARD> ?
[07:15:47 CET] <pinPoint> so I had my pc do a bench test for time. libx264 and libx265. -preset ultrafast, superfast, veryfast, faster, fast, medium, slow, slower, veryslow on both codecs. Interesting results
[07:16:10 CET] <pinPoint> a 3+ run will post the time results(benchmark) soonish
[07:16:37 CET] <pinPoint> 3 day plus run*
[07:43:17 CET] <C0nundrum>  Hello if i'm recording an flv and ffmpeg is sundenly clossed is there anyway to recover the flv ?
[08:33:38 CET] <E-TARD> C0nundrum & pinPoint i have been w8ing hrs for a reply to my ? & i feel you will end up w8ing as long or longer
[08:33:47 CET] <E-TARD> 296 user & no one replys
[08:35:45 CET] <JEEB> C0nundrum: flv should not have an index per se so highly likely that you can just pöay that flv file
[08:35:59 CET] <JEEB> *play
[08:47:21 CET] <satinder> hi
[08:47:34 CET] <satinder> I have CP Plus DVR
[08:48:39 CET] <satinder> the video stream from dvr is coming through ethernet  port to my pc
[08:49:05 CET] <satinder> I can play that stream through vlc like following
[08:49:11 CET] <satinder> rtsp://admin:admin@192.168.1.240:554/cam/realmonitor?channel=3&subtype=0
[08:49:38 CET] <satinder> but when I am trying to play with ffplay then I can not play
[08:49:51 CET] <satinder> I use following command
[08:50:02 CET] <satinder> ffplay  rtsp://admin:admin@192.168.1.240:554/cam/realmonitor?channel=3&subtype=0
[08:50:11 CET] <satinder> Is there any problem
[08:50:22 CET] <satinder> please anybody help me
[08:50:25 CET] <satinder> ??
[08:57:09 CET] <satinder> is there anyone who can help me
[08:57:11 CET] <satinder> ??
[09:22:27 CET] <satinder> Hi
[09:22:35 CET] <satinder> please anyone help
[09:27:26 CET] <durandal_1707> what to help?
[09:33:12 CET] <satinder> durandal_1707 : hi thanks for reply
[09:33:22 CET] <satinder> My query is following :
[09:33:32 CET] <satinder> hi
[09:33:33 CET] <satinder> <satinder> I have CP Plus DVR
[09:33:33 CET] <satinder> <satinder> the video stream from dvr is coming through ethernet  port to my pc
[09:33:33 CET] <satinder> <satinder> I can play that stream through vlc like following
[09:33:33 CET] <satinder> <satinder> rtsp://admin:admin@192.168.1.240:554/cam/realmonitor?channel=3&subtype=0
[09:33:34 CET] <satinder> <satinder> but when I am trying to play with ffplay then I can not play
[09:33:37 CET] <satinder> <satinder> I use following command
[09:33:39 CET] <satinder> <satinder> ffplay  rtsp://admin:admin@192.168.1.240:554/cam/realmonitor?channel=3&subtype=0
[09:33:41 CET] <satinder> <satinder> Is there any problem
[09:33:43 CET] <satinder> <satinder> please anybody help me
[09:34:10 CET] <satinder> if you have any idea thwn please help
[09:34:36 CET] <satinder> sorry for copy & Paste I hope you understand that issue
[09:34:39 CET] <satinder> :)
[09:37:47 CET] <satinder> durandal_1707 : ??
[09:42:03 CET] <durandal_1707> what ffplay reports?
[09:48:54 CET] <nindustries> Hi guys
[09:49:05 CET] <nindustries> just wanted to say thanks to all of you
[09:49:13 CET] <nindustries> because I love ffmpeg :)
[09:50:57 CET] <satinder> satinder at satinder-vsys:~/bin$ ffplay version 2.8.4 Copyright (c) 2003-2015 the FFmpeg developers
[09:51:01 CET] <satinder>   built with gcc 4.9.2 (Ubuntu 4.9.2-10ubuntu13)
[09:51:01 CET] <satinder>   configuration: --prefix=/opt/ffmpeg_build --pkg-config-flags=--static --extra-cflags=-I/home/satinder/ffmpeg_build/include --extra-ldflags=-L/home/satinder/ffmpeg_build/lib --bindir=/home/satinder/bin --enable-gpl --enable-libass --enable-libfdk-aac --enable-libfreetype --enable-libmp3lame --enable-libopus --enable-libtheora --enable-libvorbis --enable-libvpx --enable-libx264 --enable-nonfree
[09:51:02 CET] <satinder>   libavutil      54. 31.100 / 54. 31.100
[09:51:02 CET] <satinder>   libavcodec     56. 60.100 / 56. 60.100
[09:51:03 CET] <satinder>   libavformat    56. 40.101 / 56. 40.101
[09:51:04 CET] <satinder>   libavdevice    56.  4.100 / 56.  4.100
[09:51:08 CET] <satinder>   libavfilter     5. 40.101 /  5. 40.101
[09:51:10 CET] <satinder>   libswscale      3.  1.101 /  3.  1.101
[09:51:12 CET] <satinder>   libswresample   1.  2.101 /  1.  2.101
[09:51:14 CET] <satinder>   libpostproc    53.  3.100 / 53.  3.100
[09:51:16 CET] <satinder> [rtsp @ 0x7fb640004f60] method DESCRIBE failed: 404 Not Found=0/0
[09:51:18 CET] <satinder> rtsp://admin:admin@192.168.1.240:554/cam/realmonitor?channel=3: Server returned 404 Not Found
[09:51:20 CET] <satinder> durandal_1707 : please see above
[09:51:37 CET] <satinder> :(
[11:04:53 CET] <chuugar> Hi all, is there a way to add a cover art to an OGG audio file with ffmpeg?
[14:12:00 CET] <C0nundrum> Prev querstion: Hello if i'm recording an flv and ffmpeg is sundenly clossed is there anyway to recover the flv ?
[14:12:34 CET] <C0nundrum> JEEB if ffmpeg isn't closed properly for it to complete the file, it doesn't play in any player.
