[Ffmpeg-devel-irc] ffmpeg.log.20111125

[burek]

[00:10] <pythonirc101> Any suggestions for converting a mts file to high quality video that media player can read?
[00:10] <pythonirc101> I tried -i x.mts output.mp4 -- and the file size became 5 times smaller
[00:20] <pythonirc101> -sameq for same quality?
[00:48] <pythonirc101> instead of saying -b 6000k -- how can i say , use the sam quality setting as the input?
[00:59] <pythonirc101> how much difference does "-g" make on an encoding? Whats the optimum value of it for 1080p videos?
[02:07] <jesselang|laptop> Hi. I'm using ffmpeg to convert an image to a DV file. It seems to change and flutter a bit (especially the text on the image). I've tried using ffmpeg directly, and tried ffenc_dvvideo from gstreamer.
[02:08] <jesselang|laptop> The folks in #gstreamer said that the encoder must not encode identical frames deterministically.
[02:10] <jesselang|laptop> Could anyone help me out with this?
[05:12] <navatwo> Could someone help me out with a command to copy audio from a mp4 container into another mp4 container?
[05:12] <navatwo> They *are* the same length
[05:33] <burek> just use -acodec copy
[05:35] <burek> pythonirc101m
[05:35] <burek> can you please use pastebin.com, to show your command line and its output?
[05:36] <burek> you can use ffmpeg -i input -vcodec libx264 -crf 20 -acodec copy out.mp4
[07:51] <sam1> hi guys 
[07:51] <sam1> i want  to play  a movie (*.avi + *.srt) with ffplay,what's command iuse?
[08:16] <TryDent1> does ffmpeg support appleloss codec?
[08:17] <Tjoppen> you mean alac? yes, for several years I'm sure
[08:19] <TryDent1> ffmpeg -i R:\1.vob -acodec whatdoiputhereforalac R:\1.alac
[08:19] <relaxed> alac
[08:19] <Tjoppen> :)
[08:20] <TryDent1> where do i see the acodec and vcodec list
[08:20] <Tjoppen> there might be an option to spend more cpu time to improve the compression, not sure
[08:20] <sacarasc> ffmpeg -codecs
[08:21] <TryDent1> anybody know what is the command to convert any audio format to  wav
[08:22] <TryDent1> wow list is huge
[08:22] <TryDent1> wow it even support vc1
[08:22] <TryDent1> amazing
[08:24] <TryDent1> sacarasc is the "ffmpeg -codecs list" legit?
[08:26] <TryDent1> anybody know?
[08:32] <grepper> -acodec pcm_s16le -ar 48000 or somesuch
[08:32] <TryDent1> why le not be
[08:33] <TryDent1> and why 16
[08:34] <TryDent1> ffmpeg -i R:\a.wav -acodec alac R:\a.alac    does not work
[08:36] <relaxed> ffmpeg -i R:\a.wav -acodec alac R:\a.m4a
[08:38] <TryDent1> relaxed why is ffmpeg picky with  file extension
[08:39] <TryDent1> why won't this work?   ffmpeg -i R:\a.wav -acodec wmalossless  R:\a.wma
[08:42] <relaxed> what is your goal? what are you doing?
[08:42] <TryDent1> wmalossless; that's my goal
[08:43] <TryDent1> why won't it work
[08:43] <relaxed> ffmpeg doesn't support it
[08:43] <TryDent1> it's in the codec list
[08:43] <relaxed> it's not in mine
[08:43] <relaxed> who uses wmalossless anyway?
[08:43] <TryDent1> it's in mine
[08:44] <TryDent1> relaxed  then it shouldn't be in the list
[08:44] <relaxed> I have never run across one
[08:46] <TryDent1> relaxed okay ; i am confused about  converting to wav
[08:46] <TryDent1> which one do i use
[08:46] <relaxed> ffmpeg -i input output.wav
[08:46] <TryDent1> what do i put for acodec
[08:48] <relaxed> you don't need to -acodec
[08:48] <TryDent1> then what is this?  <grepper> -acodec pcm_s16le -ar 48000
[08:49] <relaxed> well, ffmpeg will decide the correct acodec/audio rate from the format.
