[Ffmpeg-devel-irc] ffmpeg.log.20120131

[burek]

[01:39] <plut0> ugh ffmpeg crashed, out of memory
[01:39] <plut0> why
[01:44] <plut0> is there a way to limit the amount of memory used?
[01:46] <undercash> how can i log ffmpeg output?
[01:46] <undercash> i dont remenber ..
[01:47] <relaxed> plut0: I've never seen that issue before. run `help ulimit`
[01:48] <relaxed> undercash: ffmpeg ... 2>ffmpeg.log
[01:48] <undercash> ok
[01:48] <undercash> thx
[01:49] <plut0> relaxed: i'm dealing with a 22G file
[01:49] <undercash> my issue is i use a cmd like : ffmpeg -re -i "input.avi" -vcodec libx264 -preset fast -crf 28 -acodec copy -threads 8 -f flv rtmp://live.justin.tv/app/xxxx
[01:49] <undercash> and it works fine, and suddenly ffmpeg stop
[01:49] <plut0> relaxed: it wasn't writing any of the buffer to disk after almost 2 hours
[01:49] <plut0> relaxed: then it crashed
[01:49] <undercash> no idea what is happening
[01:52] <relaxed> undercash: look at loglevel in the man page
[01:52] <undercash> ok relaxed
[01:52] <relaxed> plut0: ok, I've encoded files larger then that without problem.
[01:53] <plut0> relaxed: so how should i troubleshoot?
[01:53] <relaxed> first you should try a recent version to see if it still happens.
[01:54] <plut0> relaxed: what else
[02:17] <plut0> anyone got ideas on how to troubleshoot ffmpeg running out of memory?
[02:26] <plut0> anyone?
[03:02] <teratorn> plut0: ffmpeg used from command-line?
[03:02] <plut0> teratorn: yes
[03:03] <teratorn> plut0: well I'm sure that shouldn't be happening
[03:03] <teratorn> plut0: what is your command line?
[03:04] <plut0> time ffmpeg -i 00011.m2ts -vcodec libx264 -crf 22 -preset veryfast -acodec copy -scodec copy -threads 0 00011.mkv -scodec copy -newsubtitle -scodec copy -newsubtitle
[03:06] <plut0> memory usage goes up and up until it's killed eventually
[03:07] <teratorn> plut0: I would guess it's a broken encoder, but you aren
[03:07] <teratorn> aren't encoding
[03:07] <teratorn> this is latest ffmpeg HEAD?
[03:07] <plut0> teratorn: libx264 is encoding right?
[03:07] <plut0> teratorn: 0.7.8
[03:07] <teratorn> isn't that old?
[03:07] <teratorn> oh, you don't have -vcodec copy
[03:07] <teratorn> OK
[03:08] <plut0> teratorn: i downgraded from 0.9.1 because of some bugs
[03:08] <teratorn> well, try compiling git HEAD from source
[03:08] <teratorn> and latest libx264 (and make sure you actually having ffmpeg /using/ latest x264)
[03:09] <teratorn> memory errors are often fixed in newer versions of code
[03:09] <plut0> i was looking for something stable
[03:09] <teratorn> yes?
[03:09] <plut0> you want me to try latest git though?
[03:09] <teratorn> definitely
[03:09] <teratorn> I know it's a pain, but there isn't any other way to debug it
[03:10] <teratorn> the other thing I would try is to cut about 1 second of that input file, then transcode it as you are doing, but run ffmpeg under valgrind
[03:10] <teratorn> with --leak-check=full
[03:11] <plut0> will try latest git
[03:11] <teratorn> (I am not 100% sure that ffmpeg actually cleans up after itself resulting in a clean valgrind run, but you can test it against other files and with different codecs, and see if there is something especially bad about x264)
[03:12] <teratorn> there isn't any solving the problem without finding the fixing the bug, unless you encode to multiple files - small segments each - then -vcodec copy them all together
[03:12] <teratorn> which may work without leaking
[04:35] <kode54> huh, I take it there's no way to transmux a file without applying the "encoder" metadata string
[04:50] <pasteeater> kode54: why do you want to get rid of that?
[04:51] <kode54> being pedantic for some video files I'm remuxing for the sole purpose of stripping off the metadata
[04:52] <pasteeater> you could re-mux it with MP4Box -add input.mp4 output.mp4
[06:42] <SIFTU> I'm trying to convert a 4:3 640x480 video to a video with 16:9 ratio (720x480). I have done this with help from the channel (pillarbox style) but the video looks stretched. Any ideas how I can get it too look normal again?