[14:12:48 CET] <Hybridesque> Afternoon everyone
[14:20:47 CET] <JEEB> C0nundrum: that shouldn't be the case with FLV
[14:20:51 CET] <JEEB> after all, you can stream it just fine
[14:20:55 CET] <JEEB> are you sure your output is FLV :P
[14:21:10 CET] <JEEB> or if you have any data at all in the FLV file
[14:21:30 CET] <JEEB> if you run `ffprobe file.flv` does it say anything (use pastebin to copy the information and link it here)
[14:21:36 CET] <JEEB> (don't paste onto the channel)
[15:27:24 CET] <phillip> hi, I want to convert a mts file to mp4. I used the following arguments: https://gist.github.com/PhillipSz/199b3dd2b785e6f4d1dc the problem is that it adds some weird lines, when the camera moves. How can I fix that?
[15:29:01 CET] <fritsch> phillip: the original file plays without those lines?
[15:29:31 CET] <phillip> fritsch: in gnome video (totem) yes.
[15:29:37 CET] <fritsch> and in vlc no, right?
[15:29:47 CET] <fritsch> most likely it's interlaced
[15:29:49 CET] <phillip> yes
[15:29:52 CET] <fritsch> and needs deinterlacing
[15:30:03 CET] <fritsch> what does: mediainfo of this file say?
[15:30:49 CET] <phillip> fritsch: mediainfo ?
[15:31:36 CET] <phillip> fritsch: ahh wait a moment I install it
[15:32:09 CET] <Mavrik> phillip, you enable deinterlacing in the player
[15:32:13 CET] <fritsch> -vf "yadif=0:-1:0" makes it go away? (after the -i yourfile?)
[15:32:22 CET] <fritsch> Mavrik: yeah, totem does it automatically, vlc not :p
[15:32:25 CET] <fritsch> that's most likely his issue
[15:32:26 CET] <Mavrik> Also don't encode into MPEG-4 ffs -_-
[15:32:36 CET] <Mavrik> Yes, deinterlacing will affect quality though.
[15:32:39 CET] <fritsch> but if he is transcoding anyways, he can directly use yadif to deinterlace
[15:32:51 CET] <Mavrik> It introduces a blur when moving camera
[15:33:05 CET] <fritsch> not really
[15:33:11 CET] <fritsch> but with this 25 fps most likely
[15:33:24 CET] <phillip> fritsch: http://sprunge.us/YdFe
[15:33:33 CET] <fritsch> see it's interlaced
[15:33:40 CET] <fritsch> either you deinterlace in the player
[15:33:43 CET] <fritsch> or with ffmpeg
[15:34:07 CET] <fritsch> and don't use mpeg4 please
[15:34:16 CET] <fritsch> that's some 20 years too old
[15:34:43 CET] <phillip> fritsch: okey, what should I use instead?
[15:35:08 CET] <fritsch> keep h264?
[15:35:17 CET] <fritsch> you just want to convert it cause of the "lines" right?
[15:35:42 CET] <fritsch> anyways, add -vf yadif
[15:35:47 CET] <fritsch> and see if you like the output
[15:36:05 CET] <mcepl> hi, I know this is probably not a right channel (can somebody suggest a better one, please?) but can somebody suggest SIMPLE UPnP server with miniamal dependencies (just for home server with RHEL-6)?
[15:36:19 CET] <phillip> fritsch: yep
[15:38:33 CET] <phillip> fritsch: yeah it work, thank you!
[15:39:11 CET] <fritsch> can you get me a mediainfo of the resulting file?
[15:39:21 CET] <fritsch> what happens if you ommit -r 25?
[15:39:48 CET] <phillip> http://sprunge.us/NPHe
[15:40:14 CET] <fritsch> okay, now keep away that -r 25
[15:40:22 CET] <fritsch> and use "yadif=1"
[15:40:26 CET] <fritsch> and try again
[15:40:31 CET] <fritsch> do you like that even more?
[15:40:49 CET] <phillip> http://sprunge.us/GYCV
[15:40:59 CET] <phillip> ahh wait
[15:41:01 CET] <fritsch> nope, that's not it
[15:41:02 CET] <fritsch> :-)
[15:41:14 CET] <phillip> I try with "yadif=1" :)
[15:41:22 CET] <fritsch> and without -r 25
[15:43:13 CET] <phillip> fritsch: so instead of "-vcodec mpeg4" should I also just use "-vcodec h264" ?
[15:43:25 CET] <fritsch> nope - one step after the other
[15:43:38 CET] <fritsch> first see if you prefer the movement with yadif=1
[15:43:53 CET] <phillip> okey so http://sprunge.us/WiDU
[15:44:52 CET] <fritsch> how does it look the video?
[15:45:06 CET] <fritsch> can you still play it?
[15:45:56 CET] <phillip> fritsch: yes, it's perfect!
[15:46:27 CET] <fritsch> okay
[15:46:33 CET] <fritsch> now which ffmpeg version are you using?
[15:46:59 CET] <phillip> ffmpeg version 2.8.4
[15:47:04 CET] <fritsch> replace "yadif=1" with "w3fdif" (without any number)
[15:47:07 CET] <fritsch> and produce another file
[15:47:11 CET] <fritsch> do you like that better?
[15:49:16 CET] <fritsch> for video part of your encoding, use (via the wiki): -c:v libx264 -preset slow -crf 22
[15:49:32 CET] <fritsch> add the vf= yadif=1 or w3fdif
[15:49:40 CET] <fritsch> care for your audio and that's it
[15:49:52 CET] <fritsch> w3fdif is the bbc deinterlacer
[15:50:00 CET] <phillip> fritsch: so it's mostly the same, maybe a bit better :)
[15:50:16 CET] <fritsch> some say it has "much" higher quality than yadif (though - some hang with their face directly in front of the screen)
[15:50:27 CET] <fritsch> i have not seen any numbers proving that
[15:50:30 CET] <fritsch> yadif is quite efficent
[15:50:52 CET] <fritsch> oki, now assemble your ffmpeg command and see that you get something good out of the video
[15:56:38 CET] <phillip> fritsch: okey, thanks a lot! So it needs much more time to convert the file now, but that's fine :)
[15:57:04 CET] <Mavrik> and please do use "-vcodec libx264"
[15:57:16 CET] <Mavrik> Kinda pointless grabbing videos from BR and then destroying them :P
[16:00:57 CET] <phillip> Mavrik: okey so I should use https://gist.github.com/PhillipSz/199b3dd2b785e6f4d1dc this command now?