[08:50] <TryDent1> how come "ffmpeg -codecs"  does not show that i have to use  .m4a for  acodec alac
[08:50] <TryDent1> or am i missing something
[08:51] <relaxed> m4a is a format; alac is a codec
[08:51] <relaxed> ffmpeg -formats
[08:52] <relaxed> by the way, the world decided on flac unless you need it for some iProduct.
[08:53] <TryDent1> m4a is not listed in  ffmpeg -formats
[08:54] <relaxed>  D  mov,mp4,m4a,3gp,3g2,mj2 QuickTime/MPEG-4/Motion JPEG 2000 format
[08:54] <TryDent1> basically how would i know if acodec alac  has to use m4a  if you didn't told me in the channel
[08:55] <relaxed> but it's basically the mp4 format
[08:55] <TryDent1> that does not answer myq esution
[08:56] <relaxed> how about you google and read a little? http://en.wikipedia.org/wiki/Apple_Lossless
[08:57] <TryDent1> that doesn't answer my question still
[08:58] <TryDent1> and this is ffmpeg question
[09:07] <TryDent1> i found a huge bug in ffmpeg  i had original.wav  and converted to  alac and flac  and converted back to wav;    original.wav and  decodedback.wav does not have same md5sum
[09:10] <Tjoppen> not a bug (probably)
[09:11] <Tjoppen> try remuxing original.wav first, then do the same test
[09:11] <TryDent1> remuxing orignal.wav what do you mean
[09:12] <Tjoppen> original.wav could have come from anywhere, with all sorts of metadata
[09:12] <Tjoppen> or some strange structure
[09:12] <Tjoppen> ffmpeg -i original.wav -acodec copy remuxed.wav
[09:12] <Tjoppen> then compare remuxed.wav to decodedback.wav
[09:13] <TryDent1> i shouldn't have to remux it
[09:14] <TryDent1> what if somebody did the same thing with zip and md5sum didn't match
[09:14] <Tjoppen> uh.. zip?
[09:15] <Tjoppen> use vbindiff
[09:15] <Tjoppen> and compare the data atoms
[09:15] <TryDent1> no i trust md5sum
[09:15] <Tjoppen> or hell, just look att both files using vbindiff and you can clearly see the different
[09:16] <Tjoppen> or you know, compare the raw audio properly by muxing to -f s16le
[09:16] <Tjoppen> (IIRC)
[09:25] <TryDent1> okay let me convert to wav using flac program
[09:27] <TryDent1> if i convert back to wav using flac program then md5sum matches
[09:30] <TryDent1> problem with ffmpeg when convert back to wav
[09:31] <Tjoppen> just look with vbindiff. the difference will probably be obvious
[09:31] <Nagy>  Is it possible to make saw scale 
[09:31] <Nagy> Nvm, accidental send.
[09:36] <TryDent1> trydent why don't i just use filecompare app or md5sum; that's more accurate
[09:38] <Tjoppen> they don't help with diagnostics
[09:39] <Tjoppen> also, are the files the same size?
[09:39] <TryDent1> no, it's 2 bytes samller
[09:41] <Tjoppen> look at this: https://ccrma.stanford.edu/courses/422/projects/WaveFormat/  then look in vbindiff where the difference is
[09:41] <Tjoppen> specifically if the size field for the data chunk is the same
[09:42] <Tjoppen> it sounds like one sample might have gotten losst
[09:43] <TryDent1> problem is not with ffmpeg encoding
[09:45] <TryDent1> why does md5sum match when i use flac decoder then
[09:46] <TryDent1> it's a bug; please fix it
[09:48] <Tjoppen> file a ticket then
[09:49] <Tjoppen> http://ffmpeg.org/bugreports.html
[09:51] <TryDent1> vbindiff.exe R:\a.wav R:\b.wav
[09:52] <TryDent1> what am i suppose to be looking at
[09:52] <Tjoppen> do you see anything highlighted in red?
[09:52] <TryDent1> yes
[09:52] <TryDent1> what about it
[09:53] <Tjoppen> or rather, is the other file so that the whole thing is shifted two bytes?