[06:43] <SIFTU> command line - ffmpeg -y -i 9.AVI -vcodec mpeg2video -aspect 16x9 -vf "pad=720:480:40:0:black" -acodec ac3 -ab 448k -threads 0 -qscale 3 -f dvd 9.mpg
[06:52] <relaxed> SIFTU: "pad=720:480:40:0:black,setdar=16:9"
[06:52] <relaxed> remove -aspect 16x9
[06:52] <SIFTU> relaxed: thanks, I was looking at dar with no idea of what it means
[06:52] <relaxed> display aspect ratio
[06:55] <SIFTU> relaxed: still seems stretched
[06:59] <relaxed> if the source is 4:3 then set the dar to that as well
[07:01] <SIFTU> I'm not sure I follow.. is that setsar?
[07:04] <relaxed> how does "-vf scale=720:480,setdar=4:3" look"
[07:04] <SIFTU> squished
[07:04] <SIFTU> oh hang on
[07:04] <SIFTU> didnt see the scale
[07:11] <SIFTU> relaxed: still not right.. hmm
[07:24] <zro> so i gotta put some video on a website thats primaryily going to be viewed on mobile devices (ie. phones) ... should i be converting these to 3gp ?
[07:42] <SIFTU> it seemed like -vf "pad=ih*16/9:ih:(ow-iw)/2:(oh-ih)/2" works
[07:42] <SIFTU> from the man page
[09:47] <lobstery> Hi, I'm running into this issue: https://ffmpeg.org/trac/ffmpeg/ticket/309 - libx264 complains width is not divisible by 2. Anyone know a way around this?
[09:49] <JEEB> well, there's an example command there that rounds your picture to mod2
[09:49] <JEEB> libx264 by default encodes 420 YCbCr
[09:49] <JEEB> which means that your picture has to be mod2
[09:49] <JEEB> in both width and height
[09:49] <JEEB> you can also encode in 444 YCbCr, but that will most probably not be what you want :P
[09:51] <lobstery> not sure I understood half of that - very new to ffmpeg and all this.
[09:52] <lobstery> what is the example command?
[09:52] <JEEB> https://ffmpeg.org/trac/ffmpeg/ticket/309#comment:3
[09:55] <lobstery> hm.. okay.. so what I want is to keep the height at 480 and then scale the width - so a 1270x720 becomes something like 854x480
[09:55] <lobstery> so changing that to scale="trunc(oh/a/2)*2:480" should work I believe
[09:56] <JEEB> most probably
[09:56] <JEEB> not pretty syntax, but derp
[10:00] <lobstery> that seemed to scale it to 272x480 :/
[10:01] <JEEB> lol
[10:01] <JEEB> prolly ow instead of oh
[10:01] <JEEB> lol
[10:02] <JEEB> oh wait
[10:02] <JEEB> derp
[10:02] <lobstery> yeah I think it should be oh
[10:04] <lobstery> what does the trunc method do exactly?
[10:04] <JEEB> that probably makes it an int, not sure tho
[10:05] <JEEB> +Round the value of expression @var{expr} towards zero to the nearest
[10:05] <JEEB> +integer. For example, "trunc(-1.5)" is "-1.0".
[10:06] <lobstery> ah
[10:18] <lobstery> scale="trunc(oh*a/2)*2:480"  <- ftw
[10:18] <lobstery> Thanks JEEB
[10:21] <JEEB> np
[12:08] <MikeW> Hey guys, if I have a byte array of mp3 audio stream data, is there a straightforward way to construct a new MP3 file using that data and then apply ID3/metadata to that file?
[12:09] <MikeW> or would ffmpeg be a bad choice for this and would a different library be a better approach?
