[Ffmpeg-devel-irc] ffmpeg.log.20130911

[burek]

[00:35] <hans_henrik> can i use ffmpeg to check if a file is a valid jpeg/png/whatever file?
[05:33] <i336_> Hi everyone. I'm want to build FFmpeg as an output decoder for MPD. I only need to be able to play MP3/WAV/OGG/MOD(?) files so I probably don't need the other 4294967296 things FFmpeg has; what do you recommend I disable?
[05:34] <i336_> s/other \([^ ]*\)/\1 other/
[05:35] <klaxa> wow i don't speak sed that well
[05:35] <i336_> oh lol
[05:36] <klaxa> does ffmpeg even support MOD?
[05:36] Action: i336_ has no idea
[05:36] Action: i336_ will be using mikmod probbaly
[05:36] <i336_> probably*
[05:40] <klaxa> actually, if you just do ./configure && make it should be able to decode mp3 wav (pcm) and ogg (vorbis)
[05:40] <klaxa> most features are disabled by default
[05:44] <i336_> oh ok
[05:44] Action: i336_ wonders what DCT, DWT, LZP, LZO, MDCT, RDFT and FFT are
[05:44] Action: i336_ wonders what DXVA2, VAAPI, VDA and VDPAU are
[05:46] Action: i336_ will definitely be enabling --enable-libcaca :P
[05:47] <i336_> next question.
[05:47] <i336_> I downloaded an mplayer package for slackware and whenever I try to use it I get "Illegal instruction".
[05:47] <i336_> It works fine if I just want help, or provide no arguments; this happens when I try to actually *play* something (an FLV video).
[05:49] <i336_> I'm thinking this is because it uses optimizations that don't work on my CPU (an old AMD Duron).
[05:49] <klaxa> hmm yeah maybe
[05:49] <klaxa> compile from source maybe?
[05:49] <i336_> probably will be :P
[05:50] <i336_> my question is, do I have to disable optimizations that won't work on my CPU arch the same way I have to be careful not to use bad MARCH flags, or are the optimizations automatic?
[05:50] <i336_> in that, they disable if they won't work?
[05:51] <klaxa> autoconf should detect what CPU extentions are available and which are not
[05:51] <klaxa> that's pretty much the purpose of autoconf
[05:51] <i336_> ...ah :S
[05:51] <i336_> I'm compiling on one PC, but running on another
[05:52] <klaxa> oh
[05:52] <klaxa> in that case you have to pay attention
[05:52] <i336_> hmmmm
[05:52] <Plorkyeran> ffmpeg does not use autoconf
[05:52] <i336_> *this* is an AMD Duron; the laptop I'm compiling it for (a kind of "mini jukebox" type thing) is actually an AMD K6-2
[05:52] <i336_> ooh, ok
[05:53] <klaxa> wait... using ./configure is not part of autoconf?
[05:53] <Plorkyeran> no
[05:53] <Plorkyeran> it is a hand-written configure script
[05:53] <i336_> "configure" is a script name
[05:53] <klaxa> my world is breaking apart right now
[05:53] <i336_> awesome :>
[05:53] <i336_> elvis' "./configure" is auto handwritten
[05:53] <Plorkyeran> autoconf is a tool for generating configure scripts
[05:53] <klaxa> i see
[05:53] <Plorkyeran> if you're using a premade package it may have been built with runtime cpu detection disabled for some reason
[05:54] <i336_> actually it's enabled
[05:54] <i336_> I think
[05:54] <i336_> 3 lines:
[05:54] <i336_> CPU: AMD Duron(tm) (Family: 6, Model: 3, Stepping: 1)
[05:54] <i336_> CPUflags:  MMX: 1 MMX2: 1 3DNow: 1 3DNow2: 1 SSE: 0 SSE2: 0
[05:54] <i336_> Compiled for x86 CPU with extensions: MMX MMX2 3DNow 3DNowEx SSE
[05:54] <i336_> ^ running mplayer w/ no args
[05:54] Action: i336_ haz idea
[05:56] <i336_> interesting
[05:56] <i336_> I ran it through strace and it crashed after loading fonts
[05:57] <i336_> renamed the font dir out and got 234589623498527642344509234058923452901209814352 "[flv @ 0x86807a0]Unsupported video codec (7)" messages
[05:57] <i336_> lol
[05:57] <klaxa> looks like it wasn't build with unfree decoders?