[09:53] <TryDent1> tjoppen if you don't believe me; why don't you just try  encoding/decoding a wav file yourself
[09:55] <Tjoppen> I don't doubt you. anyway, reproduced locally
[09:55] <Tjoppen> went wav -> flac -> wav and the header got 2 B bigger
[09:55] <TryDent1> was it 2 bytes off?
[09:55] <Tjoppen> yah
[09:55] <TryDent1> sorry you are right; bigger not smaller
[09:55] <TryDent1> my bad
[09:56] <TryDent1> so where is the bug?
[09:57] <TryDent1> trydent1  try flac decoder
[10:01] <Tjoppen> seems to be writing 2 B of extradata. not sure why
[10:01] <TryDent1> does remuxing help?
[10:01] <TryDent1> your first idea
[10:06] <Tjoppen> yeah, the remuxed file is identical
[10:06] <Tjoppen> I see there was a fix to the wav muxer quite recently
[10:06] <Tjoppen> which might explain why the header is larger
[10:06] <Tjoppen> anyway, the data is the same which is what's important
[10:06] <TryDent1> you mean it took thing long for somebody find this 2 byte bug?
[10:06] <TryDent1> this*
[10:07] <TryDent1> when  was ffmpeg invented?
[10:07] <Tjoppen> it has to do with matroska
[10:07] <TryDent1> huh? what do you mean
[10:07] <TryDent1> that' mkv
[10:07] <Tjoppen> yeah. it shares code with the wav mxuer
[10:07] <Tjoppen> *shrugs*
[10:08] <TryDent1> so this bug is fixed now?
[10:08] <Tjoppen> I wouldn't call it a bug. the wav header is somewhat flexible - expecting two files to have the exact same header is pushing it
[10:08] <TryDent1> tjoppen flac decoder decodes exactly
[10:10] <Tjoppen> 2c4e08d89327595f7f4be57dda4b3775e1198d5e is the culprit
[10:10] <Tjoppen> in fact:
[10:10] <Tjoppen> Since fate uses wav files for the audio test a larger number of tests
[10:10] <Tjoppen> has changed checksums or shifted positions due to the 2 byte longer
[10:10] <Tjoppen> wave header.
[10:14] <Tjoppen> considering this is appearently the correct way to write the header I'd say the bug is in flac
[10:14] <Tjoppen> if you feel it's a big issue I'd file a ticket in both projects' trackers
[10:15] <TryDent1> hmm i found another bug
[10:15] <TryDent1> ffmpeg -i a.mlp -acodec flac a.flac    a.flac is smaller than a.mlp
[10:15] <TryDent1> much smaller
[10:16] <Tjoppen> mlp?
[10:16] <TryDent1> a.mlp is 24bit   but  a.flac became 16 bit
[10:16] <TryDent1> meridiallosslesspacking
[10:17] <Tjoppen> -sample_fmt s24le
[10:17] <TryDent1> huh?
[10:17] <Tjoppen> -acodec flac -sample_fmt s24le   see if that works
[10:18] <Tjoppen> *s32 even
[10:18] <Tjoppen> or.. uh
[10:18] <TryDent1> oh, why would i have to do that?  
[10:18] <TryDent1> it's 24bit not 32bit
[10:18] <Tjoppen> I know
[10:18] <TryDent1> so which one do i use
[10:18] <Tjoppen> it's supposed to not by this dumb I think
[10:18] <Tjoppen> *be
[10:19] <Tjoppen> that sounds like an actual issue. but to try -sample_fmt s32  and see what happens
[10:20] <TryDent1> still producing 16bit
[10:20] <TryDent1> let me try flac encoder
[10:20] <TryDent1> doesn't work since flac cannot do from mlp
[10:22] <TryDent1> even ffmpeg -i a.mlp a.wav  is not doing properly
[10:24] <relaxed> TryDent1: your ffmpeg version?