[12:19] <vivienschilis> hi
[12:19] <vivienschilis> Can some explain me this bufsize
[12:19] <vivienschilis> someone*
[12:20] <vivienschilis> does setting the bufsize has an impact on hardware decoding? I think about like phones
[12:20] <JEEB> maxrate and bufsize come into play when there's a certain limit to how fast data can be pushed to the decoder
[12:21] <JEEB> (and then there's the level limitations of course)
[12:21] <JEEB> in most cases if you're making something to play from an SD card etc., unless the device really sucks you don't have to care much
[12:22] <JEEB> some devices on the other hand suck really hard if you don't abide to some obscure limitations that depend on the device (I remember some blackberries having this, and maybe pre-3GS iDevices)
[12:26] <vivienschilis> thanks
[12:26] <vivienschilis> My initial issue was that
[12:26] <JEEB> most obvious use cases where you need vbv maxrate and bufsize are disc media and network streaming
[12:27] <vivienschilis> my client wants to predict his streaming billing
[12:27] <vivienschilis> so he want's bounds on bitrates
[12:28] <vivienschilis> to kind of predict the video file size
[12:28] <vivienschilis> with a min/max bitrate
[12:29] <vivienschilis> but I am not sure its the right mode, CRF would produce generally smaller files
[12:29] <vivienschilis> don't you think?
[12:29] <JEEB> CRF + maxrate and bufsize is what I'd use for general streaming
[12:29] <vivienschilis> ok cool
[12:30] <vivienschilis> how should i set the bufsize?
[12:30] <JEEB> that depends on the client :P
[12:30] <JEEB> how much does the player buffer
[12:30] <vivienschilis> yeah that's my issue
[12:33] <vivienschilis> thanks JEEB
[12:34] <vivienschilis> if you though want to use the iPhone player for example
[12:34] <vivienschilis> it's not a good choice right? cause it's not streaming anymore
[12:34] <vivienschilis> it's http delivery
[12:35] <JEEB> well, there's still network speeds to be thought of :P
[12:35] <JEEB> so even if you pre-encode the file you usually want it to abide to some vbv limits
[12:36] <JEEB> so that you know that "if the player buffers Xkb and has a speed of at least Ykbps it will play without further buffering"
[12:37] <JEEB> with non-live http delivery of video you can of course let it be more lax, as the content isn't going anywhere, but it still gives a better experience for the user
[12:39] <leandrosansilva> Hello to all. I get the lastest ffmpegt source code and it broken my app because I'm using AVFormatParamers. I'm seing it was replaced by AVDictionary. Where can I find a table which "maps" the old avformat options to the dictionary?
[12:39] <leandrosansilva> for example, width, height?
[12:40] <vivienschilis> brb
[12:41] <leandrosansilva> The avio_set_interrupt_cb was also removed
[12:41] <leandrosansilva> is it a replacement for it?
[12:41] <leandrosansilva> ops is there*
[12:54] <vivienschilis> JEEB, do you have some references I could read about this buffering?
[12:54] <vivienschilis> to understand it better
[13:05] <vivienschilis> is there a sense to put a buffer size < maxrate ?
[13:07] <JEEB> maxrate is basically the average bitrate that is calculated over the bufsize amount of bits, so it's possible
[13:08] <JEEB> (the average bitrate that is never exceeded)
[13:08] <JEEB> s/bits/bytes/
[13:09] <JEEB> the smaller the bufsize is, the less freedom the encoder has to vary the bitrate
[13:09] <JEEB> since it has to always keep within the "maximum average bitrate within the bufsize"'s limits
[13:12] <vivienschilis> make sens
[13:15] <vivienschilis> so in general if I double maxrate I get a fast buffering of about 2 sec? is that a good practice?
[13:15] <JEEB> what
[13:16] <vivienschilis> on the player side
[13:16] <JEEB> what
[13:16] <JEEB> re-phrase your question
[13:16] <vivienschilis> ok
[13:20] <vivienschilis> sorry I doesn't mean anything as it depends on the bandwdith
[13:20] <vivienschilis> my question is what is the best practice to set the buffer size limit. my client as no requirement
[13:23] <vivienschilis> a few people told me to just double the maxrate but I didn't get a valid reason from them so i don't trust those kind of formulas
[13:26] <JEEB> oh, you mean bufsize = 2 * maxrate ?
[13:27] <vivienschilis> YES
[13:27] <vivienschilis> oups capital letters
[13:29] <JEEB> the general idea vbv is that you'd either have a specification to which you abide on both sides, but yeah -- not sure if iPhone's player or whatever has those public
[13:29] <JEEB> *-either
[13:30] <JEEB> generally maxrate=bufsize or so is sane, but I don't think bufsize = 2 * maxrate is a bad idea (unless the player doesn't buffer enough and then you end up with more buffering)
[13:30] <JEEB> if you have no specifications you'll just have to try and hope :P
[13:31] <vivienschilis> better maybe it's better having a two pass + bitrate tolerance?
