[Ffmpeg-devel-irc] ffmpeg.log.20120223

burek burek021 at gmail.com
Fri Feb 24 02:05:02 CET 2012


[00:08] <tbws> ty vadim_ i didnt have the partner's listed
[00:08] <tbws> :( *sigh*
[00:09] <tbws> installed it from a debian package now anyways
[00:30] <pasteeater> you don't need partners for libfaac-dev
[00:31] <pasteeater> it's in the "multiverse" repository, and installing from a deb package isn't recommended.
[00:54] <tbws> pasteeater: then the question of why it wasn't available to me is still up in the air, i have the multiverse reps listed.
[00:54] <tbws> All is installed now
[00:54] <tbws> just having a problem executing it
[00:54] <tbws> i'm trying to use ffmpeg to encode a video for mobile safari
[00:54] <pasteeater> you could paste the message you get when you tried to: sudo apt-get install libfaac-dev
[00:54] <pasteeater> use pastebin.com
[00:54] <tbws> Couldn't find package
[00:54] <pasteeater> that's it?
[00:54] <tbws> yup
[00:54] <pasteeater> what about your sources.list and anything in /etc/apt/sources.list.d/
[00:54] <tbws> all i have are the universe and multiverse reps listed
[00:54] <tbws> my soources.list looks like vadim_'s only i've stripped the comments out.
[00:54] <pasteeater> did you run "sudo apt-get update" after editing the file?
[00:54] <tbws> yep
[00:54] <pasteeater> what about /etc/apt/sources.list.d/
[00:54] <vadim_> tbws: there was no errors while updating?
[00:54] <tbws> nope
[00:54] <vadim_> tbws: may be you have broke some links when stripped the comments?
[00:54] <tbws> vadim_ i checked it against yours
[00:54] <tbws> they are correct
[00:54] <tbws> its just plain freakin wierd
[00:54] <tbws> but i'm having an encoding problem at the moment http://pastebin.com/chR96YAh
[00:54] <pasteeater> how can we guess what you may have done wrong if you don't show your command as well?
[00:55] <tbws> true
[00:55] <tbws> "ffmpeg -i 19-2.flv -vcodec libx264 -preset hq -preset ipod640 -b 250k -bt 50k -acodec libfaac -ab 56k -ac 2 -s 480x320 19-2.mp4"
[00:55] <pasteeater> there is no preset named hq. there is no preset named ipod640 (you use that via -vpre).
[00:57] <tbws> i have installed those presets
[00:57] <pasteeater> that's not how presets work
[00:57] <pasteeater> you don't need to grab preset files from random locations. they probably won't work.
[00:58] <pasteeater> ...libx264 -preset slow -vpre libx264-ipod640 ...
[00:59] <tbws> ok cool, thank you very much
[00:59] <tbws> that seems to be encoding at the moment
[01:05] <tbws> Is there a way to maintain the 16:10 aspect ratio of the video being encoded?
[01:06] <PovAddict> is 16:10 the DAR or SAR?
[01:07] <tbws> no idea
[01:07] <tbws> lets say i wanna crop the video down until the width fits the width that i want the video to be converted to.
[01:07] <PovAddict> DAR is the aspect ratio of the video as displayed, SAR is the aspect ratio of the pixels forming the stored image; they aren't necessarily the same
[01:08] <PovAddict> erm
[01:08] <PovAddict> SAR is the aspect ratio of the stored image, width/height in pixels
[01:08] <PovAddict> (if they aren't the same, then the pixels are not square)
[01:09] <tbws> ah ok, the DAR in that case
[01:09] <pasteeater> tbws: you can use the scale filter to preserve aspect. you give it your desired width or height and it will provide a width or height to preserve aspect.
[01:10] <pasteeater> scale=480:-1 for example.
[01:10] <pasteeater> or scale="480:trunc(ow/a/2)*2" if -1 returns an odd number which x264 dislikes.
[01:11] <tbws> How would you write that into my command string though? "ffmpeg -i 19-2.flv -vcodec libx264 -preset slow -vpre ipod640 -scale="480:trunc(ow/a/2)*2" -b 250k -bt 50k -acodec libfaac -ab 56k -ac 2  -s 480x320 19-2.mp4" ??
[01:12] <PovAddict> it's -vf "scale=480..."
[01:12] <PovAddict> vf means video filter
[01:13] <tbws> thank you so much guys, you've been really helpfull
[01:13] <pasteeater> libx264 ignores -bt
[01:13] <PovAddict> what's -bt?
[01:15] <pasteeater> "bitrate tolerance"
[01:16] Action: PovAddict plays around with the x264 CRF
[01:16] <pasteeater> tbws: unless you're trying to get a specific output file size, then you should probably use -crf instead of -b
[01:16] <PovAddict> crf seems a bit non-linear...
[01:16] <PovAddict> how does it work, roughly?
[01:17] <PovAddict> using 40 instead of 20 produces something much much worse than what I'd call "half quality"
[01:18] <pasteeater> CRF is an arbitrary constant quality: a ratecontrol mode without a target bitrate
[01:19] <pasteeater> it's logarithmic, as opposed to qscale for mpeg-*
[01:19] <pasteeater> which is linear
[01:20] <pasteeater> +6 is roughly half the bitrate with crf when using 8-bit x264. i'm not familiar with 10-bit so i'm unsure how it changes for that.
