[Ffmpeg-devel-irc] ffmpeg.log.20120325

[burek]

[00:02] <__vincent> why does -acodec copy doesn't work anymore?
[00:02] <__vincent> seemed to work until i updated ffmpeg
[00:05] <ubitux> what do you mean by "doesn't work"?
[00:06] <__vincent> i get >Unknown decoder 'copy'
[00:08] <ubitux> can't reproduce here, can you share your command line and full output on pastebin?
[00:10] <__vincent> http://pastebin.com/PLrdYwCi
[00:10] <Mavrik> __vincent: you changed parameter order.
[00:10] <Mavrik> -acodec goes after -i.
[00:12] <__vincent> it worked
[00:12] <__vincent> interesting
[00:12] <__vincent> ive used that command for years, always worked in that order until now
[00:22] <Mavrik> doubt it.
[07:10] <oliv3> hi, can anyone give a hand on encoding a raw stream of PPM images to a streamable content (http/rtmp) ? I basically play around with -f image2pipe, -i pipe: , -re , -vcodec ppm and -f [mpeg|mpegts|h264|flv] - as output; streaming "kinda" works, but sometimes I get encoding errors (like images being shifted on the right, with an offset increasing randomly); if anyone does things like that please feel free to contact me :)
[08:57] <eightfold> is it possible to cut video without transcoding with ffmpeg?
[08:58] <eightfold> does ffmpeg -sameq -ss [start_seconds] -t [duration_seconds] -i [input_file] [outputfile]
[08:58] <eightfold> do this?
[09:27] <grepper> I think the codec would need to be one that supports that, like mpeg-2, using "copy" not -sameq
[09:30] <Boon> i tried to encode video for iphone but no success, tested with preset or without preset http://pastebin.com/FxvvhydT
[09:31] <Boon> any solution
[09:37] <eightfold> ffmpeg -ss 00:17:44.0 -t 00:22:05.5 -i input.flv -acodec copy -vcodec copy -async 1 output.mp4
[09:37] <eightfold> perhaps?
[09:37] <eightfold> but it doesn't seem to understand -t
[09:40] <Boon> -t for?
[09:43] <eightfold> Boon: i thought it was end time?
[09:43] <eightfold> ahhh
[09:43] <eightfold> duration
[09:44] <Boon> -s =  size
[09:46] <eightfold> Boon: i don't want to change size, i don't want to transcode, but i'd like to specify the end time of the output clip by saying stop at 00:22:05.5 in the input file
[09:48] <Boon> ffmpeg -y -i input.avi -acodec aac -ar 48000 -ab 128k -ac 2 -s 1024×768 -vcodec libx264 -b 1200k -flags +loop+mv4 -cmp 256 -partitions +parti4x4+partp8x8+partb8x8 -subq 7 -trellis 1 -refs 5 -coder 0 -me_range 16 -keyint_min 25 -sc_threshold 40 -i_qfactor 0.71 -bt 1200k -maxrate 1200k -bufsize 1200k -rc_eq 'blurCplx^(1-qComp)' -qcomp 0.6 -qmin 10 -qmax 51 -qdiff 4 -level 30 -aspect 16:9 -r 30 -g 90 -async 2 output.mp4
[09:48] <Boon> i think this command fixed the problem
[09:48] <Boon> :)
[10:55] <floater> GMplayer Playing ./Cowboyok.jpg.
[10:55] <floater> Seek failed
[10:55] <floater> libavformat file format detected.
[10:55] <floater> lold
[11:18] <nokiamaster> .m2ts files with multiple audio tracks. How do you choose tracks with ffmpeg cmd line. I have googled and nothing gives correct syntax.
[12:00] <cbreak> nokiamaster: how about -ac:1 or -codec:a:1?
[12:01] <cbreak> search for Stream Specifiers in the man page
[12:25] <igmrlm> u guys see the whitepaper for the gtx680, specifically the nvenc D:
[12:26] <igmrlm> claims to get 8x realtime encoding speed for 1080p h.264
[12:30] <igmrlm> http://www.geforce.com/Active/en_US/en_US/pdf/GeForce-GTX-680-Whitepaper-FINAL.pdf
[12:38] <Mavrik> I've yet to see hardware encoding coming close to x264 quality-wise
[12:38] <Mavrik> so let's be sceptical until we actually see at least a single encoding
[16:00] <anddam> hello
[16:01] <anddam> I'm getting this error when copying from .aac to mp4 with -absf aac_adtstoasc
[16:01] <anddam> (copying actual error)
[16:03] <anddam> [ipod @ 0x7f836403fc00] Application provided invalid, non monotonically increasing dts to muxer in stream 0: 6710272 >= 6710272
[16:05] <anddam> is there a way to "fix" the source aac stream?
[17:04] <sprzybilla> does the configure script for ffmpeg look for the libraries according to my environment? I am getting a libx264 not found error but have the library configured..