[10:24] <TryDent1> C:\>ffmpeg --version
[10:24] <TryDent1> ffmpeg version N-34906-g4e7b3ef, Copyright (c) 2000-2011 the FFmpeg developers
[10:24] <TryDent1>   built on Nov 16 2011 12:35:07 with gcc 4.6.2
[10:27] <TryDent1> relaxed; i thought  ffmpeg -i a.mlp a.wav   is smart enough to use same as source
[10:27] <TryDent1> it keeps using 16bit
[10:30] <TryDent1> again, flac decoder  decodes properly and uses 24bit  if the source is 24bit
[10:30] <TryDent1> but not  ffmpeg
[10:31] <TryDent1> Tjoppen you going to blame this on mkv too?
[10:34] <Tjoppen> to wav you can use -acodec pcm_s24le
[10:35] <Tjoppen> it might be that the mlp demuxer isn't setting some field in AVCodecContext correctly
[10:36] <Tjoppen> try ffmpeg -i a.mlp -acodec pcm_s24le a.wav   and then ffmpeg -i a.wav a.flac   and see what happens
[10:37] <TryDent1> what about decode back to wav from  24bitflac
[10:38] <Tjoppen> just see what ffprobe says about the resulting flac first
[10:40] <TryDent1> ffmpeg -i a.wav a.flac  is producing 16bit flac
[10:42] <Tjoppen> is a.wav 24-bit?
[10:42] <TryDent1> yes
[10:43] <Tjoppen> flacenc.c:221: if (avctx->sample_fmt != AV_SAMPLE_FMT_S16)
[10:43] <Tjoppen> return -1;
[10:44] <TryDent1> huh
[10:44] <Tjoppen> so only 16-bit is supported
[10:44] <Tjoppen> it should probably nah or something
[10:44] <Tjoppen> *nag
[10:44] <TryDent1> flac app  does 24bit fine
[10:45] <Tjoppen> writing a general purpose encoder tool like ffmpeg is harder, so no big surprise that there are corner cases where it behaves in unexpected ways
[10:46] <Tjoppen> doesn't it print a warning though?
[10:46] <TryDent1> i have a question  for lossy compression formats;   how do you know if  it's using  24bit or 16bit or even 8 bit
[10:51] <TryDent1> any idea?
[10:51] <Tjoppen> it depends. most don'thave bitdepths per se
[10:51] <Tjoppen> in some cases you can tell the decoder to use float instead of 16-bit
[10:52] <Tjoppen> you'd have to be more specific
[10:52] <TryDent1> lossy compression shows  sample rate such as 44.1khz
[10:56] <relaxed> TryDent1: it's listed in the stream info. Stream #0.0: Audio: mp3, 44100 Hz, stereo, s16, 252 kb/s
[10:57] <relaxed> s16 = 16bit  Look at `ffmpeg -sample_fmts`
[11:02] <TryDent1> relaxed  what is the command to see the stream info
[11:04] <relaxed> ffmpeg -i input
[11:05] <TryDent1> wow 
[11:06] <relaxed> there's also ffprobe
[11:06] <TryDent1> okay i am going to test this with  24bit mp3
[11:06] <TryDent1> this better work
[11:07] <tdr> for osx, is it "better" to just compile ffmpeg or use macports to get it?
[11:07] <TryDent1> lol fail
[11:07] <TryDent1> Stream #0:0: Audio: vorbis, 96000 Hz, stereo, s16, 0 kb/s
[11:08] <TryDent1> wow so many bugs
[11:10] <relaxed> You want libvorbis, not vorbis. Until you know what you're doing please stop with the fail/bug comments.
[11:11] <TryDent1> i didn't use ffmpeg to  convert  24bitaudio to  vorbis-ogg
[11:12] <relaxed> If you have a problem, go to pastebin and paste your command and all output.
[11:13] <TryDent1> R:\>ffprobe 24bit.mp3
[11:13] <TryDent1> Stream #0:0: Audio: mp3, 44100 Hz, stereo, s16, 224 kb/s
[11:14] <TryDent1> let me test 8bit.mp3
[11:16] <Tjoppen> unless I'm mistaken mp3 doesn't have bitdepth
[11:18] <TryDent1> not just mp3;  all the lossy format;  i cannot seem to find  bitdepth
[11:18] <Tjoppen> also, there's the mp3float decoder
[11:19] <TryDent1> R:\>ffprobe 8bit.mp3     Stream #0:0: Audio: mp3, 44100 Hz, stereo, s16, 128 kb/s
[11:20] <TryDent1> relaxed failed again
[11:22] <Tjoppen> it should be possible to use mp3float if you need more precision. I don't know how though
[11:24] <TryDent1> am i just using the wrong/buggy version or what
[11:26] <ubitux> 24 bit mp3? heh?