[01:20] <PovAddict> so 20 has be N times the quality of 21? (where N may very well be < 2)
[01:20] <PovAddict> s/be//
[01:21] <pasteeater> basically just use the highest value that still gives you an acceptable quality.
[01:21] <pasteeater> a good start is 24.
[01:22] <PovAddict> I wish I had a good way to compare the quality of original and encoded
[01:24] <PovAddict> pasteeater: 23 is the default, so that gives me an idea already
[01:24] <PovAddict> pasteeater: my source material is low quality to begin with, but I don't want to make it any worse
[01:24] <pasteeater> youll just have to look at it and compare for yourself.
[01:25] <PovAddict> re: comparison, I meant I wish there was a way to play them in sync and switch between them with one click :)
[01:28] <PovAddict> hmm... starting two players at the same time gave me surprisingly good sync
[01:30] <PovAddict> wow, it's encoding faster than realtime
[01:30] <PovAddict> faster than mpeg2 too, due to better threading :)
[01:41] <PovAddict> pasteeater: trying crf 25, the encoded video looks better than the source because it smoothed out the analog color noise :D
[01:43] <pasteeater> what is the original format?
[01:44] <PovAddict> 8mm analog tape, which I transcoded into deinterlaced .dv around a year ago
[01:46] <nextime> is there any particular option to tell ffmpeg to use neon on omap cpu?
[01:46] <pasteeater> PovAddict: i have some of that here. really noisy due to low light. yadif and hqdn3d helped a bit.
[01:47] <PovAddict> unfortunately DV is not lossless
[01:48] <PovAddict> and removing noise *might* have been easier on the interlaced data
[01:48] <PovAddict> but going back to the source tapes isn't an option, I don't have a working device for that anymore :P
[01:49] <pasteeater> might find one for $5 at a thrift store.
[01:50] <PovAddict> our camera died around 10 years ago, so last year I borrowed one to do these conversions, gave it back already
[01:50] <PovAddict> there is also one particular tape that took me several tries to rip
[01:51] <PovAddict> after a few minutes the image slowly started to get garbled; somehow I'd get it to work again and copy another piece
[01:52] <PovAddict> concatenating the pieces was some work too :)
[01:53] <PovAddict> I'm afraid that if I touch that tape again it will suddenly break down into fine rust
[03:18] <slackyman> hi
[03:19] <slackyman> is there a mode to reverse a strip command?
[03:19] <slackyman> is there a way tyo revert stripping?
[03:23] <slackyman> anyone?
[04:02] <slackyman> is there a way tyo revert stripping?
[04:04] <PovAddict> what's stripping?
[04:09] <slackyman> strip an executable, a binary file
[04:11] <PovAddict> err no
[04:11] <PovAddict> strip deletes stuff from inside the executable
[04:12] <slackyman> ok
[04:12] <PovAddict> you undo it by restoring your pre-strip backup :P
[04:12] <slackyman> thanks
[04:12] <PovAddict> or compiling the binary again
[04:12] <slackyman> I'm "making-install" again :)
[04:13] <slackyman> thanks
[04:13] <slackyman> bye
[04:13] <Ultimatespirit> So this is probably a stupid question? but does ffmpeg include libx264?
[04:14] <PovAddict> ffmpeg is distributed as source code, and it doesn't "include" anything that is elsewhere (it's bad practice to embed other libraries since they'd go inevitably out of date)
[04:14] <PovAddict> if you mean a ffmpeg binary you got somewhere, that's not something we'd know :)
[04:15] <Ultimatespirit> ah, since i'd seen examples of using libx264 as a codec and I was wondering if I can actually use it
[04:15] <PovAddict> yes, ffmpeg can use libx264
[04:15] <PovAddict> and many other libraries
[04:15] <PovAddict> use != include
[04:15] <Ultimatespirit> then I guess i messed something up since it gives me errors when i try it...
[04:15] <Ultimatespirit> != include
[04:15] <Ultimatespirit> !=include
[04:16] <PovAddict> I meant "use isn't the same as include"
[04:16] <Ultimatespirit> ah
[04:16] <PovAddict> are you compiling ffmpeg from source?
[04:16] <Ultimatespirit> yes
[04:16] <Ultimatespirit> i compiled it from freedom
[04:16] <Ultimatespirit> but can use a git build too if it's something that wasn't in freedom
[04:16] <PovAddict> what's "freedom"?
[04:17] <Ultimatespirit> 0.10 release
[04:17] <PovAddict> ah
[04:17] <PovAddict> well
[04:17] <PovAddict> install libx264, pass --enable-libx264 to the configure script when compiling ffmpeg :)
[04:17] <Ultimatespirit> that gives error saying it isn't found
[04:17] <Ultimatespirit> which is why I was worried that I was missing something
[04:17] <PovAddict> what linux distro do you have?
[04:17] <Ultimatespirit> using mac osx so that's probably a problem?
[04:18] <PovAddict> ah okay
[04:18] <PovAddict> so you installed libx264 from source as well?
[04:18] <Ultimatespirit> ah, so i need to do that? because that gave me entirely different errors that I cannot figure out
[04:18] <PovAddict> ffmpeg doesn't include x264
[04:18] <PovAddict> it can use x264 if you have it
[04:19] <Ultimatespirit> ah, that explains a lot. Thanks.
[04:19] <Ultimatespirit> Now to find out what went wrong wiht libx264 huh...