[13:31] <JEEB> dunno how that is better :P
[13:32] <JEEB> you still have the same problem
[13:36] <vivienschilis> true
[13:44] <JEEB> vivienschilis, well -- having some kind of vbv will at least let you have control over how much data goes over at max. :P
[13:45] <vivienschilis> sure but I don't want to be constrained to set a buffer information on the player
[13:46] <JEEB> what
[13:46] <vivienschilis> you said I have to set the buffer size on the player as weel
[13:46] <vivienschilis> well*
[13:47] <JEEB> yes, or just abide to whatever default behavior it has
[13:47] <JEEB> otherwise you can't stream really :P
[13:48] <JEEB> vbv is needed for network streaming and so on, to be able to set a minimum network speed your stream can be watched with
[13:48] <JEEB> (or well, to make it sure that a person with a speed of Xkbps or more can watch it without hick-ups)
[13:57] <vivienschilis> They have their spec from their Sorenson software
[13:57] <vivienschilis> they have 1 pass VBR
[13:57] <vivienschilis> with a Data Rate...
[13:57] <vivienschilis> how is that possible
[14:00] <vivienschilis> with min and max in percent urggh!!
[14:32] <ultramantis> hello is it possible to remove or change these lines in the logfile to something diffrent or custom  frame=176115 fps= 49 q=27.0
[14:33] <Mavrik> they're hardcoded
[14:34] <ultramantis> can they be changed?
[14:34] <Mavrik> if you change them in the source
[14:34] <ultramantis> which file in the source would do that
[14:35] <Mavrik> grep over them and see?
[14:35] <ultramantis> grep over them?
[14:39] <iive> try to find the string "frame=" in the source files.
[14:40] <ultramantis> right i under stand that part but which source file exactly is there a specific one and would i just # it out
[15:16] <ultramantis> ok i think i figured it out
[15:16] <ultramantis> whats the proper way to make a -vstats_file
[15:16] <ultramantis> just add that line?
[15:25] <ultramantis> figured that otu
[15:25] <ultramantis> but im do i need to recompile n configure after i mess with the C file
[15:35] <LexSfX> is there a build of ffmpeg that has a "libx26410bit" or something like that for encoding yuv444p10 h264 video with ffmpeg CLI?
[15:35] <LexSfX> i thought it would work with libx264, but i was disappointedly surprised
[15:35] <LexSfX> libx264 doesn't even do yuv444p
[16:12] <evil_andy> I'm trying to run some code using the ffmpeg libs.  Every time I try to decode some frames and convert them, I get the following error: http://pastebin.ca/2108154
[16:13] <evil_andy> http://pastebin.ca/2108155 is updated with the output of ldd pertaining to the ffmpeg libs
[16:58] <Zeranoe> Could anyone tell me how FFmpeg uses GnuTLS? I see it's included now
[17:54] <peace_> guys but wtf ; ffprobe is bad => I have worked on output with awk but ... it's a pain
[17:54] <peace_> coudl someone take care of that ?
[17:55] <peace_> look at this http://paste.ubuntu.com/824067
[18:01] <saste> peace: read the manual and you'll find plenty output options
[18:07] <peace_> saste:  uou mean this http://ffmpeg.org/ffprobe.html
[18:08] <saste> peace_: same as man ffprobe
[18:09] <relaxed> peace_: it supports JSON output which is nice
[18:10] <peace_> i see
[18:17] <peace_> relaxed: is this correct? fprobe -print_format JSON ~/output.mov
[18:17] <peace_> it gives me an error
[18:17] <peace_> mah
[18:17] <peace_> ok -print_format json
[18:18] <relaxed> or, ffprobe -print_format JSON -show_streams ~/output.mov
[18:20] <peace_> ah damn... i hate fork i guess i have libav stuff
[18:20] <peace_> -.-
[18:20] <peace_> relaxed:
[18:20] <peace_> Copyright (c) 2007-2011 the Libav developers
[18:21] <peace_>   built on Jan 24 2012 07:22:02 with gcc 4.6.2
[18:21] <peace_> i guess it's not the standard one?
[18:21] <peace_> right ?