[05:57] <i336_> heh
[05:58] <i336_> MPlayer crashed by an 'Illegal Instruction'. It usually happens when you run it on a CPU different than the one it was compiled/optimized for. Verify this!
[05:58] <i336_> XD
[05:58] <i336_> probably an instruction issue. it's okay.
[06:36] Action: i336_ runs "bash-4.2$ ./configure --disable-dxva2 --disable-vaapi --disable-vdpau --disable-doc --disable-everything --enable-outdev=alsa --enable-outdev=oss --disable-iconv --enable-lto --enable-hardcoded-tables --enable-shared --enable-decoder={mp3,mp3adu,mp3adufloat,mp3float,aac,theora} --enable-demuxer={aac,mp3,wav,ogg} --enable-parser={aac,aac_latm,mpegaudio,vorbis} --enable-protocol={file,tcp,pipe}" and hopes it works
[06:37] <i336_> still mysteriously thinking :P
[06:37] <i336_> O.o
[06:37] <i336_> s/bash-4\.2\$ //
[06:38] Action: i336_ writes a progress bar for ./configure
[06:38] <i336_> apparently it needs one o.o
[06:39] <i336_> ooh
[06:39] <i336_> *output*
[06:40] <i336_> hm.
[06:41] <i336_> "*various CPU flags* enabled        yes"
[06:41] Action: i336_ tries to figure out which ones will run and which won't
[06:41] <i336_> (on an AMD K6-2)
[08:08] <i336_> meep
[08:08] <i336_> I left ffmpeg building while I went out for a little while figuring it would take a while
[08:08] <i336_> I was wondering how broken it'd be when I got back :P
[08:09] <i336_> here's the whole scrollback of the compile (warning, 121KB): http://pastebin.com/wXzy7K0L
[08:09] <i336_> basically I'm getting a bunch of
[08:09] <i336_> `ff_bgr2UVOffset' referenced in section `.text' of /tmp/ccGFqtOq.ltrans9.ltrans.o: defined in discarded section `.text' of libswscale/x86/swscale.o (symbol from plugin)
[08:09] <i336_> hm
[08:10] Action: i336_ tries disabling the stripper
[08:14] <i336_> hrmmmmm
[08:14] <i336_>   --enable-lto             use link-time optimization
[08:14] Action: i336_ wonders what happens when you use that
[08:14] Action: i336_ tries removing it
[08:14] <i336_> :P
[09:12] <CoJaBo> While this bitrate and its obvious implication on my free disk space is appealing, will this cause a black hole if I leave it run?
[09:12] <CoJaBo> frame=  101 fps=  5 q=0.0 size=      -0kB time=4.04 bitrate=  -2.0kbits/s
[09:48] <i336_> agh
[09:48] <i336_> I keep getting compile errors
[09:48] <spaam> what kind of errors?
[09:48] <i336_> I've probably enabled some wrong combination of options
[09:48] <i336_> libavformat/libavformat.so: undefined reference to `ff_guess_image2_codec'
[09:49] <i336_> that's the entire error
[09:49] <i336_> *grabs configure string*
[09:49] <i336_> ../configure --disable-dxva2 --disable-vaapi --disable-vdpau --disable-doc --disable-everything --enable-outdev=alsa --enable-outdev=oss --disable-iconv --enable-hardcoded-tables --enable-shared --enable-decoder={mp3,mp3adu,mp3adufloat,mp3float,aac,theora} --enable-demuxer={aac,mp3,wav,ogg} --enable-parser={aac,aac_latm,mpegaudio,vorbis} --enable-protocol={file,tcp,pipe} --disable-amd3dnowext --disable-mmxext --disable-sse{,2,3,4}
[09:49] Action: i336_ probably did something wrong
[09:49] <i336_> I'm just building it (..!) as an output plugin for MPD.
[09:52] <t4nk931> hi all
[09:53] <t4nk931> does anyone have idea about the code :get_packet_send_clock() from ffserver.c
[09:53] <t4nk931> can anyone explain the code?