[04:23] <edrx> I am puzzled by this behavior: when I run this to extract the 1-minute
[04:23] <edrx> segment between 0:30 and 1:30 from a very heavy .MOV file (20 mins, almost 1GB),
[04:23] <edrx> "ffmpeg -i DSC_0004.MOV -ss 0:00:30 -t 0:01:00 /tmp/clip.MOV"
[04:23] <edrx> I am getting a low-quality .MOV, with about 3.7MB, instead of a .MOV with the original quality and about 50MB...
[04:23] <edrx> the manpage says "By default ffmpeg tries to convert as losslessly as possible: It uses the same audio and video parameters for the outputs as the one specified for the inputs" - so this is a case where the default is _not_ to be lossless. What is happening? Any hints (esp. on how to do lossless extraction)? 8-(
[04:23] <PovAddict> edrx: huh, that quote seems crap :P
[04:24] <edrx> is there an option to ask for losslessness? =)
[04:24] <PovAddict> if you want it to be *really* lossless, use -vcodec copy -acodec copy, but then you can only extract down to nearest keyframe, which may be quite a few seconds away from the time offsets you want
[04:25] <PovAddict> otherwise, just increase the video and audio bitrate; the default video bitrate in ffmpeg is ridiculously low
[04:27] <edrx> how do I increase it to, hmm, 50MB per minute -> 1MB per second?
[04:27] <PovAddict> -b 8000k
[04:27] <PovAddict> (8 megabits per second)
[04:28] <edrx> trying
[04:43] <edrx> PovAddict: excellent, thanks =)! and I've just solved my other problem too, which was to find a way to reduce the framerate to 6 FPS to save more space... I am converting the recordings of a maths talk, so 6 FPS is ok, but the blackboard needs to appear in hi-res
[04:44] <edrx> this did the job: "ffmpeg -i DSC_0004.MOV -b 10000k -r 6 -ss 0:00:30 -t 0:01:00 /tmp/clip-high6.avi"
[05:26] <tbws> i'm trying to write a bash script that will encode a file for all of the various html5 video formats and i'm having difficulty choosing what the params should be
[05:26] <tbws> http://pastebin.com/
[05:26] <tbws> could someone please help me out?
[05:26] <tbws> *http://pastebin.com/TYj2qBa9
[05:28] <relaxed> tbws: "ls $1 | while read file; do" is wrong, use --> for i in "$@"; do
[05:29] <edrx> I prefer this: "for file in *; do ...; done"
[05:30] <edrx> oops, that was practically the same as relaxed's
[05:31] <edrx> (I am a newbie, I can say very little about parameters)
[05:36] <relaxed> tbws: also, that's a sh script
[05:38] <CarlFK> how do I see what the codec names are?
[05:39] <zap0> ffmpeg -codecs
[05:39] <zap0> or    --codecs
[05:39] <CarlFK> thanks
[06:09] <kevin_> hi all. so i have a video file and i need to extract the video stream to one file and the audio stream to another file and i want the video to have no audio in it. if i use the -an flag, it errors because then the audio stream doesn't exist to put in.. any ideas? command-line https://gist.github.com/1890453
[06:09] <kevin_> using that will cause cut/$file.mpg to have both the audio and video streams, but i only need it to have the video stream
[06:13] <relaxed> kevin_: why do you want mpeg2 video and mp3 output?
[06:14] <kevin_> relaxed: i'm normalizing the audio using normalize-mp3
[06:15] <kevin_> then rejoining the two streams
[06:16] <relaxed> output to wav and use normalize-audio instead
[06:16] <relaxed> you won't lose as much quality
[06:20] <relaxed> I would encode both streams together (with raw audio), dump the wav, normalize it and the mux the normalized audio and mpeg2 video together.
[06:20] <kevin_> hmm
[06:22] <relaxed> well, encode the normalized audio to ac3/mp2 while muxing with the video.
[06:23] <kevin_> ok yeah
[06:23] <relaxed> the day ffmpeg gets a normalization filter will be a wonderful day indeed.
[06:27] <kevin_> well, any idea on how to just have the intermediate video file not have the audio stream? seems unnecessary to have ffmpeg to have to encode it to ac3 when it's not gonna be used
[06:29] <relaxed> not ac3, use pcm_s16le so there won't be any quailty lose. This will help you maintain sync as well.
[06:30] <relaxed> s/lose/loss/
[06:30] <kevin_> which is the raw audio correct?
[06:30] <relaxed> yes
[06:31] <kevin_> even if the intermediate uses mpeg2video+raw, it still seems unnecessary to have the audio in there...
[06:32] <kevin_> ie, why have one mpeg2video+wav, then dump that wav, when i could just have one mpeg2video and one wav...?
[06:32] <kevin_> (if i'm explaining that right)
[06:34] <relaxed> then use -f mpeg2video and -f wav to separate the two
[06:36] <kevin_> ah, ok. great.
[11:12] <rainmaker1> Hello I am trying to caputre HD live stream (h264) to disk using ffmpeg -re -i udp://231.0.0.124:5000 -vcodec copy -acodec copy -y testing.mpg
[11:13] <rainmaker1> using vlc there is no video just audio
[11:13] <rainmaker1> Is h264 capturing supported using ffmpeg?
[11:14] <kriegerod> yes
[11:14] <rainmaker1> kriegerod: any hint?
[11:14] <kriegerod> i didn't understand exactly what's your problem
[11:15] <rainmaker1> kriegerod: When I capture live video stream (h264) to disk using ffmpeg -re -i udp://231.0.0.124:5000 -vcodec copy -acodec copy -y testing.mpg command
[11:15] <rainmaker1> kriegerod: and when I try to play that file using vlc, there is no video and I can hear audio
[11:16] <Mavrik> rainmaker1: provide ffmpeg output in a pastebin as ChanServ notified you when you joined please.
