[Ffmpeg-devel-irc] ffmpeg.log.20130708

[burek]

[00:18] <Hafydd> Where are the -b, -vb, and -ab options of ffmpeg documented?
[00:50] <brontosaurusrex> Hafydd, you mean -b:v and -b:a ?
[00:59] <MachinaeWolf> Hmm tried someone's perfectly working ffmpeg on debian that they compiled and still an error: http://pastebin.com/g8mtuG62
[01:00] <djahandarie> Hmmm, does anyone know what controls this? "Aspect ratio is 1.38:1 - scaling to correct movie aspect."
[01:01] <djahandarie> Because that results in some weird scaling, "VO: [xv] 1920x1080 => 1920x1394 Planar YV12", which I don't want.
[01:01] <brontosaurusrex> djahandarie, aspect ratio flag of the input stream
[01:01] <djahandarie> I crop and scale using -filter:v, does it not automatically correct the aspect ratio?
[01:02] <djahandarie> (Those pastes were from mplayer on the output file.)
[01:05] <djahandarie> Nevermind, error in the commandline, the crop wasn't getting applied properly.
[01:10] <Hafydd> brontosaurusrex: I'm not sure, because I haven't seen the documentation yet.
[01:11] <Hafydd> Let's just say I mean -b (or -b:a and -b:v, if you prefer), for now.
[01:14] <brontosaurusrex> Hafydd, well, scroll around the docs for a while and you should spot it
[01:15] <Hafydd> brontosaurusrex: which docs?
[01:15] <brontosaurusrex> man ffmpeg
[01:17] <Hafydd> I searched for "bitrate" and "bit rate" and "-b" in `man ffmpeg` and `man ffmpeg-filters` and `man ffmpeg-bitstream-filters`, to no avail.
[01:17] <Hafydd> I'd be interested to see if you can find it there, and tell me what phrase to search for.
[01:17] <Hafydd> (Because I also scrolled through all of the options and didn't see it)
[01:22] <Hafydd> Or indeed anybody who knows were it's documented - my question still stands.
[01:22] <brontosaurusrex> hmm, i have the html version of docs
[01:23] <brontosaurusrex> and its on the main page
[01:23] <brontosaurusrex> or the first page
[01:24] <Hafydd> What does it say, brontosaurusrex?
[01:25] <Hafydd> Ah, I found this ticket, actually: http://trac.ffmpeg.org/ticket/613
[01:25] <Hafydd> Maybe it was fixed in a later version than I have.
[01:28] <brontosaurusrex> dunno, they keep changing cli all the time is all I know
[01:28] <Hafydd> So it's in the output of ffmpeg -help, but not the manual page.
[01:29] <Hafydd> How could such a basic option go poorly documented for so long. Maybe I should open another ticket.
[01:33] <Hafydd> brontosaurusrex: what does your HTML documentation say about -b?
[01:34] <brontosaurusrex> Hafydd, i have you a link
[01:34] <brontosaurusrex> *gave
[01:35] <Hafydd> brontosaurusrex: you did no.
[01:35] <Hafydd> *not
[01:37] <Hafydd> Oh, excuse me, my client didn't notify me of the notice.
[01:37] <brontosaurusrex> relaxed, running your Linux.Encoding.install
[01:38] <Hafydd> Where on that page does it describe the -b option, brontosaurusrex?
[01:38] <brontosaurusrex> relaxed, it would be cool to add fdk-aac
[01:38] <brontosaurusrex> Hafydd, no, the -b:v option and -b:a
[01:39] <Hafydd> That's what I mean by "the -b options". :v and :a are additional arguments to an option name,
[01:39] <Hafydd> So, where are they described?
[01:41] <brontosaurusrex> Hafydd, wtf is wrong with you?
[01:41] <brontosaurusrex> i told you 5 times allready
[01:42] <Hafydd> All you've told me are a number of false statements.
[01:42] <brontosaurusrex> Hafydd, ok, just put me on ignore, ok?
[01:42] <Hafydd> I appreciate the effort to help, but I've put more effort into trying to find this than you think.
[01:43] <Hafydd> So if there actually isn't any documentation, it's better to admit that.
[01:43] <brontosaurusrex> Hafydd, iam not a ffmpeg dev, just a user like you
[01:43] <brontosaurusrex> Hafydd, so i have really nothing to "admit"
[01:44] <Hafydd> You can admit that you didn't even look in the documentation before wasting my time directing me to it.
[01:49] <brontosaurusrex> i want my lawyer
[02:25] <brontosaurusrex> trying to compile ffmpeg with x264 and i constantly get : ERROR: libx264 not found
[02:25] <brontosaurusrex> i've tryed all different possible approaches, including relaxed superscript
[02:25] <brontosaurusrex> always the same error
[02:28] <brontosaurusrex> any clues?