[16:01:17 CET] <shincode> holy crap people are alive
[16:01:18 CET] <Mavrik> -vcodec and -c:v are obviously the same parameter.
[16:01:25 CET] <Mavrik> But yeah.
[16:01:30 CET] <shincode> Building static libraries of ffmpeg collection with optimizations
[16:01:32 CET] <phillip> hehe
[16:01:47 CET] <shincode> link all statics in my shared library whines about Recompile with fPic
[16:01:58 CET] <shincode> I did --enable-pic on static ffmpegs and without
[16:02:02 CET] <shincode> same with my shared libraries
[16:02:04 CET] <shincode> arm processor
[16:02:12 CET] <shincode> doesnt seem to do this with x86 32 bit
[16:02:26 CET] <phillip> so thanks and bye Mavrik and fritsch !
[16:03:06 CET] <shincode> Lies!
[16:03:13 CET] <shincode> someone said you cant mix statics with shared
[16:03:22 CET] <shincode> I can certinally do this on windows
[16:03:34 CET] <shincode> unless its because theres assembly code it cant mix them
[16:04:00 CET] <shincode> im certain if i disable the assembly instructions it will link and be fine but then i loose the optimizations or will I?
[16:04:42 CET] <shincode> Since when was ffmpeg devs better than a optimizations made by compiler which isa collection of nerds in a cave that alls they do is optimize unless Ox options was invented by ffmpeg
[16:06:33 CET] <durandal_1707> what are you talking about?
[16:08:00 CET] <DHE> I dunno but gcc 5.2.0 with -flto and inline ASM disabled performs worse than gcc 4.3 without -lfto and inline ASM enabled
[16:08:10 CET] <DHE> so I'm inclined to say the hand-written ASM is actually pretty good
[16:09:12 CET] <shincode> i read about that with ftlo
[16:09:47 CET] <shincode> my problem now is its whining that some in libswscale is not relocatable
[16:10:07 CET] <shincode> so I'm guessing enable-pic and the cflags in config.mak are not carrying to libswscale
[16:11:14 CET] <DHE> I think the issue is the local labels in the asm are being duplicated and then the linker has a fit about duplicate symbols
[16:11:35 CET] <DHE> or something to that effect
[16:13:23 CET] <shincode> im only linking in avformat, avcodec, avutil, and swscale if that helps.
[16:16:50 CET] <shincode> VideoStreamer.o: relocation R_ARM_THM_MOVW_ABS_NC against `__pthread_key_create' can not be used when making a shared object; recompile with -fPIC
[16:16:54 CET] <shincode> is what it whines about
[16:22:28 CET] <JEEB> rebuild all of your dependencies with PIC
[16:22:40 CET] <JEEB> like, make sure the whole chain is built with PIC
[16:22:46 CET] <JEEB> with ffmpeg the option is --enable-pic
[16:23:52 CET] <JEEB> if it still fails then, make your own thingy build with PIC as well
[16:24:02 CET] <JEEB> usually it's just the dependencies that need the PIC if it tells you like that
[16:24:06 CET] <JEEB> (IIRC)
[16:24:34 CET] <shincode> I did this
[16:24:51 CET] <shincode> i even when in config.mak and stuck -fPICS on AS flags lol
[16:25:27 CET] <shincode> i even verify cflags gets populate with fPIC when using enable-pic as well
[16:25:37 CET] <JEEB> well I mean your dependencies
[16:25:58 CET] <JEEB> and then finally your code
[16:26:02 CET] <JEEB> if it then still fails then gj
[16:28:50 CET] <shincode> you mean gl?
[16:28:53 CET] <shincode> lol
[16:29:41 CET] <shincode> reality is if i can find the linker on the make file Id link it all to the same library shared or static
[16:52:57 CET] <shincode> --enable-pic --disable-asm used
[16:53:01 CET] <shincode> hope it works
[17:05:47 CET] <shincode> this is wierd....
[17:06:28 CET] <shincode> with x86 I can link just fine statics with asm compiled and optimizations into a shared library which makes use of avcodec but
[17:06:33 CET] <shincode> arm can't....
[17:06:38 CET] <shincode> with the same configure line
[17:11:30 CET] <shincode> there a way to ditch ffmpegs pthread.c objects
[17:11:38 CET] <shincode> like --disable-reinvent-the-wheel
[17:11:55 CET] <shincode> I think this is the problem considering pthread_create_key is whatis whining about
[17:29:32 CET] <shincode> I got closer to this issue
[17:29:40 CET] <shincode> some reason x86 uses pthread.so in the end
[17:29:50 CET] <shincode> but on arm its either compiling a pthread.c into it
[17:29:58 CET] <shincode> Or linking with some pthread.a somewhere
[18:04:01 CET] <shincode> so it appears when doing a diff of config.mak of x86 and arm build it seems like YASM is missing -DPIC
[18:35:42 CET] <Franciman> Hello, I don't pretty much understand what libavfilter is for
[18:35:54 CET] <Franciman> can you give me some pointer to an explaination?
[18:36:05 CET] <Franciman> maybe with examples
[18:41:02 CET] <DHE> the -vf and -af parameters for ffmpeg invoke filters from there
[18:43:39 CET] <Franciman> DHE, where can I find some examples?
[18:44:17 CET] <DHE> I'm not the guy to ask API questions about
[18:44:53 CET] <Franciman> :P
[18:44:58 CET] <Franciman> sorry then
[18:47:33 CET] <c_14> Franciman: doc/examples/filter*
[18:48:26 CET] <Franciman> c_14, would creating a waveform image from audio be a suitable example of filters use?
[18:49:13 CET] <Franciman> anyways thanks
[18:50:06 CET] <c_14> Well, that's basically what showwaves does
[18:50:41 CET] <Franciman> oh great, thanks
[18:53:47 CET] <shincode> i got it!
[18:54:00 CET] <shincode> for the arm problem on shared library add -mword-relocations arm specific..
[18:54:09 CET] <shincode> they say -fPIC should do same thing but its a lie of lies
[18:54:16 CET] <shincode> someone somewhere probablys says option is bad
[18:54:20 CET] <shincode> but wont say why
[18:54:27 CET] <shincode> anyone can sahre any input?