[18:21] <saste> peace_: yes the output changes have not been ported to libav
[18:21] <relaxed> correct, avprobe doesn't have JSON support
[18:22] <peace_> wtf who the hell is has done the fork now everytime i have to check this shit
[18:22] <\DSAFEW\> hey I figured out why ffmpeg was giving the file not found error for every rtmp address, networking wasn't compiled in
[18:30] <relaxed> peace_: my avprobe doesn't output the trailing comma
[18:30] <peace_> relaxed: i dunno but this fork is a pain i had my software broken for that
[18:31] <relaxed> what broke?
[18:31] <peace_> i did a service menu for dolphin to convert file with profiles
[18:32] <peace_> now i have to analyze every files to get the proper bitrate for video and audio to use on ffmpeg
[18:33] <relaxed> that tells me nothing
[18:33] <peace_> mydistro provide some sort fo ffmpeg
[18:33] <peace_> ffmpeg
[18:33] <peace_> ffmpeg version 0.8-4:0.8-1ubuntu1, Copyright (c) 2000-2011 the Libav developers
[18:34] <peace_> it's not ffmpeg i guess
[18:34] <peace_> This program is not developed anymore and is only provided for compatibility. Use avconv instead (see Changelog for the list of incompatible changes).
[18:35] <relaxed> again, what specific problem did it cause?
[18:35] <peace_> :) it doesn't provide the json stuff
[18:35] <peace_> that is the problem
[18:36] <relaxed> uh, didn't you just learn about that 5 minutes ago?
[18:37] <peace_> i have an old version of ffmpeg in another machine tht worked fine with my awk script
[18:37] <peace_> and .... now i see 2 version of ffmpeg
[18:37] <peace_> that doesn't respect the old style for output
[18:38] <peace_> so my stuff is broken
[18:39] <peace_> i have to check if the user uses ffmpeg standard of he use the avstuff
[18:39] <peace_> bah
[18:51] <relaxed> it should be trivial to add another awk gsub for trailing commas
[18:51] <\DSAFEW\> I'm trying to capture the sound from my mic and my normal computer sounds with ffmpeg and alsa, any help would be appreciated
[18:52] <\DSAFEW\> right now I can only capture my mic
[19:09] <peace_> relaxed: :)
[19:09] <peace_> awk  '
[19:09] <peace_> /Video:/{$4 gsub(/\,/,"");gsub(/\(Simple Profile\)/,"");gsub( /\[[^\)]*]/, "" ) ;videocodec = $4 ; pixelformat = $5 ; resolution= $6 ;videobitrate = $7; fps = $9 }
[19:09] <peace_> /Duration:/{$4 gsub(/\:/," ") ;realseconds= $4; seconds = int($4); minutes= $3 ;hours= $2;totalsoundtrack = (seconds+(minutes*60)+(hours*3600)); totalbitrate = $8 }
[19:09] <peace_> /Audio:/{$4 gsub(/\,/,"");gsub(/\(Simple Profile\)/,"");gsub( /\[[^\)]*]/, "" ); audiocodec =$4; freq= $5 ; channel =$7 ; abitrate= $9} END {print "Video Codec: " videocodec ;print "Audio Codec: " audiocodec;print "Video Bitrate: " videobitrate ; print "Audio Bitrate: " abitrate;print "Total Bitrate: " totalbitrate ;print "Resolution: " resolution ; print "Fps: " fps ;   print "Frequency: " freq ; print "Channel: " channel;print "Pixel
[19:09] <peace_> Format: " pixelformat ;print "Total Duration: " hours ":" minutes ":" realseconds ;print "Total seconds " totalsoundtrack } '
[19:09] <peace_> ops
[19:09] <peace_> :D
[19:09] <peace_> sorry
[19:12] <peace_> relaxed:
[19:12] <peace_> awk  '
[19:12] <peace_> /Video:/{$4 gsub(/\,/,"");gsub(/\(Simple Profile\)/,"");gsub( /\[[^\)]*]/, "" ) ;videocodec = $4 ; pixelformat = $5 ; resolution= $6 ;videobitrate = $7; fps = $9 }
[19:12] <peace_> /Duration:/{$4 gsub(/\:/," ") ;realseconds= $4; seconds = int($4); minutes= $3 ;hours= $2;totalsoundtrack = (seconds+(minutes*60)+(hours*3600)); totalbitrate = $8 }
[19:12] <peace_> /Audio:/{$4 gsub(/\,/,"");gsub(/\(Simple Profile\)/,"");gsub( /\[[^\)]*]/, "" ); audiocodec =$4; freq= $5 ; channel =$7 ; abitrate= $9} END {print "Video Codec: " videocodec ;print "Audio Codec: " audiocodec;print "Video Bitrate: " videobitrate ; print "Audio Bitrate: " abitrate;print "Total Bitrate: " totalbitrate ;print "Resolution: " resolution ; print "Fps: " fps ;   print "Frequency: " freq ; print "Channel: " channel;print "Pixel