[09:57] <t4nk931> is ther any tolerance about rtp send packet when Audio + video in ffserver
[09:59] <t4nk931> because I see the code in ffserver.c function http_send_data() it says:
[09:59] <t4nk931> if((get_packet_send_clock(c) - get_server_clock(c)) > 0{ 	  /* nothing to send yet: we can wait */                     return 0; }
[10:00] <t4nk931> why there is no any tolerance if the frame duration is bigger than current server clock?
[10:23] <t4nk931> does anyone have idea about the time check in ffserver?
[10:24] <t4nk931> which criterion used by ffserver to decide send the data or not/
[12:33] <izemize> hy
[12:34] <izemize> i use vlc media player for stream webcam to rtmp server in h264, but im get "late buffer for mux caching" error. This vlc or ffmpeg problem? C:\Users\User\vlc>vlc -I dummy -vvv dshow:// :dshow-vdev="Logitech HD Pro Webcam C920" :dshow-size="320x240" --sout "#transcode{vcodec=h264}:standard{access=rtmp,mux=ffmpeg{mux=flv},dst=rtmp://192.168.56.10/deejay/1}" --sout-mux-caching="30"
[12:59] <Zaga> hi ppl
[12:59] <Zaga> has anyone been able to put ffserver working with ffmeg out-of-the-box?
[12:59] <Zaga> it always says to me connection reset by peer
[14:06] <IFN10EBrain> I unmerged ffmpeg to have xbmc pull it in. I hoped xbmc would  set the right use flags for it to work with, but the flags I  see in emerge -pv xbmc seem to be exactly the same as when  ffmpeg was installed. xbmc works fine but the streaming of  content is very glitchy and I think this is because of  ffmpeg. Any pointers?
[14:47] <lentferj> when trying to compile current git master I get an error with xvid. This didn't happen back in July
[14:47] <lentferj> libavcodec/libxvid.c:362: error: .xvid_plugin_lumimasking_t. undeclared (first use in this function)
[14:47] <lentferj> libavcodec/libxvid.c:539: error: .masking_l. undeclared (first use in this function)
[14:47] <lentferj> libavcodec/libxvid.c:550: error: .masking_v. undeclared (first use in this function)
[15:00] <durandal_1707> lentferj: what xvid version?
[15:07] <durandal_1707> IFN10EBrain: what ffmpeg version? note that this is not xbmc channel
[15:46] <IFN10EBrain> durandal_1707: it's ffmpeg-7.0.1
[15:47] <durandal_1707> IFN10EBrain: there is no such ffmpeg (yet)
[15:47] <durandal_1707> latest is 2.0.1
[15:48] <IFN10EBrain> pardon me :) it's ffmpeg-1.0.7
[15:49] <durandal_1707> and it worked fine before?
[15:51] <IFN10EBrain> ffmpeg works fine. and xbmc too on this box with mint installed. now I have gentoo and the stream is very quirky.
[15:52] <IFN10EBrain> Are there any options that ffmpeg would need to meet with xbmc's requirements?
[15:52] <durandal_1707> dunno, better ask xbmc folks
[15:53] <IFN10EBrain> Guess so. Thanks anyhow.
[15:53] <lentferj> durandal_1707: libxvidcore4                        2:1.2.2-0.1
[15:54] <lentferj> durandal_1707: this is debian squeeze with deb-multimedia.org packages
[15:58] <durandal_1707> i dunno why that happens headers should be there...
[16:41] <roger21> hey
[16:42] <roger21> if i want to swith the order of to audio stream, i just add -map 0:a:1 -map 0:a:0 ?
[16:42] <roger21> two*
[16:50] <vl4kn0> Hi, is there a way to get average luminance of a frame in libav?
[16:57] <JoeyJoeJo> I've got a quick question about compressed audio files. Say I have a 1 minute long MP3 that's just silence. Is that file the same size as one that's not silent at the same bit rate?
[16:57] <JoeyJoeJo> Sorry for being a little off topic. I just don't know of a better place to ask this sort of question
[16:58] <Mavrik> now that's a very good question O.o
[16:59] <JoeyJoeJo> My thought is that filesize = bitrate * length of file, so it shouldn't matter if the file is just silence or full of sound
[17:00] <Mavrik> JoeyJoeJo, the catch is that encoders for some formats will not use full bitrate if they have nothing to store
[17:00] <JoeyJoeJo> Ah, ok
[17:00] <Mavrik> but honestly, for mp3 I really don't know
[17:01] <JoeyJoeJo> I was just using mp3 as an example.