[19:21:53 CET] <shincode> anything
[20:15:30 CET] <lmat> I'm playing a .wav that has 8 channels, and ffmpeg isn't impressed :-) The error is...
[20:16:25 CET] <lmat> pcm_s24le @ 0x7fee....] Channel layout '5.1' with 6 channels does not match specified number of channels 8: ignoring specified channel layout"
[20:16:42 CET] <lmat> I'm guessing that means that the .wav file is specifying itself as a 6-channel audio, but it actually has 8 channels?
[20:16:47 CET] <lmat> It's not playing correctly, too.
[20:16:51 CET] <JEEB> that's not an error
[20:16:54 CET] <JEEB> oh wait
[20:17:15 CET] <JEEB> you are setting a 5.1 layout somewhere yourself
[20:17:21 CET] <JEEB> or the API is
[20:17:34 CET] <JEEB> and you have 8 channels in that wav as data
[20:17:37 CET] <lmat> My pulse audio profile is 7.1
[20:17:47 CET] <JEEB> please pastebin full command line and terminal output
[20:17:48 CET] <JEEB> and link here
[20:17:59 CET] <lmat> oh boy, I'll try! just a sec
[20:18:28 CET] <lmat> I'm using computer A to ssh to computer B for irc. All this is happening on computer C ^_^
[20:18:40 CET] <lmat> I think I'll use computer A, and ssh to computer C, copy this into computer B...
[20:19:31 CET] <JEEB> you could just use ssh as a ProxyCommand
[20:19:44 CET] <lmat> http://sprunge.us/QfFP
[20:20:36 CET] Action: lmat never heard of that...reading
[20:22:21 CET] <shincodex> Hey thats what i was doing earlier
[20:22:36 CET] <JEEB> Host target_boxen\n  ProxyCommand ssh -q -W %h:%p jeeb at boxen
[20:22:43 CET] <shincodex> im xrdping from windows into x86 mate to irc while xrdping into arm processor doing all the work
[20:22:50 CET] <JEEB> in your ~/.ssh/config
[20:23:16 CET] <JEEB> that way when you ssh to target_boxen, you will automagically proxy it through boxen
[20:25:12 CET] <shincodex> the only sad thing is i cant copy paste between sessions :((((
[20:25:14 CET] <lmat> JEEB: ah, I see.
[20:25:33 CET] <lmat> okay, pasted output
[20:25:50 CET] <JEEB> does ffprobe say the same?
[20:26:27 CET] <lmat> JEEB: dunno, just a sec
[20:27:26 CET] <lmat> yes
[20:27:51 CET] <JEEB> ok
[20:27:52 CET] <lmat> hmm, maybe I can hexdump the file and read it...I used to know the format ^_^
[20:28:05 CET] <JEEB> so basically the wav says something and the data doesn't agree, which is funky
[20:28:14 CET] <lmat> JEEB: not terribly ^_^
[20:28:34 CET] <JEEB> there's a WAVE extended something that has the channel layout in it
[20:28:41 CET] <JEEB> don't remember how the structure was called :/
[20:28:50 CET] <lmat> JEEB: ooohhhh, I'm just looking at the fmt section.
[20:30:11 CET] <lmat> within the fmt header, the first field is "audio fmt"...the value in the file is  fe ff 08 00  (LE)
[20:32:51 CET] <lmat> http://soundfile.sapp.org/doc/WaveFormat/ says that the fmt chunk should have a length of 16. Mine reports 0x28...
[20:33:04 CET] <lmat> Maybe I'll just write my own wave file instead of this one I found online
[20:33:13 CET] <lmat> Here's the deal: VLC plays it as expected.
[20:33:49 CET] <E-TARD> How to i set things like hls_segment_filename or segment_list or segment_list_type in ffserver.conf? do i use avoptionvideo or avpresetvideo?
[20:34:15 CET] <lmat> I'll just bugger off unless anyone's interested in the nitty gritty here.
[20:36:38 CET] <E-TARD> so no one know how to set settings like hls_segment_filename or segment_list or segment_list_type in ffserver.conf?
[20:37:16 CET] <JEEB> E-TARD: you are using ffserver
[20:37:17 CET] <JEEB> good luck
[20:37:30 CET] <JEEB> it's mostly unmaintained and only few know how to use it
[20:37:35 CET] <E-TARD> i have got far with it & know soo much more now then i did before
[20:37:37 CET] <JEEB> if you can get your stuff done with ffmpeg itself
[20:37:42 CET] <JEEB> then do so
[20:37:44 CET] <E-TARD> i just hit a wall with this
[20:37:47 CET] <JEEB> if not, my condolences
[20:37:54 CET] <E-TARD> i need ffserver
[20:38:02 CET] <JEEB> then I'm sorry
[20:38:06 CET] <E-TARD> i have got ffmpeg down pat
[20:38:20 CET] <E-TARD> but i need a ffserver to put on my server
[20:38:31 CET] <JEEB> yeah, ffmpeg is OK. it's just ffserver that is as "dead" as it is
[20:38:40 CET] <JEEB> if you get it to work, great
[20:38:42 CET] <E-TARD> not from what i have seen
[20:38:44 CET] <JEEB> if not, you get to keep both pieces
[20:38:47 CET] <E-TARD> it can do soo much
[20:38:59 CET] <JEEB> right, but that still doesn't make it not dead
[20:39:10 CET] <JEEB> it's only very lightly poked at by a very limited set of people
[20:39:40 CET] <E-TARD> all i need is to set things like -hls_flags single_file in ffserver.conf
[20:39:43 CET] <JEEB> I wouldn't be against removing ffserver altogether because nobody knows how to use it and nobody knows if it's broken or not :P
[20:39:57 CET] <JEEB> but I'm not in control of that :)
[20:40:07 CET] <E-TARD> i feel that to do that i have to set it with something like avoptionvideo or avpresetvideo?