[02:28] <drv> check config.log and see why the test fails
[02:32] <brontosaurusrex> ok
[02:34] <brontosaurusrex> same thing at the bottom : ERROR: libx264 not found
[02:34] <brontosaurusrex> i guess for some reason libx264.so is not generated
[04:42] <mapreduce> Hi.  I can't seem to build the most recent FFmpeg on msys (Windows) but it builds ok on Linux.
[04:42] <mapreduce> make gives: common.mak:139: *** missing separator.  Stop.
[04:43] <mapreduce> line 139 of that file is $(eval $(RULES)), looks reasonable to me but I don't know make very well.  Any suggestions?
[04:59] <mapreduce> Nevermind, I think git config core.autocrlf false fixed it.
[07:04] <nownot> trying to convert flac to m4a and I keep getting Input stream #0:0 frame changed from rate:44100 fmt:s16 ch:2 chl:stereo to rate:44100 fmt:s16p ch:2 chl:stereo
[07:04] <nownot> using command for f in *.flac; do avconv -i "$f" -c:a alac "${f%.*}.m4a"; done & ideas?
[07:32] <highgod> Hi, I want to ask a question, I use 64bit ffmpeg libs and dlls, it cursh on my code, but I use the 32bit dlls and libs and using the same code ,the code run correct, can anyone give some help? thanks
[07:35] <robpitt> Hi, how can I use ffmpeg to capture audio from the microphone on android?
[08:50] <igmrlm> g'day
[08:51] <igmrlm> what would be the command line to generate random video noise with ffmpeg, like the snow from a tv set?
[08:51] <igmrlm> apparently one could use the noise filter but i'm unsure the syntax
[08:52] <igmrlm> i found a site that tells how to do this from /dev/urandom but i'm not on a linux machine at the moment
[11:10] <mark4o> igmrlm: ffmpeg -f lavfi -i "nullsrc=s=640x480,geq=random(1)*255:128:128" ...
[11:11] <mark4o> the filters are documented at http://ffmpeg.org/ffmpeg-filters.html
[12:18] <robpitt> On android, how can I use ffmpeg to record the microphone?
[14:31] <boys21> hi, Im trying to enable libx264 with ffmpeg in Mingw environment, but I keep getting error: libx264 not found
[14:32] <boys21> i have x264 lib. include and bin folders in
[14:32] <zap0> is your mingw build system really an ffmpeg problem?
[14:33] <boys21> can someone help me with suggesstions
[14:34] <boys21> no i guess, i compiled ffmpeg with out x264 more than thrice, but I get troubles enabling x264
[14:35] <Mavrik> boys21, check config.log
[14:35] <Mavrik> and see why ffmpeg thinks you have no libx264
[14:37] <boys21> ffconf.wpjIoPmL.o_converted.c c99wrap link -L/usr/local/lib -nologo -out /tmp/ffconf.DdtrQknc.exe /tmp/ffconf.wpjIoPmL.o x264.lib psapi.lib advapi32.lib shell32.lib LINK : warning LNK4044: unrecognized option '/LC:/MinGW/msys/1.0/local/lib'; ignored LINK : fatal error LNK1181: cannot open input file 'x264.lib' ERROR: libx264 not found
[14:37] <boys21> these were my last lines
[14:41] <Mavrik> is x264.lib stored in your MinGW/msys/local/lib?
[14:43] <boys21> yes i have x264.a
[14:44] <boys21> i dont have x264.lib
[14:44] <Mavrik> hmm
[14:44] <Mavrik> .lib is usually extension for MSVC generated static libraries
[14:44] <Mavrik> .a for gcc generated static libraries
[14:45] <boys21> okay
[14:45] <Mavrik> did you compile your x264 with mingw beforehand?
[14:45] <Mavrik> and are you compiling ffmpeg with MSVC or mingw gcc?
[14:45] <boys21> i compiled x264 before hand
[14:45] <boys21> with mingw
[14:45] <boys21> and now im compiling ffmpeg with msvc with x264 option
[14:46] <Mavrik> ah
[14:46] <Mavrik> compile it with gcc.
[14:46] <Mavrik> and it'll look for the .a library
[14:46] <boys21> okay
[14:46] <boys21> how can i port them to msvc later?
[14:49] <Mavrik> no idea
[14:50] <boys21> ok, thanks anyways. I will give it a try in this direction.