[11:27] <Tjoppen> that's the sentiment I've been trying to get across
[11:28] <TryDent1> ubittux  bitdepth exist for lossless formats
[11:29] <ubitux> can mp3 be lossless? :p
[11:29] <TryDent1> Format/Info                      : Free Lossless Audio Codec
[11:29] <TryDent1> Duration                         : 8s 169ms
[11:29] <TryDent1> Bit rate mode                    : Variable
[11:29] <TryDent1> Bit rate                         : 101 Kbps
[11:29] <TryDent1> Channel(s)                       : 2 channels
[11:29] <TryDent1> Sampling rate                    : 44.1 KHz
[11:29] <TryDent1> Bit depth                        : 8 bits
[11:29] <Tjoppen> again, there's mp3float. just figure out how to use it
[11:33] <TryDent1> tjoppen link?
[11:34] <Tjoppen> I don't know
[11:35] <Tjoppen> I just know there's a float decoder and it should be possible to use it somehow
[11:36] <TryDent1> tjoppen which version fixes but 2 byte bug
[11:37] <Tjoppen> you could use 2c4e08d89327595f7f4be57dda4b3775e1198d5e~1 if you don't want the fix that increased the size of the header
[11:38] <Tjoppen> which is 582f231142c62a0bd6391efbd5a2ac119d73bb40
[11:38] <TryDent1> huh
[11:40] <relaxed> to decode with mp3float you would use `ffmpeg -acodec mp3float -i input.mp3 ...`
[11:40] <TryDent1> and what exactly does that do
[11:42] <ubitux> from my understanding, you decide how to decode the mp3: default is using 16 bits samples, but you could use 24 bits (i'm not sure it would make sense) or float; the above command says: use mp3float to decode the input mp3
[11:42] <ubitux> afaiu, but i may be wrong
[11:43] <TryDent1> what if you have 8bit.mp3
[11:44] <spyworldxp> how to select all the files inside the directory and convert it?
[11:44] <relaxed> what kind of files?
[11:44] <spyworldxp> mov to avi
[11:45] <ubitux> for f in dir/*.mov; do ffmpeg -i $f ${f%mov}.avi; done
[11:45] <ubitux> ?
[11:45] <spyworldxp> yes
[11:46] <spyworldxp> why linux video converter faster than windows?
[11:46] <ubitux> TryDent1: mp3 is lossy, so i don't think the bit depth actually matters
[11:46] <ubitux> i mean i'm not sure the information is actually in it
[11:47] <TryDent1> spyworldxp how much faster
[11:47] <spyworldxp> Windows take 1hour 30mins. Linux just 20mins
[11:47] <TryDent1> wow big difference; are you converting exactly same video
[11:48] <ubitux> "Technically speaking, bit depth is only meaningful when applied to pure PCM devices. Non-PCM formats such as lossy compression systems like MP3, have bit depths that are not defined in the same sense as PCM."
[11:48] <TryDent1> with exact same settings
[11:48] <ubitux> (https://en.wikipedia.org/wiki/Audio_bit_depth)
[11:55] <Tjoppen> spyworldxp: asm disabled perhaps?
[11:56] <spyworldxp> no
[11:56] <spyworldxp> i know the different
[11:56] <spyworldxp> linux using mpeg4 and windows using h264 (high)
[11:58] <spyworldxp> for f in dir/*.mov; do ffmpeg -i $f ${f%mov}.avi; done --> How to add h264?
[12:23] <spyworldxp> hi
[12:23] <spyworldxp> Unrecognized option c:a
[12:23] <spyworldxp> how?
[12:54] <torii> Hi.
[12:55] <torii> I have a question, install of ffmpeg.
[12:56] <torii> I installed libx264 that version is "0.118.x", "x264-snapshot-20111124-2245-stable".