[11:17] <kriegerod> makes sense also to publish the output file of duration ~10 seconds
[11:17] <Mavrik> he's probably just remuxing it wrong
[11:17] <kriegerod> rainmaker1: did you try to playback the file using ffplay?
[11:18] <kriegerod> Mavrik: you mean container may be incompatible? Or just VLC support of mpg format has issues?
[11:18] <Mavrik> probably incompatible container
[11:19] <kriegerod> rainmaker1: try saving with extension .ts, or mp4
[11:21] <rainmaker1> Here is pastebin http://pastebin.com/ewLfTPsv
[11:22] <Mavrik> rainmaker1: your input is a MPEG2-TS stream which you're remuxing to MPEG, which could be the cause of your problems
[11:22] <Mavrik> try using -f mpegts output.ts for output file
[11:22] <Mavrik> also it seems your input is rather lossy :)
[11:25] <rainmaker1> Mavrik: When I try to play live stream (not from file) it says it's H264-MPEG-4 AVC(part10)(h264)
[11:26] <Mavrik> yeah
[11:26] <Mavrik> packaged into a MPEG-2 TS
[11:26] <Mavrik> that's standard for IPTV
[11:27] <Mavrik> also, "-bufsize" after "-i" is the wrong way to set input buffer size
[11:29] <rainmaker1> Mavrik: hmmm nope, still cirucula buffer overun error :( http://pastebin.com/nwcCccn8
[11:30] <rainmaker1> Mavrik: now I have video but everytime I get cirucular buffer overun error
[11:30] <Mavrik> ok good
[11:31] <Mavrik> that's probably because your datarate is too high
[11:31] <Mavrik> try using
[11:31] <rainmaker1> Mavrik: you mean I need a faster cpu?
[11:31] <Mavrik> udp://231.0.0.1?pkt_size=1316
[11:31] <Mavrik> for input
[11:32] <Mavrik> the circular buffer error seems quite common (And no you don't need a faster CPU)
[11:33] <rainmaker1> Mavrik: same :( http://pastebin.com/zt0i90uV
[11:33] <Mavrik> also try using ?buffer_size=150000 or something like that
[11:34] <Mavrik> to see if you still get buffer overruns
[11:35] <rainmaker1> nope
[11:35] <rainmaker1> same
[11:36] <rainmaker1> should I set -bufsize or this is the same as ?buffer_size ?
[11:37] <Mavrik> no
[11:37] <Mavrik> -bufsize is for the encoder and means something totaly different
[11:38] <kriegerod> try  udp://231.0.0.1?fifo_size=0buffer_size=10000000
[11:38] <kriegerod> try  udp://231.0.0.1?fifo_size=0&buffer_size=10000000
[11:38] <Mavrik> UDP input reads buffer_size tag after URL
[11:41] <rainmaker1> so far so good, but bitrate is now about 2000k and was 8000-11000k
[11:43] <rainmaker1> I could not imagine it's this hard to make for living from video :)
[11:43] <Mavrik> it's harder :P
[11:43] <rainmaker1> :D
[11:43] <Mavrik> 2Mbit/s is pretty standard bitrate for IPTV H.264 broadcasting
[11:43] <Mavrik> if it's SD
[11:44] <rainmaker1> no, it's HD
[11:44] <Mavrik> we even went as low as 1Mbit
[11:45] <rainmaker1> that's exactly what I want to do, but as far as I can't capture sample video to make some benchmarks I want go far away :)
[11:53] Action: nextime ha issues with mpeg4 decoding
[11:53] <nextime> *has
[11:54] <Mavrik> which mpeg4? :)
[11:58] <nextime> Mavrik : from an ip cam
[11:58] <nextime> http://nopaste.info/88d2a1bda6_nl.html
[12:00] <nextime> but, it has those errors only on a omap3 cpu ( with neon )
[12:00] <nextime> on x86 it work
[12:00] <nextime> ( same version of ffmpeg and whole things )
[12:01] <rainmaker1> Mavrik: strange, but when I remove -re I have no circular buffers errors...
[12:01] <Mavrik> -re?
[12:02] <Mavrik> nextime: hmm... mpeg4 decoder tends to spew out those errors when it can't keep up with the input
[12:02] <Mavrik> nextime: try specifying a bigger buffer... is your CPU fast enough for that transcode?
[12:02] <rainmaker1> Mavrik: yes, -re: Read input at native frame rate. Mainly used to simulate a grab device.
[12:02] <nextime> Mavrik : so, do you think it is a performances issue>
[12:02] <nextime> ?
[12:02] <Mavrik> nextime: it's possible, why aren't you passing speed preset to libx264?
[12:03] <Mavrik> -preset ultrafast
[12:03] <nextime> Mavrik : i was, this is a "simplified" command line
[12:03] <nextime> cause i was trying to understand the bug
[12:04] <Mavrik> well, the encoder is complaining about broken input
[12:04] <Mavrik> I've had those errors when the buffers were overrun
[12:04] <Mavrik> which could be your problem since you're switching to an obviously slower CPU
[12:05] <Mavrik> do you get correct input if you just dump it to a file?