[14:50] <Mavrik> boys21, depends on how you want to use them
[14:50] <boys21> it worked disabling the msvc
[14:50] <Mavrik> I had no problems using gcc compiled ffmpeg/x264 .dll from MSVC and python projects
[14:50] <boys21> ok
[14:50] <Mavrik> for static linking you should probably compile everything with MSVC
[14:51] <boys21> previously, i used the msvc tool chain option and used the staticlibs in the additional dependencies to compile ffmpeg projects
[14:52] <boys21> i mean the additional dependencies of project properties in  vs 2012
[15:02] <boys21> Thanks for suggestion, i wil try to recompile libx264 using msvc tool chain option
[15:56] <robpitt> How can I record microphone audio in android?
[16:08] <zap0> use android api
[16:20] <robpitt> zap0 I'm not an android coder is there not some way i can do it on the command line with the ffmpeg binary?
[16:21] <zap0> i dont know. i suspect you'd find it in a simple google query;  if not, then probably not.
[16:22] <robpitt> I've been googling for hours so maybe it can't be done. I see references to using a pipe but don't know how to make one
[16:23] <zap0> you do know that access to hardware goes via the permissions system
[16:24] <zap0> so unless its an obvious device you can open linux style,  you likely need permissions to access the hardware
[16:24] <Mavrik> robpitt, you cannot.
[16:24] <Mavrik> microphone is especially protected on Android
[16:25] <Mavrik> on most devices it's not possible to record conversations even with root access
[16:25] <robpitt> Shame. I've got root and still can't do it
[16:26] <zap0> apps can acquire permissions. so API's can access it.
[16:28] <robpitt> There's a patched ffmpeg that has a device called lsfaudio but I really wanted it to work with common builds of ffmpeg
[16:30] <zap0> it's not a ffmpeg issue.  teh platform doesn't allow it
[16:36] <robpitt> ffmpeg could do it though, the patched version proves that
[16:38] <zap0> that doesn't prove `FFMPEG` is doing it... it proves some android specific API stuff has been grafted in.
[16:38] <zap0> you said yourself, you wanted a `native` ffmpeg to work. not some franken-ffmepg
[16:39] <JEEB> robpitt, if you want to see the feature get into ffmpeg, then feel free to grab the patch, see that its license is LGPL and not GPL, and then update it to the current ffmpeg master -> post on the mailing list
[16:40] <JEEB> (and lurking around ffmpeg-devel is a good idea)
[17:04] <robpitt> I don't have a PC so I tried to compile it on my phone but it requires the android ndk. Sigh.
[17:55] <curfont> Hi I am trying to save an RTSP stream with ffmpeg, and it works in VLC but I can't find the arguments to use in actual ffmpeg
[17:56] <curfont> It's an IP CAM streaming h264 over RTSP/554
[17:56] <Fjorgynn> ffmpeg -i ?
[17:56] <curfont> Doesn't work, neither do examples I find online
[17:56] <curfont> It says rtsp://192.168.0.220:554/live/h264/ch0: Invalid data found when processing input
[18:51] <diegoviola> hi
[18:54] <diegoviola> i have a home theater device that is not reading 720p mkv files, but it reads fine xvid videos with lower quality/format/res, is there a way to change the container format from the 720p files to avi, so the home theater would read them?
[18:54] <diegoviola> without reencoding the audio/video in the process
[18:56] <curfont> Ok guys, found the problem, this camera does not accept the "DESCRIBE" in the RTSP protocol that is sent by ffmpeg
[18:58] <curfont> But it is fine with the one sent by LVC
[18:58] <curfont> VLC*
[19:03] <xlinkz0> diegoviola: -c copy out.avi ?
[19:04] <diegoviola> can avi carry H.264 video and AAC audio?
[19:05] <diegoviola> xlinkz0: thanks
[19:05] <xlinkz0> one way of finding out is testing it :)
[19:05] <diegoviola> right, thanks
[19:05] <diegoviola> i'll give it a try
[19:24] <curfont> Ok, so this specific camera needs a User-Agent in the DESCRIBE, is there any way to get ffmpeg to add a User-Agent to the RTSP packets?
[19:58] <curfont> fyi https://ffmpeg.org/trac/ffmpeg/ticket/2761
[20:27] <diegoviola> crap
[20:27] <diegoviola> this home theater won't accept H.264 also
[20:27] <diegoviola> nor AAC
[20:27] <diegoviola> not even mkv
[20:27] <diegoviola> it seems to only accept avi/xvid/mp3
[20:28] <diegoviola> 2-pass encoding only affects file size or also quality?
[20:28] <diegoviola> when it comes to xvid
[20:28] <Fjorgynn> :(
[20:29] <Plorkyeran> xvid doesn't have a good 1-pass mode, if that's what you're asking
[20:29] <Fjorgynn> 2 pass affects the bitrate
[20:30] <diegoviola> Plorkyeran: xvid is 2-pass encoding always?