[12:57] <torii> But cannot install last-ffmpeg.
[12:58] <torii> Console display, http://pastebin.com/mxzUvtQY.
[12:58] <sacarasc> Do you have any other versions of x264 installed? And did you install the dev files?
[12:59] <torii> yes, ihave installed another version of x264.
[13:01] <torii> cannot override x264 install?
[18:01] <njbair> is ffmpeg built to take advantage of 64 bits?
[18:35] <Diogo> hi
[18:36] <Diogo> hi i have installed ffmpeg and i want to stream a mp4 file to rtmp..(wowza server)
[18:36] <Diogo> this is possible..?
[18:40] <kakapo> yes
[18:45] <Diogo> and using mjpeg to rtmp
[18:45] <Diogo> ?
[18:46] <Diogo> i want o get the video fom a ip camera..
[18:47] <spm-Draget> I have an interlaed mpeg2 video source and encode it to xvid using 'ffmpeg -i test.mpg -f avi -vcodec mpeg4 -vtag xvid -flags +ilme+ildct -g 50 -qmin 1 -qmax 2 -ab 256 test.avi'. Especially the flags I am confused about. I want the resulting xvid video to be flagged and encoded as interlaced, since I want to deinterlace it later.
[18:48] <Diogo> i'm using this configuration but don't stream..
[18:48] <Diogo>   built on Nov 25 2011 16:48:34 with gcc 4.6.1
[18:48] <Diogo>   configuration: --enable-gpl --enable-libfaac --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libtheora --enable-libvorbis --enable-libx264 --enable-nonfree --enable-postproc --enable-version3 --enable-x11grab
[18:49] <Diogo> i'm using this configurations
[18:55] <spm-Draget> Okay, different question: Can I extract only the videostream from a file without reencoding?
[18:57] <mateo`> spm-Draget: ffmpeg -i source -vcodec copy -an dest ?
[19:01] <spm-Draget> mateo`: Hmm, this makes fcchandlers mpeg2 plugin more happy, indeed. Thanks
[19:05] <Diogo> anyone use ffmpeg to do a restreming?
[20:33] <Takyoji> Just need to switch container formats. How do I preserve the video and audio without re-encoding?
[20:33] <burek> -acodec copy -vcodec copy
[21:54] <pythonirc101> is there a way to tell ffmpeg to output in standard form instead of using fancy output (curses etc)
[22:35] <jimlkl> Here's  the ffmpeg- i information from a video I downloaded. (pastebin will follow after this post) It
[22:35] <jimlkl> transfers and plays on the iPod BUT the width of the video is partially
[22:35] <jimlkl> cut off on the iPod screen. Any idea why?  Please let me know how to fix it.
[22:36] <jimlkl> Here's the pastebin
[22:36] <jimlkl> http://pastebin.com/2TbLeVq9
[22:39] <pasteeater> jimlkl: does it look cut off with ffplay too? ffplay rp.mp4
[22:42] <jimlkl> hang on .....I'll give it a try.
[22:43] <jimlkl> No, it doesn't.  It plays perfectly.
[22:46] <pasteeater> what does quicktime say about the video? there's a menu with "video information" or something like that
[22:47] <jimlkl> I don't know.....I mostly run on Ubuntu 10.04 and would have to reboot to go to Windoz
[22:48] <jimlkl> It plays fine in iTunes 10.5 but gets cut off on the iPod
[22:48] <jimlkl> Wait....
[22:48] <jimlkl> geez.....I can't remember if it gets cut off in iTunes.
[22:48] <jimlkl> My head is a sive.
[22:49] <pasteeater> looks like ffmpeg is fine with it then. blame quicktime/itunes/iDevice/apple.
[22:49] <jimlkl> 'Had a disk crash and had to reinstall Windows and restore all of my music and videos to stupid itunes.
[22:50] <jimlkl> I download the videos from Youtube and try and put 'em on the iPod.
[22:50] <jimlkl> Previous to THE CRASH of my disk drive I had no problem at all.
[22:50] <jimlkl> Now....nothing but problems.