[12:05] <nextime> Mavrik : it isn't so slow
[12:05] <nextime> as it is 1Ghz clock
[12:05] <nextime> and also, it work well with more huge streams that aren't mpeg4
[12:06] <Mavrik> 1GHz ARM is slow when it comes to video encoding.
[12:06] <Mavrik> especially if you're trying to use x264
[12:06] <nextime> Mavrik : on the x264side it work great with neon
[12:07] <nextime> i can encode in realtime from an mjpeg stream that is as double resolution than the mpeg4 one
[12:07] <nextime> and also from the same resolution with an mpeg2 stream
[12:07] <nextime> with very low cpu load
[12:07] <Mavrik> hmmf
[12:07] <Mavrik> interesting that
[12:07] <Mavrik> there IS a chance that mpeg4 decoder is broken on ARM
[12:08] <Mavrik> however I'd check input buffers first
[12:09] <nextime> ( and -preset ultrafast doesn't solve )
[12:10] <nextime> what make me think to an mpeg4 decoder issue is than with every other stream than mpeg4 with the very same options on the encoding side
[12:10] <nextime> it work well enough
[12:10] <Mavrik> mhm
[12:10] <Mavrik> nextime: do you have the same transport for all? meaning, could it be a demuxer issue?
[12:11] <nextime> Mavrik : yes, it could be
[12:11] <nextime> the transport isn't the same
[12:12] <nextime> wait, i have the same stream on rtsp over http inside an afs file other than the rtsp i'm trying
[12:13] <Mavrik> (I've had rather bad experience concerning FFMpeg's network muxers/demuxers)
[12:15] <nextime> uhmm
[12:15] <nextime> same stream but with rtsp encapsulated in http and asf, it seems to work
[12:17] <Mavrik> bbl, need to install some hardware to my workstation :P
[12:17] <nextime> ok no
[12:18] <nextime> it *seems* to work at first
[12:18] <nextime> but same issue on the second try
[14:19] <Qcoder00> Morning
[14:19] <Qcoder00> Or Afternoon
[14:19] <Qcoder00> :)
[14:20] <Qcoder00> Does anyone know how you set up ffmpeg on Windows XP?
[14:20] <Qcoder00> I get an error from Windows Media Player saying it can't play DVD's
[14:20] <Mavrik> uum
[14:20] <Qcoder00> And I am sure this is a missing codec issue
[14:20] <Mavrik> ffmpeg isn't what you want probably
[14:20] <Qcoder00> What do I need then?
[14:21] <Mavrik> try using an alternative player (Media Player Classic Home Cinema, VLC, etc.) or install a decoder (ffdshow is a popular one)
[14:22] <Qcoder00> The error I get in Windows Media Player is that it claims there is no 'compatible' decoder
[14:22] <Qcoder00> which is silly as I thought XP could play DVD out of box
[14:22] <Mavrik> afaik it doesn't
[14:22] <Mavrik> you're using an obsolete old OS anyway
[14:22] <Mavrik> so grab a player that can do it
[14:23] <Qcoder00> VLC works
[14:23] <Qcoder00> but it takes a VERY long time to decode stuff
[14:24] <Qcoder00> I'll try upgrading VLC and see it that cures the problem
[14:25] <Qcoder00> In the meantime where do I ask for ffdshow support?
[14:26] <JEEBsv> doom9's thread
[14:26] <JEEBsv> clsid and friends do not IRC
[14:27] <Mavrik> puny humans.
[14:34] <Qcoder00> Is doom9's thread a 'safe' site?
[14:34] <JEEBsv> doom9 is one of the oldest DVD ripping/video encoding related sites :P
[14:34] <JEEBsv> and doesn't have ads as far as I know
[14:34] <Qcoder00> OK
[14:35] <Qcoder00> So it's not a site that's going to get busted for showing people how to break the encryption NASA uses on downlinks or anything like that? ;)
[14:35] <JEEBsv> they only do video-related DRM research
[14:35] <JEEBsv> and it hasn't been busted yet
[14:35] <JEEBsv> during the... probably over 10 years of existence it has had?
[14:36] <Qcoder00> OK
[14:36] <Qcoder00> I have ethical issues with DRM in any case
[14:36] <Qcoder00> DRM =Don't Rip me!!
[14:37] <Qcoder00> DRM= doom9 reversed me ;)
[14:37] <Qcoder00> (Joke)
[14:42] Action: nextime hates gstreamer, but it seems that i'll have to use it instead of ffmpeg for what i'm doing...
[14:43] <ayaka> does ffmpeg support acceleration in some arm cpu like msm
[14:44] <nextime> ayaka : it has "some" support for neon on omap
[14:44] <nextime> but i'm going to gstreamer cause it doesn't support the TI DSP
[14:46] <ayaka> nextime, but i think i can't run a x with acceleration in it
[14:47] <ayaka> nextime only omap why neon is obly in it
[14:47] <nextime> ayaka : of course no, you need accelleration support in the X driver
[14:48] <ayaka> nextime, do you have sone info about that
[14:48] <nextime> ayaka : i don't know if it is only omap, i just know omap cause i'm using it
[14:49] <nextime> ayaka : just what you can find on google about linux on beagleboard and similar things
[14:50] <ayaka> nextime, it means you have install a linux on mobile, do you have time and offtopic channel, i want to ask something
[14:57] <nextime> ayaka : i just have some beagleboards and similar boards with omap and my lovely n900 :P
[15:05] <ayaka> nextime, sorry yaaic shock down, mine is htc evo 3d, i'm installing X for it in chroot, do you know fastboot?