[20:40:26 CET] <E-TARD> without ffserver there would only be pay shit
[20:40:30 CET] <E-TARD> f that
[20:40:44 CET] <JEEB> not really
[20:40:45 CET] <E-TARD> i dont have $800 and up to blow on this
[20:40:58 CET] <JEEB> there's icecast, and for HLS/DASH you use normal httpds
[20:41:09 CET] <JEEB> because it's literally shit on a file system
[20:41:32 CET] <E-TARD> can it do live flv over http
[20:41:57 CET] <JEEB> why not? although nothing supports that
[20:42:05 CET] <JEEB> because flv is only used in rtmp(e)
[20:42:16 CET] <JEEB> or well, not only but you get my point :P
[20:42:21 CET] <E-TARD> i need ffserver to take a feed from me & pump out a flv over http a hls & a dash
[20:42:21 CET] <JEEB> it's flash-only
[20:42:49 CET] <E-TARD> i have got it to do everything i want but the hls and dash because of plays
[20:43:04 CET] <E-TARD> playlists
[20:43:08 CET] <JEEB> the receiving side is what I'd do with ffmpeg I guess
[20:43:25 CET] <JEEB> or did ffmpeg have listen mode so you could push into it?
[20:43:32 CET] <E-TARD> does any one other then JEEB know anything that can help me
[20:43:34 CET] <JEEB> because I know you can push it
[20:43:47 CET] <JEEB> uhh, got it weird again
[20:43:56 CET] <JEEB> *I know ffmpeg can push things through to an HTTP thing
[20:44:19 CET] <E-TARD> i can't use ffmpeg on the server side
[20:44:27 CET] <E-TARD> i need ffserver
[20:45:22 CET] <E-TARD> ok then back to looking at lines and lines of code
[20:45:27 CET] <JEEB> ok, seems like you can listen with ffmpeg
[20:45:38 CET] <JEEB> so yeah, I would have one ffmpeg on the server listening and muxing all those three
[20:45:48 CET] <JEEB> and then another actual transcoder pushing stuff through http
[20:45:55 CET] <JEEB> to the muxer
[20:46:18 CET] <E-TARD> but ffmpeg has not config, ffmpeg will not push out mroe thenone streams from one feed
[20:46:23 CET] <JEEB> what
[20:46:45 CET] <E-TARD> does it have a config file
[20:47:04 CET] <JEEB> ffmpeg -i input -map 0 out.m3u8 -map 0 -f dash out.mpd -map 0 out.flv
[20:47:09 CET] <JEEB> there, one input, three outputs
[20:47:35 CET] <JEEB> you'd of course set more options, but that's the basic gist of it
[20:48:04 CET] <E-TARD> but are they listening
[20:48:23 CET] <JEEB> you'd use an actual httpd for that
[20:48:46 CET] <E-TARD> ok then how to i get my files from my pc to the server
[20:48:58 CET] <E-TARD> ftp
[20:49:18 CET] <prelude2004c> hEY GUYS.. GOOD DAY
[20:49:20 CET] <prelude2004c> oops..
[20:49:29 CET] <prelude2004c> using ffmpeg and vdpau . the M4000 does h264 decoding right ? i am using vdpau and i finally got it going but for some reason any source that is mpeg2video decodes fine with GPU but h264 content does not ? Any idea why that would be ?
[20:49:32 CET] <JEEB> you put one remuxer on the server listening for the transcoder, and push transcoded stuff via HTTP to the muxer
[20:49:57 CET] <JEEB> [ffmpeg transcoding one output stream] => [ffmpeg muxing those formats on the server (no transcoding)]
[20:50:27 CET] <E-TARD> blah there would be mad lag doing it that way
[20:50:36 CET] <JEEB> why?
[20:50:44 CET] <E-TARD> i think ffmpeg ftp to server would be faster
[20:50:49 CET] <JEEB> no
[20:50:52 CET] <E-TARD> this is a live feed
[20:51:06 CET] <E-TARD> i need low lag
[20:51:06 CET] <JEEB> this shouldn't have more lag than what you have between ffmpeg and ffserver :P
[20:51:26 CET] <JEEB> also you'd by now be noticing the lag that libavformat and libavcodec give by themselves
[20:51:44 CET] <JEEB> like the fact that libavformat/-codec have an analyzeduration of five seconds
[20:52:20 CET] <JEEB> anyways, not my job so have fun with whatever you have chosen
[20:52:47 CET] <E-TARD> at this point i wish i had $800 for evostream
[20:53:38 CET] <furq> 19:51:06 ( E-TARD) i need low lag
[20:53:40 CET] <furq> then don't use HLS
[20:53:51 CET] <E-TARD> that and i miss the days of flash and crtmpserver
[20:54:06 CET] <E-TARD> hls is for the iphone dumb shits
[20:54:22 CET] <E-TARD> i dont care soo much about them
[20:54:57 CET] <E-TARD> i just want to have it so i dont have to hear there emails anymore
[20:59:05 CET] <E-TARD> well i'm going to cut out thanks for helping JEEB. i will leve you all with this i feel ffserver is good (I like it for what i got it to do) but i feel more of the same settings found in ffmpeg need to be added to ffserver as well as better docs/tuts
[20:59:56 CET] <shincodex> None
[21:00:01 CET] <furq> did someone in here say that ffserver is a candidate for deprecation
[21:00:09 CET] <furq> that can't happen soon enough
[21:00:22 CET] <shincodex> I thought it was all a candidate for that
[21:00:23 CET] <shincodex> :O
[21:00:36 CET] <JEEB> I definitely would be for it, since so few people know what the flying fuck it does
[21:00:51 CET] <JEEB> and it keeps making people think they can use it as well as ffmpeg
[21:00:51 CET] <furq> i know what it does
[21:00:53 CET] <furq> it sucks
[21:01:10 CET] <JEEB> I think the worst part is that it "works" for some people
[21:01:27 CET] <JEEB> some people who have cut throats of goats on the right moon
[21:01:29 CET] <JEEB> etc
[21:01:53 CET] <furq> yeah i used to use it a bit and then it just stopped working completely after an upgrade
[21:02:20 CET] <furq> and by "working" i mean sort of working
[21:06:24 CET] <shincodex> i think the entertainment of this channel helps me over look those goat cutters
[21:07:07 CET] <furq> there's nothing wrong with a bit of ritual sacrifice
[21:07:31 CET] <furq> as long as it's for a good cause like divining a working configuration
[21:17:32 CET] <romano2k> hello there! is this a good place to ask for help with file transcoding?
[21:18:20 CET] <romano2k> i'll give it a try
[21:19:39 CET] <romano2k> i have a movie file with a FLAC audio track and 4 subtitles tracks. i'd like to copy video, transcode the audio track to AAC and keep only 2 of the subtitles tracks. is ffmpeg suited for this operation?