[20:31] <Plorkyeran> you can do 1-pass encodes, but it'll be lower quality at the same filesize as a 2-pass encode
[20:31] <diegoviola> ok
[20:32] <diegoviola> ty
[20:33] <LithosLaptop> diegoviola: http://ffmpeg.org/trac/ffmpeg/wiki/How%20to%20encode%20Xvid%20/%20DivX%20video%20with%20ffmpeg
[20:34] <diegoviola> is it best to use the xvid encoder or the native mpeg4 ffmpeg encoder?
[20:35] <diegoviola> i'll try the ffmpeg encoder first
[20:37] <Fjorgynn> :)
[20:38] <diegoviola> ty folks
[20:58] <MrE> Hey ive used this guide http://trac.ffmpeg.org/wiki/How%20to%20burn%20subtitles%20into%20the%20video for burning picture based dvb subtitles onto the video, it works quite good, but when I use a 720p hd source as input without rezise or anything else the subtitles is small and placed in the middle of the picture, using 1080p hd as source and resizing to 720p its even worse. Here is some samples :
[20:58] <MrE> http://10only.explode.dk/screenshot003.png
[20:58] <MrE> http://10only.explode.dk/screenshot004.png
[20:59] <diegoviola> wth?
[20:59] <diegoviola> Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height
[20:59] <diegoviola> mpeg4/xvid requires a specific width/height?
[20:59] <diegoviola> could i preserve the same 720p quality of h264?
[20:59] <JEEBsv> see the lines before that
[20:59] <MrE> Tried serveral options with overlay=main_w and all that, nothing seems to work, any help would be appreciated :)
[20:59] <JEEBsv> that should contain the actual error
[21:00] <JEEBsv> the "maybe incorrect parameters" thing just means that you will probably want to look at the previous lines for an error
[21:10] <diegoviola> http://pastie.org/8122051
[21:11] <diegoviola> bitrate too low i guess?
[21:11] <diegoviola> well it seems like it
[21:11] <diegoviola> tried 800k and it works now
[21:12] <diegoviola> but my first pass was with 555k
[21:13] <diegoviola> i'll try again with 1k both passes
[21:55] <someone-noone> Hello!
[21:55] <someone-noone> ./configure --enable-gpl --enable-libx264 --prefix=/usr --enable-libopus --enable-libvpx
[21:56] <someone-noone> ERROR: opus not found
[21:56] <someone-noone> opus headers are in /usr/include/opus and /usr/include
[21:56] <someone-noone> opus libs are in /usr/lib/
[21:56] <someone-noone> in config.log last 2 lines are:
[21:56] <someone-noone> check_pkg_config opus opus_multistream.h opus_multistream_decoder_create
[21:56] <someone-noone> ERROR: opus not found
[21:57] <someone-noone> What is going on and how to fix it?
[21:57] <someone-noone> ffmpeg version is latest from git
[23:29] <semme> guys i've got a really strange problem: im remuxing an h264 f4v to mp4 (copying the streams)
[23:29] <semme> and the output is ok (picture, sound), but the output's fps are 320fps
[23:30] <semme> im doing r_frame_rate = av_d2q(_fps, INT_MAX);
[23:31] <JEEBsv> you should be using PTS with such containers, really
[23:31] <semme> i am
[23:31] <semme> im copying the pts/dts values from the FLV, but thats not working either
[23:32] <semme> is there a diff between MP4/FLV pts/dts?
[23:33] <JEEBsv> there's o9k things you could be doing wrong there, and I have no idea what you might be doing wrong :P use L-SMASH's boxdumper or something to look at the CTS (almost PTS in "MP4" lingo, but not exactly) values of your samples
[23:33] <JEEBsv> so you can see if you have the correct X/Y rate, as in if your PTS are correct on the timescale
[23:34] <semme> ok thanks, im going to double check the pts/dts values
[23:35] <JEEBsv> and as I said, you can use L-SMASH's boxdumper tool to dump a list of CTS values for the samples
[23:35] <JEEBsv> of the output mov-like ("mp4" etc.)
[23:39] <semme> might be a dumb question, but is l-smash tested on osx?
[23:43] <semme> hmm guess not
[23:44] <JEEBsv> it is
[23:44] <JEEBsv> compiling it should be rather straightforward
[23:45] <JEEBsv> it doesn't really require anything
[23:45] <JEEBsv> ./configure with possibly maybe --cc=clang or something if you want to use clang
[23:45] <JEEBsv> and then make
[23:46] <mark4o> works for me on os x
[23:47] <JEEBsv> yes, there have been people using OS X related to the L-SMASH development, and in general it's just so simple it'd be surprising if it'd broke
[00:00] --- Tue Jul  9 2013