[17:01] <Mavrik> for most encoders the file will be noticably smaller
[18:33] <vl4kn0> Hi, where can I find api-example of libavfilter? I tried github repository but there's nothing like that.
[18:49] <Plorkyeran> ffmpeg.c is probably it
[18:49] <Plorkyeran> I don't know of any external projects which use libavfilter
[19:01] <JEEB> LAV Video from LAV Filters does actually use libavfilter, and I think mpv has a libavfilter interface?
[19:10] <theo_> I want to compile FFmpeg with libfdk_aac, but where to put the - "--enable-libfdk_aac --enable-nonfree"? I'm using FFmpeg compile guide https://ffmpeg.org/trac/ffmpeg/wiki/UbuntuCompilationGuide
[19:33] <spaam> theo_: configure line
[19:46] <theo_> spaam: Thanks! Just for clarification, is it the configure line when I install fdk-aac? or ffmpeg?
[19:47] <spaam> theo_: ffmpeg
[19:48] <spaam> theo_: since you want to add support for libfdk_aac in ffmpeg
[19:49] <theo_> So maybe I don't need fdk-aac? I want to convert mp3 to m4a =)
[19:54] <klaxa> m4a is not a codec though
[20:10] <theo_> klaxa: I'm always forgetting that. Mp3 to aac, isn't it?
[20:10] <klaxa> i'd think so
[20:11] <klaxa> and fdk_aac is the recommended aac encoder
[20:15] <theo_> klaxa: Do I also need libmp3lame? I never going to convert to mp3, just from it.
[20:15] <klaxa> i don't think you need libmp3lame to decode mp3
[20:16] <theo_> klaxa: That was really helpfull. thanks you!
[21:09] <christian-c> allright.. so I have followed every step in this guide, and I still get "-bash: ffmpeg: command not found" when typeing "ffmpeg" in the cmd. My server is running Debian 7.0 64-bit. Am I missing something?
[21:09] <christian-c> the guide: https://trac.ffmpeg.org/wiki/UbuntuCompilationGuide#CompilationInstallation
[21:30] <relaxed> christian-c: ~/ffmpeg_build/bin/ffmpeg
[21:31] <relaxed> wait, it should be in ~/bin- try ~/bin/ffmpeg
[22:18] <Patrickk> hi all , please help me... i try to install ffmpeg on centos 6....
[22:18] <Patrickk> and if i make ./configure --enable-version3 .......
[22:18] <Patrickk> i have this error
[22:18] <Patrickk> If you think configure made a mistake, make sure you are using the latest version from Git.
[22:19] <Patrickk> admin on?
[22:21] <spaam> what error?
[22:21] <spaam> with configure line ?
[22:21] <spaam> can you show us the whole output?
[22:23] <Patrickk> [root at youtubecrestin ffmpeg-0.9.1]# ./configure --enable-version3 --enable-shared --enable-gpl --enable-nonfree --enable-postproc --enable-libmp3lame --enable-libvorbis --enable-libxvid --enable-libfaac --enable-libtheora --enable-libx264 ERROR: libtheora not found  If you think configure made a mistake, make sure you are using the latest version from Git.  If the latest version fails, report the problem to the ffmpeg-user at ffmpeg.
[22:23] <spaam> Patrickk:  ERROR: libtheora not found
[22:23] <spaam> Patrickk: you need to install the -devel packet for libtheora
[22:23] <Patrickk> it is installed
[22:24] <spaam> then you might have a old version
[22:24] <Patrickk> [root at youtubecrestin ffmpeg-0.9.1]# ls -l /usr/local/lib/libtheora* -rw-r--r-- 1 root root 447526 Sep 11 20:13 /usr/local/lib/libtheora.a -rwxr-xr-x 1 root root    941 Sep 11 20:13 /usr/local/lib/libtheora.la lrwxrwxrwx 1 root root     19 Sep 11 20:13 /usr/local/lib/libtheora.so -> libtheora.so.0.3.10 lrwxrwxrwx 1 root root     19 Sep 11 20:13 /usr/local/lib/libtheora.so.0 -> libtheora.so.0.3.10 -rwxr-xr-x 1 root root 353884 Sep 1
[22:25] <spaam> i have no idea have to read what version that is
[22:25] <Patrickk> libtheora?