[22:51] <burek> jimlkl
[22:51] <burek> can you please use pastebin.com, to show your command line and its output?
[22:52] <jimlkl> Oh boy.....that's a tough one......I don't even remember it nor could I recreate it.  When it comes to FFMPEG I'm a moron and the best I can do is ask people for a code line to convert to stuff to ipod compatible format.
[22:53] <pasteeater> did you try the command i gave you on ubuntuforums yet?
[22:53] <jimlkl> Surely in this wicked world there has to be one simple method to ffmpeg a video file (mp4) to a ipod format.
[22:53] <jimlkl> Are you FakeOutdoorsman?
[22:54] <pasteeater> that bastard?
[22:54] <pasteeater> yes
[22:54] <pasteeater> http://ubuntuforums.org/showpost.php?p=11486441&postcount=6
[22:54] <jimlkl> YOU ARE THE MAN!!!!!!
[22:54] <jimlkl> Hang on....let me get the code line...
[22:56] <jimlkl> Is this it......ffmpeg -i cba.mp4 -vcodec libx264 -preset medium -vpre ipod640 -crf 24 -acodec libfaac -aq 100 FAKEOUTD.mp4
[22:57] <jimlkl> I've tried so many different code lines I don't remember if yours worked.  I think it did....but not sure.
[22:58] <jimlkl> Hang on....I'll check it again.
[22:58] <jimlkl> By the way.....are you really in Alaska?
[22:58] <pasteeater> yes
[22:58] <jimlkl> Wow.
[22:58] <pasteeater> if your input is too big for your ipod then add: -vf scale="640:trunc(ow/a/2)*2"
[22:58] <pasteeater> it will auto scale to an acceptable size.
[22:59] <jimlkl> Yes....the code line above worked perfectly.
[23:00] <jimlkl> ....that is....the FAKEOUTD.mp4 line
[23:00] <jimlkl> What do you mean by "if your input is too big?
[23:02] <pasteeater> you ipod probably can't play back videos larger than 640x480 or sometihng like that
[23:02] <jimlkl> Using the auto scale you mentioned just above is this how the line should look?  ffmpeg -i cba.mp4 -vcodec libx264 -preset medium -vpre ipod640 -crf 24 -acodec libfaac -aq 100  -vf scale="640:trunc(ow/a/2)*2" FAKEOUTD.mp4
[23:03] <Diogo> hi anyone know how can i restream to a ip camera to wowza server..
[23:03] <Diogo> using ffmpeg?
[23:05] <pasteeater> jimlkl: yes
[23:05] <jimlkl> Alright.....I'll try it.
[23:05] <jimlkl> Is there a way to speed up the conversion process as it takes a long time to convert a 50 meg file?
[23:08] <pasteeater> use a faster preset. you can see a preset list with x264 --help
[23:10] <jimlkl> Oh boy....that stuff is so confusing to me.
[23:12] <pasteeater> presets in order of speed: ultrafast, superfast, veryfast, faster, fast, medium, slow, slower, veryslow
[23:13] <jimlkl> Fake.....thanks so much for your help.  You really are a great help here and over at ubuntuforums.
[23:14] <jimlkl> Peace.
[23:14] <jimlkl> Paste......would you re-send that last line you wrote to me?
[23:17] <Hyperi> 00:12:53 < pasteeate> presets in order of speed: ultrafast, superfast, veryfast, faster, fast, medium, slow, slower, veryslow
[23:17] <Hyperi> yw :P
[23:18] <jimlkl> thanks
[23:30] <wolfman2000> Afternoon. I am trying to compile ffmpeg n0.8.7, and I get failures on the linking process. My issue is in the pastebin. http://pastie.org/2921531  If it helps, I'm trying to build on Mac OS X 10.7 Lion.
[23:32] <pasteeater> wolfman2000: i'm ignorant of osx, but your ./configure would probably be useful for others who are familiar with osx.
[23:33] <wolfman2000> pasteeater: same paste link has been updated. If it will also help, I'm starting with a configure provided by an old build script for a project I help contribute to.