[15:35] <nextime> uhmm
[15:36] Action: nextime has found some posts in the developement ml of ffmpeg about the ti dspbridge support
[15:58] <julienT_> hello, i'm just doing to make ./configure on the ffmpeg project but nothing
[15:59] <julienT_> how install the git project on my pc?
[15:59] <julienT_> not "./configure" "./make install" ?
[16:06] <drno_> julienT_: ./make?
[16:06] <drno_> looks like a path issue.
[16:07] <julienT_> oh yes i forgot ./make
[16:07] <julienT_> ty :p
[16:11] <julienT_> return error with make
[16:11] <julienT_> cmdutils.c line 48
[16:11] <julienT_> print_program_info
[16:11] <drno_> what version of ffmpeg are you trying to compile?
[16:12] <julienT_> last
[16:12] <julienT_> git clone
[16:12] <julienT_> from git://source.ffmpeg.org/ffmpeg.git
[16:13] <andrees> hi
[16:13] <julienT_> hi
[16:14] <andrees> j^
[16:14] <andrees> julienT_
[16:14] <julienT_> drno_: does it correct git?
[16:14] <andrees> do you know how to encode avi file to flv with vbr using 2 pass?
[16:15] <andrees> do you know how to encode avi file to flv with vbr using 2 pass? in windows ffmpeg :P
[16:15] <julienT_> no i'm trying to compile ffmpeg but i don't
[16:15] <julienT_> you do?
[16:15] <andrees> ok
[16:15] <andrees> and in linux bulid?
[16:15] <julienT_> no windows
[16:16] <andrees> ok
[16:16] <andrees> i want to learn how to encode .avi or any other file to .flv or .mp4 (avc,aac) with vbr using two pass
[16:17] <andrees> but i still dont understand how two pass works :S
[16:20] <andrees> if someone wants to help me, i'd be super gretful :)
[16:22] <julienT_> how build FFmpeg on windows plz
[16:22] <andrees> so julienT_, you need help with something?
[16:23] <julienT_> yes
[16:23] <andrees> can you tell me? i may help you
[16:23] <julienT_> how build on windows
[16:23] <andrees> you men, you have the source code, and you want to build it to use in windows?
[16:24] <julienT_> yes
[16:24] <andrees> mmm, i have no idea
[16:24] <andrees> lol
[16:24] <julienT_> and there is some errors
[16:24] <andrees> why dont you just donwload an already buit windows version?
[16:27] <andrees> barhom_ hi
[16:46] <julienT_> in which file we can found the main plz?
[16:50] <andrees> hi scoobertron codewrangler
[16:51] <andrees> i want to learn how to encode .avi or any other file to .flv or .mp4 (avc,aac) with vbr using two pass
[16:54] <codewrangler> Andrees: I'm just a beginner using ffmpeg&  Lately I've been using iVI on my Mac for video conversion lately.
[16:54] <andrees> ok
[16:54] <andrees> lol
[16:55] <andrees> so codewrangler you need help with anything? i may help
[16:55] <codewrangler> andrees: Just my 2011 taxes is all. ;-)
[16:55] <codewrangler> I'm pretty set at the moment for video, but thanks so much for asking.
[17:15] <andrees> hi
[17:15] <andrees> is there anyone that can help me?
[17:15] <andrees> i want to learn how to encode .avi or any other file to .flv or .mp4 (avc,aac) with vbr using two pass
[17:28] Last message repeated 1 time(s).
[18:10] <Alasdairr> Hi all
[18:11] <Alasdairr> Trying to build ffmpeg 0.10, getting a gcc error about redeclaration of enumerator 'VAR_SAR' - this is with gcc 4.4.5
[18:11] <Alasdairr> There are indeed two occurances of VAR_SAR in the var_name enum
[18:12] <lkiesow> Hi everyone, I have a problem with ffmpeg 0.10
[18:12] <lkiesow> After calling "ffmpeg -i INPUT -acodec aac -vcodec libx264  OUTPUT.mp4" I get the error messages "[NULL @ 0x1dc8da0] Codec is experimental but experimental codecs are not enabled, see -strict -2" and "Error while opening encoder for output stream #0:1 - maybe incorrect parameters such as bit_rate, rate, width or height"
[18:12] <lkiesow> Adding "-strict experimental" or "-strict -2" changes nothing
[18:13] <lkiesow> On another system with ffmpeg version 0.8.9 it worked just fine
[18:13] <lkiesow> Has anyone an idea what might be wrong?
[18:17] <relaxed> lkiesow: what does this return?  ffmpeg -codecs 2>&1| grep  ^..E.*aac*
[18:20] <lkiesow> "ffmpeg -codecs 2>&1| grep  ^..E.*aac*" returns nothing, but "ffmpeg -codecs 2>&1| grep  aac" will give you "DEA D  aac             Advanced Audio Coding" and " D A D  aac_latm        AAC LATM (Advanced Audio Codec LATM syntax)"
[18:21] <lkiesow> relaxed: Thus the codec should be there if I'm not mistaken
[18:22] <relaxed> lkiesow: yes, I was hoping you had a better aac encoder to use
[18:22] <relaxed> like libvo-aacenc
[18:23] <lkiesow> I build it myself ’ I can build it with other encoders. Bt as I said: It worked with the old ffmpeg&
[18:24] <relaxed> I know but you would be better off using libvo-aacenc anyway
[18:25] <lkiesow> Nevertheless: Which aac encoder would you recommand? faac, libvo-aacenc (are there more)
[18:25] <lkiesow> Why? :-)
[18:27] <teratorn> are there any established image file formats that store uncompressed or compress YUV data?