[21:19:59 CET] <furq> sure
[21:21:25 CET] <furq> https://ffmpeg.org/ffmpeg.html#Advanced-options
[21:21:31 CET] <romano2k> i've tried ffmpeg -i XXX -c:v copy -c:a aac -c:s copy XXX, but it kept only the first subtitle track
[21:21:58 CET] <furq> use -map for stream selection
[21:22:32 CET] <romano2k> so this is a normal behaviour not keeping subtitles with -c:s copy ?
[21:22:40 CET] <furq> i can't say i've ever tried
[21:22:46 CET] <furq> but -map 0:s:0 -map 0:s:1 should work
[21:22:56 CET] <furq> you may need to also explicitly map the audio and video tracks before that
[21:23:06 CET] <romano2k> ok
[21:24:31 CET] <JEEB> romano2k: ffmpeg by default selects "best" tracks in its opinion from all types and only uses those
[21:24:46 CET] <JEEB> you have to use map as furq noted to select more or less
[21:24:50 CET] <romano2k> JEEB: good to know, thank you
[21:25:02 CET] <JEEB> as soon as you put a map parameter into the command line it stops the automagic selection
[21:25:15 CET] <JEEB> so yes, you have to list all tracks you want to put into the output file
[21:25:22 CET] <furq> romano2k: you can also exclude tracks with -map 0 -map -0:s:2a
[21:25:28 CET] <furq> -map 0 -map -0:s:2
[21:27:28 CET] <romano2k> you mean "-map 0" or "-map -0:x:x"?
[21:27:42 CET] <romano2k> i'm reading the map help section, it's pretty new to me
[21:28:05 CET] <furq> both
[21:28:09 CET] <furq> -map 0 maps all streams
[21:28:19 CET] <romano2k> oh ok, i get it
[21:28:25 CET] <romano2k> include all then exclude some
[21:28:28 CET] <furq> right
[21:28:54 CET] <romano2k> i don't understand the 0:s:0 format
[21:29:08 CET] <romano2k> let me share the result of a ffprobe on my file
[21:30:07 CET] <romano2k> http://pastebin.com/Q89mwh40
[21:30:40 CET] <romano2k> there are also chapters which i didn't pasted in here
[21:30:54 CET] <furq> you can just use 0:4 and 0:5
[21:31:08 CET] <romano2k> that's what i was thinking, ok
[21:31:21 CET] <romano2k> and do i need to specify anything to keep the chapters?
[21:31:23 CET] <furq> 0:s:x refers to subtitle streams in particular
[21:32:43 CET] <furq> there's -map_chapters 0 but i think that's done by default
[21:33:22 CET] <romano2k> anyway, "ffmpeg -i XXX -c:v copy -c:a aac -b:a 256k XXX" returns "Unable to find a suitable output format for 'copy'"
[21:33:30 CET] <romano2k> is it a problem with my ffmpeg version?
[21:34:09 CET] <furq> what's the extension of the output file
[21:34:57 CET] <romano2k> .mkv
[21:35:05 CET] <furq> that's weird
[21:36:53 CET] <JEEB> full command line and terminal output of ffmpeg in a pastebin, kthx
[21:40:05 CET] <romano2k> JEEB furq: http://pastebin.com/ChEXLu05
[21:41:31 CET] <JEEB> with -v debug
[21:42:41 CET] <JEEB> I recommend 2> stuff.log
[21:42:43 CET] <JEEB> at the end
[21:42:48 CET] <JEEB> so it gets written into a file :P
[21:42:52 CET] <Prelude2004c_> hey guys.. can anyone give me help with vdpau ? i am totally stuck
[21:42:57 CET] <JEEB> and you don't have to copypasta from the windows cmd.exe
[21:44:40 CET] <romano2k> JEEB: i'll have to find how to output the result into a file, it's not ">" with PowerShell ;-)
[21:45:35 CET] <romano2k> weird, the same command works within cmd
[21:45:47 CET] <romano2k> i mean not only the ">", the whole command
[21:45:51 CET] <furq> fun
[21:46:07 CET] <shincodex> libswscale.v are these the version stamps
[21:46:09 CET] <furq> i'm surprised -c:a aac works without -strict -2
[21:46:16 CET] <romano2k> it doesn't
[21:46:16 CET] <shincodex> so if i delete the files like this they will get regenerated
[21:46:19 CET] <furq> oh ok
[21:46:25 CET] <furq> that does work now but it only changed this month
[21:46:41 CET] <furq> there's been a lot of work on the builtin aac encoder so you might want to upgrade anyway
[21:46:45 CET] <romano2k> but in PowerShell i have the weird error message pasted above, with cmd it gets past it
[21:47:07 CET] <furq> i was going to suggest PS because it looks like an option parsing error
[21:47:10 CET] <furq> and the options look fine to me
[21:47:24 CET] <romano2k> you mean that with the latest version, encoder aac doesn't require the extra parameters anymore?
[21:47:31 CET] <furq> i believe so
[21:47:33 CET] <furq> i've not tried it yet
[21:47:34 CET] <romano2k> cool
[21:47:52 CET] <furq> i already went to the trouble of compiling a version with fdk_aac
[21:48:03 CET] <furq> but apparently the builtin one is much better now than it used to be
[21:48:58 CET] <JEEB> it is
[21:49:07 CET] <furq> speaking of new features i should try out that fancy new deinterlace filter
[21:49:09 CET] <JEEB> atomnuker did a great job with that other guy :3
[21:49:15 CET] <JEEB> "new deinterlacer"?
[21:49:24 CET] <romano2k> yeah i just read the wiki, it puts "aac" above all others
[21:49:26 CET] <JEEB> also you should build your ffmpeg with zscale
[21:49:34 CET] <JEEB> or well, zimg
[21:49:40 CET] <furq> w3fdif
[21:49:49 CET] <JEEB> so you get the zscale resizer/colorspace conversion filter :)
[21:49:50 CET] <furq> someone was talking about it earlier
[21:49:53 CET] <JEEB> hmm
[21:50:24 CET] <furq> i've been meaning to get avisynth and QTGMC working but i sort of hate avisynth
[21:50:53 CET] <JEEB> I've moved all my transcoding that used avisynth to vapoursynth now vOv
[21:51:39 CET] <JEEB> also I'll have to poke w3fdif
[21:56:43 CET] <shincodex> visual studio has trouble importing folders with things like Makefile because its extension based.