[17:12] <deke> could anyone help me w/ converting a .wmv to a nook compliant mp4? i use cyanogenmod 7.1 on my nook... afaik it'd be the same
[18:24] <sprzybilla> does anyone know of a guide that better describes the cropping video filter? I can't seem to wrap my head around the docs
[20:44] <_polto_> hi
[20:44] <_polto_> http://2012.rmll.info/en/participate/call-for-papers - would be nice to see FFMPEG at Geneva ;)
[20:46] <sprzybilla> does the latest build of ffmpeg not come with "verylong", "veryshort", first pass, etc. presets?
[20:48] <JEEB> you don't need separate files any more
[20:48] <JEEB> you pass profile names straight into libx264
[20:48] <JEEB> http://mewiki.project357.com/wiki/X264_Settings#preset
[20:48] <JEEB> and as for "fast first pass", it gets automatically set if you set that you are doing a first pass
[20:49] <JEEB> (as in, explicitly set whatever is the equivalent of --pass 1 in x264)
[20:49] <sprzybilla> JEEB: Ahhhhhh, thanks, I missed the memo, haven't updated in ages and everything is different!
[20:54] <sprzybilla> JEEB: if I want to use veryslow for both pass 1 and two, how can I expicity state that (from FFPEG) -pass 1 --preset-veryslow shpouts an error
[20:56] <JEEB> sprzybilla, --pass 1 is x264 syntax (ffmpeg mostly uses only one dash with options, thus it was more like "try to find out whatever is the equivalent to this in x264 in ffmpeg")
[20:56] <JEEB> -preset:v veryslow
[20:56] <JEEB> ^ sets video preset to veryslow
[20:58] <sprzybilla> JEEB: ahh - got it, I have always been a bit spacey when it comes to ffmpeg opts vs x264
[20:58] <sprzybilla> ty
[20:59] <leontopod> I am trying to convert ogg to mp3
[20:59] <leontopod> ffmpeg -i jupiter.ogg -acodec libmp3lame jupiter.mp3
[21:00] <leontopod> doesn't work
[21:00] <leontopod> says libmp3lame not found
[21:00] <leontopod> even though I have lame installed
[21:00] <leontopod> what am I doing wrong?
[21:01] <sacarasc> Did you compile ffmpeg with libmp3lame support?
[21:02] <leontopod> oh, no didn't =/
[21:02] <leontopod> I think this ffmpeg is what came with Slackware 11
[21:02] <Tim-Work> leon: confirm that it is installed by running "ffmpeg -formats -codecs"
[21:04] <leontopod>  DE mp3             MPEG audio layer 3
[21:05] <leontopod> apparently it is
[21:05] <leontopod> but how do I invoke it?
[21:05] <sacarasc> That's probably the mp3 container, not the codec.
[21:06] <leontopod> ok so I need to compile ffmpeg from scratch, including the libmp3lame encoder?
[21:06] <JEEB> yes
[21:06] <leontopod> ok will do
[21:06] <leontopod> thanks
[21:16] <grepper> leontopod: just a note that running "ffmpeg -formats -codecs" as written will only use/show "-formats", you would need to do "ffmpeg -codecs" as well
[21:18] <leontopod> oh
[21:18] <leontopod> let me try that thanks
[21:19] <sprzybilla> is there an x264 preset that is even more exhaustive than veryslow? It seems that the parametes specified in that preset aren't as exhuastive as possible? (maybe I should take this to the other irc)
[21:19] <leontopod> interesting
[21:19] <leontopod> it says it is there
[21:19] <leontopod> audio codec for mp3
[21:19] <leontopod>  D A    mp3             MP3 (MPEG audio layer 3)
[21:19] <leontopod> so what got compiled in?
[21:20] <Tim-Work> leon: you also need the E to encode to MP3
[21:20] <leontopod> drat
[21:20] <grepper> should show libmp3lame
[21:20] <leontopod> double drat
[21:20] <leontopod> looks like I will have to compile from scratch =/
[21:20] <leontopod> I do have lame installed
[21:20] <leontopod> but I guess that's not enough
[21:20] <Tim-Work> the good news is once you have the environment all setup it only take about 3 minutes to full compile
[21:20] <grepper> doesn't use the binary
[21:21] <leontopod> grepper, it's what came with Slackware 11
[21:21] <leontopod> or 12
[21:21] <leontopod> err
[21:22] <leontopod> PARDON ME
[21:22] <leontopod> 48 => cat slackware-version
[21:22] <leontopod> Slackware 13.1.0
[21:22] <leontopod> 13.1!
[21:22] <leontopod> =)
[22:52] <anddam> I experienced the same error as http://ffmpeg.org/trac/ffmpeg/ticket/222 "non monotonically increasing&"
[22:53] <anddam> I noticed that using a ffmpeg version built with gcc rather than clang make ffmpeg works while remuxing the same file
[22:54] <anddam> thought it was worth mentioning but the ticket says the issue is solved
[23:15] <hi117> E5=B09 < i can read japanese in russian
[23:16] <hi117> er wrong channel
[23:16] <hi117> >.<
[23:45] <xenome> hi, anyone know why I can't receive a udp stream unless I use multicast and use a URL like udp://@224.0.0.1:6666.... if I use udp://my.ip:6666 i get nothing
[23:47] <xenome> i'm not sure I understand why the @ needs to be there
[00:00] --- Mon Mar 26 2012