[23:34] <burek> Diogo, what command did you try so far
[23:37] <wolfman2000> pasteeater: what may help is that Lion is a 64 bit OS by nature, but can also build 32 bit
[23:39] <burek> wolfman2000, did you try: https://www.google.com/search?q=%22Undefined+symbols+for+architecture+x86_64%22
[23:39] <wolfman2000> burek: yes. right now reading
[23:40] <ghostbar> hey! I want to grab a rtsp stream a put it on x.jpg but the only way I found is to use -vframes 1 and re-launch ffmpeg. Is there a way to make it rewrite the jpeg without re-lunching ffmpeg?
[23:40] <burek> yes
[23:41] <burek> add -y
[23:41] <burek> wait
[23:41] <ghostbar> i used something like ffmpeg -i foo.avi -r 1 -s WxH -f image2 foo.jpeg
[23:41] <burek> can you please use pastebin.com, to show your command line and its output?
[23:41] <burek> ok
[23:41] <ghostbar> burek: it doesn't works
[23:41] <burek> so
[23:41] <burek> ffmpeg -loop 1 -i foo.avi -r 1 -s WxH -f image2 foo.jpeg
[23:41] <ghostbar> let me tell you what gives me
[23:41] <burek> that works
[23:41] <burek> wait wait.. my bad
[23:42] <burek> ffmpeg -re -i foo.avi -r 1 -s WxH -f image2 foo.jpeg
[23:43] <burek> that should work
[23:44] <wolfman2000> burek: going to try http://ffmpeg.org/trac/ffmpeg/ticket/469  and make it use clang
[23:45] <ghostbar_> burek: it gives me [image2 @ 0x9c20820] Could not get frame filename number 2 from pattern 'what.jpg'
[23:45] <ghostbar_> that's the same output with -y
[23:45] <ghostbar_> and without -y
[23:45] <ghostbar_> ffmpeg -re -y -i shell-20110908-1.webm -r 1 -f image2 -s qvga what.jpg
[23:46] <ghostbar_> that's the line i'm trying with
[23:46] <wolfman2000> burek, pasteeater: --cc=clang to ./configure did the job. At least I got no errors.
[23:46] <burek> ghostbar, it seems that ffmpeg expects at least one % param (sprintf)
[23:46] <burek> so we need to trick it somehow
[23:53] <burek> ghostbar_
[23:53] <burek> you know what you could do
[23:53] <pythonirc1011> can ffmpeg use nvidia CPUs? seems like for my AVCHD to mpeg4 conversion, its pretty slow on my quadcore
[23:53] <burek> since ffmpeg doesn't allow you to just set the same filename
[23:53] <burek> it requires you to put %d into your filename
[23:54] <burek> you could set filename like 'output%d.jpg'
[23:54] <burek> that would presumably go from 1-9 (or 0-9)
[23:54] <sacarasc> pythonirc1011: CPUs or GPUs?
[23:54] <burek> and then loop
[23:54] <burek> so, you just set -r 10 (instead of 1)
[23:54] <burek> and read one of those 10 files you want
[23:54] <burek> for example output1.jpg
[23:55] <burek> do you get my point?
[23:55] <wolfman2000> ...sad part here is that I'm now stumped. I don't know which of the .a files I now need for my project.
[23:55] <burek> wolfman2000 :beer: :)
[23:55] <burek> well, at least you are never bored, right? :D
[23:55] <wolfman2000> ...something like that
[23:57] <burek> pythonirc1011, im not sure really, i think ffmpeg cannot (so far) use gpus
[23:57] <burek> but, vlc has got an option for ffmpeg, namely --ffmpeg-hw
[23:57] <burek> This allows hardware decoding when available. (default disabled)
[23:57] <burek> so, I don't know what to answer :)
[23:58] <pythonirc1011> i'm looking to convert a AVCHD to MPEG4 that is playable on win 7
[23:58] <pythonirc1011> its using all my quad cpus and still takes a lot of time because the bitrate is high
[23:58] <burek> are you using libx264
[23:59] <pythonirc1011> yes
[23:59] <pythonirc1011> my camera claims to output 24Mbps. I'm using 12Mbps
[23:59] <burek> did you read x264 --help, especially for -preset and -tune
[00:00] --- Sat Nov 26 2011