[21:58:20 CET] <furq> i think there's another reason why visual studio has trouble with that
[22:02:21 CET] <shincodex> not sure who to be upset with in this case
[22:02:27 CET] <shincodex> Its import files from folder
[22:02:35 CET] <shincodex> seems to be very broken
[22:03:45 CET] <furq> you should be upset with yourself for using visual studio
[22:11:05 CET] <shincodex> its best ide
[22:11:10 CET] <shincodex> i died when i used code blocks
[22:11:21 CET] <shincodex> and it spun in infinite loop when derefence pointer
[22:11:24 CET] <shincodex> i was so sad.
[22:11:34 CET] <shincodex> but then again the pointer was AVFrame so
[22:27:29 CET] <jonascj> Is it realistic to have 4 hours of 720p video h264 encoded on a dvd (4.7GB)?
[22:27:47 CET] <furq> sure
[22:27:55 CET] <furq> whether or not it'll look any good depends on the content
[22:28:07 CET] <jonascj> It means a bitrate of 4.7*1024^2/(4*60*60) = 350 kbit/s
[22:28:27 CET] <furq> that's 350KB/s
[22:28:33 CET] <jonascj> KB/s, yes
[22:28:36 CET] <furq> 2611kbps
[22:29:14 CET] <furq> so probably about 2500 for the video
[22:29:20 CET] <furq> that's plausible
[22:30:09 CET] <jonascj> kbps, is that kilo bit per second (not byte)?
[22:30:14 CET] <furq> yes
[22:30:53 CET] <jonascj> is it always kbps that is in the thousands (1000-3000) when you see specs of modern day h264 video? I always thought it was KB/s
[22:31:05 CET] <jonascj> So 360KB/s is not so bad
[22:31:57 CET] <furq> i've seen 720p crf 20 encodes come out smaller than that
[22:32:03 CET] <furq> it depends on the source
[22:40:39 CET] <jonascj> The source is this http://paste.linuxassist.net/view/0a4c4446 - a 36GB 1080p output from iMovie on Mac (I didn't opt for iMovie, a friend did, now I am trying to compress it a bit).
[22:41:46 CET] <jonascj> First of all I was thinking to rescale it to 720p.
[22:48:52 CET] <prelude2004c> anyone here good with vdpau or can give me direction on how i should proceed ? CPu doesn't have hardware decoding built in so i am trying to use my GPU card... works if the source is mpeg2video but if the source is h264 it doesnt' use the GPU.. not sure why
[22:48:58 CET] <prelude2004c> looking for some guidance
[22:57:30 CET] <fritsch> prelude2004c: vdpauinfo | pastebinit
[22:57:32 CET] <fritsch> please
[23:02:52 CET] <prelude2004c> http://pastebin.com/QSqW7MYY
[23:03:40 CET] <fritsch> all fine
[23:03:45 CET] <fritsch> you talk of decoding, right?
[23:04:54 CET] <fritsch> use a player of your choice and playback the files
[23:06:30 CET] <furq> h264 baseline not supported?
[23:07:01 CET] <fritsch> H264_CONSTRAINED_BASELINE
[23:07:14 CET] <fritsch> not to be confused with h264_baseline
[23:07:22 CET] <furq> they both say not supported
[23:07:31 CET] <furq> also how can it support main and not constrained baseline
[23:08:25 CET] <fritsch> constrained baseline is a nice mixture of all of those a subset
[23:08:41 CET] <furq> i thought CBP was a subset of main
[23:08:51 CET] <fritsch> https://en.wikipedia.org/wiki/H.264/MPEG-4_AVC
[23:08:53 CET] <jonascj> When you see 190 and 320 "kbit/s" what is the unit they mean, is it "kilobits/s", i.e. it is 190/8 KB/s = 24 KB/s and 320/8 KB/S = 40 KB/s ?
[23:08:57 CET] <furq> yeah that's what i'm looking at
[23:08:58 CET] <fritsch> that's what they say
[23:09:09 CET] <furq> It corresponds to the subset of features that are in common between the Baseline, Main, and High Profiles.
[23:09:25 CET] <fritsch> so what do baseline, main and high have in common?
[23:09:49 CET] <fritsch> that's written quite funny
[23:09:59 CET] <fritsch> if you think of sets, the answer would be baseline
[23:10:09 CET] <fritsch> but that would make constrained baseline obsolte
[23:10:16 CET] <furq> look at the feature matrix at the bottom
[23:10:39 CET] <fritsch> nice
[23:10:47 CET] <furq> prelude2004c: run ffprobe on an h264 input file
[23:10:50 CET] <furq> and pastebin the result
[23:11:11 CET] <prelude2004c> yup doing that now
[23:11:15 CET] <prelude2004c> yes i am decoding the input one sec
[23:11:25 CET] <furq> granted if it is baseline i have no idea how you'd fix it
[23:12:15 CET] <prelude2004c> http://pastebin.com/Mddu11Yf
[23:12:31 CET] <prelude2004c> so basically its an h264 high input
[23:12:36 CET] <prelude2004c> i am transocing it using nvenc
[23:12:45 CET] <prelude2004c> the encoding is fine.. uses encoder on the card just fine
[23:12:53 CET] <prelude2004c> the issue is the cpu is not fast enough to decode and i get errors
[23:13:02 CET] <prelude2004c> so i want to offload to GPU decoder which is at 0% usage
[23:13:27 CET] <prelude2004c> if the source is SD and mpeg2video i see the decoder go up and be used for that source.. but the h264 for some reason never uses the decoder and prefers to use the CPU instead
[23:13:33 CET] <prelude2004c> i can't have that
[23:13:56 CET] <fritsch> do you do that from code? or via commandline?
[23:15:07 CET] <furq> prelude2004c: do you get any useful vdpau error messages with -v debug
[23:16:00 CET] <C0nundrum> JEEB are you still there ?
[23:17:04 CET] <prelude2004c> nothing really
[23:17:10 CET] <prelude2004c> i use this code.. let me get it for you
[23:17:31 CET] <Mavrik> Uh.