[18:27] <relaxed> ffmpeg's aac encoder still needs work, libfaac is non-free and unmaintained, so that leaves libvo-aacenc
[18:29] <lkiesow> relaxed: ok, I try it. It vannot get worse anyway ^^
[18:30] <relaxed> lkiesow: I use it and have been happy with it.
[18:31] <relaxed> teratorn: using the output "%03d.Y" will output 001.Y, 001.U and 001.V if that's what you're after.
[18:32] <teratorn> relaxed: ah no, I need to produce said files in code
[18:32] <teratorn> such that some kind of image viewer could view them
[18:33] <relaxed> are you talking lossless?
[18:40] <andrees> i want to learn how to encode .avi or any other file to .flv or .mp4 (avc,aac) with vbr using two pass
[18:40] <andrees> is there anyone that can help me?
[19:08] Action: nextime cry and want to have tidspbridge support in ffmpeg
[19:08] <jamiel> Hi all, is it possible to apply an avfilter at specific time intervals within the stream rather than have it applied all the time?
[19:08] <jamiel> Can't find any documentation on this
[19:12] <relaxed> jamiel: look at the movie filter in the man page
[19:16] <relaxed> I'm not sure if it gives you the control you're looking for.
[19:20] <jamiel> Hmm, it looks like I can only specify a single time interval ... I'm using the overlay filter and I would like it to appear every X number of seconds
[19:20] <jamiel> Anyone know a better way to achieve this if there is no way to make the filter repeat itself?
[19:46] <jamiel> I see there is a select filter ... can I combine the select filter and color filter ... so apply the color filter to the bits of the stream that match the select?
[20:27] <Mattias> I have a headset, and I want to stream, with ffserver/ffmpeg, what I hear. It's an usb soundcard and it only has "Speaker" and "Capture" which can be seen in alsamixer. How do I use "Speaker" as audio input? is that possible?
[20:37] <pasteeater> burek probably can figure it out if he's around. i'm alsaignorant.
[21:04] <o_0oo> hi
[21:04] <jamiel> hi o_0oo , how can we help?
[21:04] <vadim__> hi 2 all
[21:06] <o_0oo> well, i was wondering - i have setup a filtergraph which splits and in one of the split chains i do a select - when i overlay it back with the original, the whole in->out chain seems to be trimmed to what's in the select. I would have expected the in->out chain to be unaffected, and only the selected part of the second chain would be overlaid
[21:08] <o_0oo> weird, huh?
[21:08] <jamiel> hmmm
[21:09] <jamiel> what's your -av line ?
[21:10] <jamiel> I mean -vf
[21:11] <o_0oo> essentially, it is [in] split [nobox], fifo, [withbox] overlay [out]; [nobox] fifo, drawbox, select [withbox]
[21:11] <o_0oo> the select gets just seconds 10-20
[21:11] <o_0oo> so i would have thought that the drawbox would be overlay'd for just seconds 10-20
[21:12] <o_0oo> but, it doesnt do tht
[21:16] <o_0oo> what i want to achieve is to get a colour'd overlay on top of an rtmp at specified intervals (e.g., every 20 seconds).
[21:22] <o_0oo> if anyone could offer any advice on how to do that, i'll give them a hug
[21:25] <andrees> i want to learn how to encode .avi or any other file to .flv or .mp4 (avc,aac) with vbr using two pass
[21:25] <andrees> is there anyone that can help me?
[21:28] <JEEB> are you trying to get the exact same file size for same length, or do you just want to do good quality video?
[21:48] <andrees> jeeb are you asking me?
[21:48] <JEEB> whom else?
[21:48] <andrees> lol
[21:48] <andrees> yes
[21:49] <JEEB> so you would rather have exactly the same file sizes compared to stable quality?
[21:49] <andrees> i want to convert an avi file to flv, that keep all the quality but the less size posible
[21:49] <JEEB> ...which means you want some certain quality .-.
[21:49] <JEEB> and not exactly the same file size
[21:49] <andrees> yes
[21:50] <andrees> the same quality than the avi, but the less size posible
[21:50] <andrees> i thought that with vbr the file size could be less
[21:50] <JEEB> it's pretty hard to get cbr with libx264 :P
[21:50] <andrees> ok, i want to vbr
[21:50] <JEEB> (possible but definitely not what you'd want in 99% of all cases)
[21:51] <JEEB> anyways, the same algorithm is used for two vbr rate control modes in libx264
[21:51] <andrees> is it posible to encode to vbr flv?
[21:51] <andrees> ok im listening
[21:51] <JEEB> crf is basically "quality level"
[21:51] <JEEB> and then there's a bitrate mode for "I want this exact size/average bit rate"
[21:52] <andrees> ok
[21:52] <JEEB> for bitrate mode you need two passes for best quality, but for crf you only need one because you aren't really limiting the encoder in a similar way
[21:52] <andrees> ok i understand
[21:53] <andrees> but
[21:53] <JEEB> ffmpeg -i input.derp -vcodec libx264 -preset:v slowest_libx264_preset_you_can_take -crf highest_crf_that_still_looks_good_to_you -acodec libfaac (out.flv or out.mp4)
[21:53] <JEEB> ^ this is what you want
[21:54] <andrees> with this, will a have a variable bitrate?
[21:54] <JEEB> yes
[21:54] <andrees> o a constant bit rate?