[19:12] <peace_> Format: " pixelformat ;print "Total Duration: " hours ":" minutes ":" realseconds ;print "Total seconds " totalsoundtrack } '
[19:12] <peace_> -.-
[19:12] <peace_> =>>>http://paste.ubuntu.com/824175/
[19:21] <relaxed> you just awked all over my channel
[19:22] <peace_> relaxed: :D sorry pasted for mistake on channel
[19:26] <relaxed> sub(/,$/,"")
[20:33] <LiroXIV> Okay, I'm trying to build a little VB-based FFmpeg front end for making WebM videos. I downloaded some 64-bit binaries of FFmpeg, and hooked it into my program
[20:34] <LiroXIV> But, when I try to convert an OGV video to WebM, it gives me an "ffmpeg.exe has stoped working" error. We're running Windows 7 SP1 x64
[22:42] <hachi> hello folks
[22:44] <hachi> does anyone happen to know what raw video format I might want to use for output if my input file format is still image jpegs using image2? I'd like to start assembling them into containers so I can address each frame rather than having to have a directory with a few thousand files in it :)
[22:47] <hachi> hmm, perhaps my best bet is to simply request a container and say -vcodec copy... I don't really know :)
[22:48] <burek> use -f mjpeg
[22:55] <hachi> right... so does mjpeg hold the same data as a jpeg image input for each frame?
[23:02] <hachi> can I cycle an mjpeg encoded stream back out to jpeg files on disk using no recompression?
[23:02] <hachi> with ffmpeg I mean
[23:03] <relaxed> yes, use -vcodec ocpy
[23:03] <relaxed> copy*
[23:04] <hachi> I'm not getting an idempotent result
[23:04] <hachi> if I input one file into an avi with copy, and then cycle it back out
[23:04] <hachi> I get four images
[23:05] <hachi> looking at what they contain now though
[23:05] <hachi> first two are damaged... or not really images
[23:06] <hachi> third is the exif thumbnail, I guess... it's a tiny copy of the image
[23:06] <hachi> fourth one is possibly the original, but corrupted
[23:07] <hachi> ffmpeg -i 104MEDIA/FILE0001.JPG -vcodec copy test2.avi
[23:07] <hachi> ffmpeg -i test2.avi -f image2 -vcodec copy foo%04d.jpg
[23:07] <hachi> should that be giving me an idempotent result... in theory at least?
[23:07] <relaxed> ffmpeg -i 104MEDIA/FILE0001.JPG -vframes 1 -vcodec copy -f mjpeg test2.avi
[23:09] <hachi> when I try to do the output from that, now I get an MP3 codec error :)
[23:10] <hachi> http://nopaste.snit.ch/113720
[23:11] <hachi> well... hum
[23:11] <hachi> debian appears to have some ancient ffmpeg
[23:13] <burek> :)
[23:13] <burek> just compile it
[23:13] <hachi> I'm gonna assume that could be part of the problem...
[23:13] <burek> git clone ..
[23:13] <hachi> yeah :)
[23:13] <burek> and compile
[23:13] <pozican> not surprising tbh
[23:13] <hachi> amazingly, this is a step up from some other version I was attempting this with
[23:14] <burek> you can download static binary
[23:14] <burek> just to try that
[23:14] <burek> it will be lot faster
[23:14] <hachi> is an OSX binary available by chance?
[23:14] <hachi> I didn't see one
[23:15] <burek> I'm not on that user account where I bookmarked that page, you can ask relaxed for a link
[23:15] <hachi> meh, I'll just compile it... it's not exactly hard
[23:15] <hachi> thank you both for the input.... on inputs
[23:15] <burek> ok :) this way it would be download and run
[23:15] <burek> but ok
[23:16] <burek> https://sites.google.com/site/linuxencoding/builds
[23:17] <burek> just to try something
[23:17] <burek> oukey
[23:24] <relaxed> hachi: omit '-f mjpeg', just use -vcodec copy for lossless conversion back and forth
[00:00] --- Wed Feb  1 2012