[23:17:46 CET] <Mavrik> Are you trying to use both decoding and encoding on the same card?
[23:17:51 CET] <prelude2004c> http://pastebin.com/RAchymRY
[23:17:58 CET] <prelude2004c> yup encoding and decoding on same card
[23:18:09 CET] <prelude2004c> the new cards have hardware decoders and cuda cores
[23:18:16 CET] <prelude2004c> so they work independantly.
[23:18:26 CET] <prelude2004c> again ,, when i use a source of mpeg2video it works perfectly as intended
[23:18:54 CET] <xintox> anyone hacking on neulion?
[23:19:05 CET] <xintox> join us in #h4kr
[23:19:08 CET] <prelude2004c> oh wait.. the pastbin i just posted shows ${ffmpeg} -hwaccel vdpau -i "$stream" -c:v h264 -pix_fmt yuv420p -c:a copy -map 0:p:$6 -f mpegts -  .. but the c:v should be mpeg2video
[23:19:15 CET] <jonascj> In the two-pass example here https://trac.ffmpeg.org/wiki/Encode/H.264 , how do the two ffmpeg instances communicate. The first one (pass 1) does not seem to produce an output file and something is going to /dev/null ...
[23:19:47 CET] <c_14> jonascj: there's an ffmpegpass-0.log or something
[23:19:52 CET] <c_14> it's created in the cwd
[23:19:52 CET] <prelude2004c> i convert to raw video and then taake that raw video and transcoding back to h264 .. no easier way to do it because if i just send the h264 to the encoder it has to decode too which uses more CPU
[23:21:18 CET] <C0nundrum> When recording an flv, if ffmpeg isn't closed properly, the file isn't playable and ffprobe returns http://pastebin.com/bgWwunma Is there anyway to avoid that or recover the file ?
[23:21:51 CET] <furq> that's an mp4, not an flv
[23:21:51 CET] <flux> c0nundrum, record to some other format that recovers better and convert it to flv afterwards?
[23:22:02 CET] <furq> an flv should be playable if ffmpeg aborts
[23:22:08 CET] <jonascj> c_14: sorry, I did not understand that, unless you are saying "-pass 1" instructs ffmpeg to pick a temp location and filename, write it to a log file, and "-pass 2" instructs ffmpeg to look for any recent pass 1's in the log...
[23:22:31 CET] <c_14> jonascj: it writes to a log file in cwd
[23:22:38 CET] <c_14> You can explicitly set the file with --passlogfile
[23:22:38 CET] <C0nundrum> the reason i don't do that is because multiple streams are being recorded and converting it while recording takes to much cpu time
[23:22:41 CET] <c_14> eh --
[23:23:02 CET] <C0nundrum> furq that what evryone says but it this is the problem i have one sec iwill post the ffprove of a completed file
[23:23:20 CET] <furq> 22:21:50 ( furq) that's an mp4, not an flv
[23:23:24 CET] <furq> does everyone say that as well
[23:23:44 CET] <jonascj> c_14: oh, Current Working Directory.
[23:24:07 CET] <jonascj> c_14: what about "-i input" is that a special keyword as well or should I actually specify myinputfile.mp4 in both commands?
[23:24:28 CET] <furq> you specify the input filename
[23:25:24 CET] <c_14> jonascj: actually specify the file in both commands
[23:25:36 CET] <jonascj> and now I get it, pass1 does not do anything except record statistics into the log file. That is why there is no output filename in the first command (except for /dev/null)
[23:26:28 CET] <prelude2004c> i want to send snapshot files through
[23:26:36 CET] <prelude2004c> thats the best way to do iT so you guys can see the 3 screen shots ?
[23:26:44 CET] <C0nundrum> Nice sarcasm lol, my mistake i meant is there anyway to avoid it with this filetype ? http://pastebin.com/tGhZsNSi
[23:28:24 CET] <furq> no
[23:28:36 CET] <furq> the moov atom isn't written until the encoding is done
[23:28:45 CET] <prelude2004c> http://www.megafileupload.com/e738/shot1.png & http://www.megafileupload.com/e739/shot2.png & http://www.megafileupload.com/e73a/shot3.png
[23:28:49 CET] <prelude2004c> see the 3 files
[23:29:00 CET] <furq> prelude2004c: imgur.com tia
[23:29:23 CET] <C0nundrum> furq so what are my options ?
[23:29:35 CET] <prelude2004c> scratch that
[23:29:40 CET] <furq> use a format which will work if the encoder breaks
[23:29:42 CET] <furq> such as flv
[23:30:32 CET] <prelude2004c> http://imgur.com/a/Af4Zq
[23:30:33 CET] <C0nundrum> Will that require re-encoding the input ?
[23:30:34 CET] <prelude2004c> ok see images there
[23:30:40 CET] <furq> C0nundrum: no
[23:31:11 CET] <prelude2004c> so the issue is the first one using 100% cpu
[23:32:17 CET] <C0nundrum> furq this is the format of the stream ffmpeg is saving http://pastebin.com/tNxudNrv
[23:34:53 CET] <MachinaeWolf> There's some slight discrepency when recording with ffmpeg -f alsa -i hw:0,0,0 file.wav when recording from mic line in
[23:35:12 CET] <MachinaeWolf> [alsa @ 0x6500e0] ALSA buffer xrun.
[23:35:15 CET] <MachinaeWolf> [alsa @ 0x6500e0] ALSA buffer xrun.trate=1497.0kbits/s
[23:35:47 CET] <MachinaeWolf> At one point in the song I'm recording from my 2ds running korg dsn12 it skipped a slight bit
[23:36:53 CET] <C0nundrum> furq Would i just do 'ffmpeg -i "'+link+'" -vcodec copy -acodec copy -bsf:a aac_adtstoasc "'+fileName+'.flv"'; ?
[23:43:37 CET] <jonascj> can I rescale at the same time I do the two pass compression?
[23:45:02 CET] <jonascj> or should I scale before running pass1? It seems to me that the stats will be wrong if it performs pass 1 on a file which is of the wrong resolution
[23:53:48 CET] <jonascj> I've found some examples with "-vf scale=-1:720" passed to both "-pass 1" and "-pass 2" commands. Is that what is needed?
[00:00:00 CET] --- Wed Dec 23 2015