[22:25] <spaam> yes.
[22:26] <Patrickk> libtheora.so.0.3.10
[22:26] <spaam> thats not the version number.
[22:27] <spaam> looks like libtheora 1.1.1
[22:27] <Patrickk> http://downloads.xiph.org/releases/theora/libtheora-1.1.1.tar.gz
[22:27] <spaam> maybe configure didnt find it? check config.log
[22:28] <Patrickk> bow?
[22:30] <spaam> open it ?
[22:30] <spaam> cat config.log
[22:30] <spaam> and look and the bottom and see if you find somthing about theora
[22:32] <Patrickk> checking for Ogg... no *** Could not run Ogg test program, checking why... *** The test program failed to compile or link. See the file config.log for the *** exact error that occured. This usually means Ogg was incorrectly installed *** or that you have moved Ogg since it was installed. configure: error:
[22:35] <mittens> do you have libogg?
[22:35] <Patrickk> i reinstalled now... libogg.. i think , now work
[22:43] <Patrickk> Found yasm 1.1.99.HEAD Minimum version is yasm-1.2.0
[22:43] <Patrickk> a tutorial to install yasm 1.20 on centos ?
[22:46] <JEEB> grab the source code, run the configure with --prefix set to where you want to install it, make, make install
[22:46] <JEEB> yasm is very straightforward
[22:55] <Patrickk> 2013-09-11 20:51:04 (500 KB/s) - `yasm-1.2.0-1.el6.x86_64.rpm' saved [753752/753752]  [root at youtubecrestin tmp]# rpm -ivh yasm-1.2.0-1.el6.x86_64.rpm Preparing...                ########################################### [100%]    1:yasm                   ########################################### [100%]
[22:56] <Patrickk> and then my ERROR [root at youtubecrestin x264-snapshot-20130910-2245-stable]# ./configure --enable-shared Found yasm 1.1.99.HEAD Minimum version is yasm-1.2.0
[22:57] <Patrickk> AND UP I INSTALLED YASM 1.2.0-1
[22:57] <spaam> good
[22:58] <spaam> maybe you have an older version of yasm installed somewhere else?
[22:58] <spaam> yasm --version say ?
[23:00] <Patrickk> http://dl.fedoraproject.org/pub/epel/6/x86_64/yasm-1.2.0-1.el6.x86_64.rpm
[23:00] <Patrickk> ir is not old
[23:02] <spaam> as i said
[23:02] <spaam> you might have an old version installed already
[23:02] <Patrickk> it is not old..... i installed now from this link
[23:03] <Patrickk> http://dl.fedoraproject.org/pub/epel/6/x86_64/yasm-1.2.0-1.el6.x86_64.rpm
[23:03] <Patrickk> it is 1.2.0
[23:03] <Patrickk> not 1.9
[23:04] <Patrickk> [root at youtubecrestin tmp]# rpm -ivh yasm-1.2.0-1.el6.x86_64.rpm Preparing...                ########################################### [100%]    1:yasm                   ########################################### [100%]
[23:05] <spaam> what does yasm --version say?
[23:05] <Patrickk> [root at youtubecrestin x264-snapshot-20130910-2245-stable]# ./configure --enable-shared Found yasm 1.1.99.HEAD Minimum version is yasm-1.2.0 If you really want to compile without asm, configure with --disable-asm.
[23:06] <spaam> what does yasm --version say?
[23:06] <Patrickk> i installed 1.20-1
[23:06] <spaam> but what does yasm --version say=
[23:08] <Patrickk> yasm 1.1.99.HEAD Compiled on Sep 11 2013.
[23:09] <spaam> there you have it
[23:10] <spaam> somewhere in your $PATH you have an old version
[23:12] <Patrickk> do you know how i can make this wihout yasm ?
[23:12] <Patrickk> cd /tmp/x264-snapshot-20120214-2245-stable/ ./configure --enable-shared make make install
[23:13] <spaam> you want yasm
[23:13] <Patrickk> i dont whant
[23:14] <spaam> if you going to encode stuff using x264 you _really_ want it. if you dont have it it will be soooo sloow
[23:15] <Patrickk> i whant to install libx264
[00:00] --- Thu Sep 12 2013