[21:54] <JEEB> ...
[21:54] <JEEB> <JEEB> it's pretty hard to get cbr with libx264 :P
[21:54] <andrees> ok
[21:54] <andrees> good
[21:54] <JEEB> you'd have to actually work harder to get CBR out of x264 :P
[21:54] <andrees> its ok
[21:54] <andrees> i dont want cbr
[21:54] <JEEB> because it's certainly NOT RECOMMENDED and NOT WANTED for most cases
[21:54] <andrees> i want the variable bit rate
[21:55] <pasteeater> just do what JEEB said and you'll get what you want.
[21:55] <andrees> yes i will
[21:55] <andrees> but first
[21:55] <JEEB> http://mewiki.project357.com/wiki/X264_Settings#preset
[21:55] <JEEB> ^ list of presets
[21:55] <pasteeater> also add '-q:a 100' for audio VBR when using libfaac
[21:56] <andrees> why do i need a slow preset?
[21:56] <JEEB> you don't need a slow preset
[21:56] <andrees> i mean i have this ones
[21:56] <JEEB> medium is default
[21:56] <JEEB> and then if you want it to compress less and be faster you go faster
[21:56] <JEEB> and if you want it to compress more and be slower you go slower
[21:56] <JEEB> which is why you use the slowest preset YOU CAN TAKE
[21:56] <andrees> ok
[21:56] <andrees> i understand
[21:56] <andrees> so can i use this preset?: libx264-slow_firstpass.ffpreset
[21:57] <JEEB> ...
[21:57] <JEEB> use the libx264's internal presets
[21:57] <JEEB> I just listed them
[21:57] <JEEB> also start with crf around 24 or so
[21:57] <JEEB> if it looks good, raise it
[21:57] <JEEB> if it looks bad, lower it
[21:57] <JEEB> also
[21:58] <JEEB> you most probably don't want to use the fastest and slowest presets in libx264
[21:58] <JEEB> because those are mostly for testing
[21:58] <JEEB> placebo is lolslow
[21:58] <JEEB> and ultrafast sacrifices everything to be fast
[22:02] <andrees> ok
[22:02] <andrees> it is ocnverting now
[22:02] <andrees> it is converting now
[22:02] <andrees> another question, when using this :ffmpeg -i input.derp -vcodec libx264 -preset:v slowest_libx264_preset_you_can_take -crf highest_crf_that_still_looks_good_to_you -acodec libfaac (out.flv or out.mp4)
[22:02] <andrees> is one pass or two pass?
[22:03] <JEEB> one pass
[22:03] <JEEB> crf only needs one
[22:03] <JEEB> ...which, btw, I already said
[22:10] <pasteeater> andrees: you forgot '-q:a 100'.
[22:11] <pasteeater> ...but default at least isn't bad.
[22:11] <pasteeater> so either way...
[22:12] <andrees> it keep giving me this error
[22:12] <andrees> Error while opening encoder for output stream #0:1 - maybe incorrect parameters such as bit_rate, rate, width or height
[22:15] <andrees> pasteeater no i dont forgot it, i include it in the cmd, but it give me arror
[22:15] <andrees> i believe its because im using -acodec aac insted of libaac
[22:15] <andrees> but i dont have libaac library
[22:16] <andrees> libfaac
[22:18] <andrees> JEEB pasteeater are you there?
[22:19] <JEEB> -acodec aac is the ffaac encoder (ff* are internal ones), which is still in experimental status
[22:19] <JEEB> are you sure it doesn't tell you to add the experimental enabling setting before that final error?
[22:20] <andrees> it doesnt askme to write -strict experimental
[22:20] <andrees> but i just try and it works thanks
[22:21] <andrees> i will try con encode now and tell how it gon
[22:31] <andrees> jeeb
[22:31] <Mattias> JEEB: !
[22:31] <Mattias> You are everywhere :)
[22:31] <andrees> at first look it seems it work good
[22:32] <andrees> so you said, lower cfr will result in more quality and higer cfr in worse quality, is that correct?
[22:32] <andrees> mattias do you speak spanish?
[22:33] <Mattias> andrees: I'm Swedish
[22:33] <Mattias> I only speak Japanese
[22:33] <andrees> ok
[22:33] <andrees> i asked because of your name
[22:33] <andrees> it is a very common name here in argentina
[22:34] <andrees> but with one t only
[22:34] <Mattias> Very common in Sweden with 2x t's or th combo
[22:35] <Mattias> ok, so, I'm trying to stream with ffserver + ffmpeg, when I run ffmpeg I get this: http://pastebin.com/aqdFAgTy   broken ffmpeg something -.- Did I forget a conf file somewhere?
[22:35] <JEEB> <andrees> so you said, lower cfr will result in more quality and higer cfr in worse quality, is that correct? <- yes, thus the thing we're aiming is the highest crf value that looks good for you
[22:36] <Mattias> I have x264 library installed and such -.-
[22:36] <JEEB> which is why I said, if it looks good -> go higher
[22:36] <andrees> yes, thanks a lot :)
[22:37] <Mattias> JEEB: any experience with ffserver + ffmpeg? well, I guess this error is more related to ffmpeg alone
[22:37] <JEEB> ffserver is someething deeply generally unknown and unused
[22:38] <JEEB> arcane black magic
[22:38] <Mattias> Got any other ideas on how to stream on linux?
[22:38] <Mattias> to pages like twitch.tv etc
[22:38] <Mattias> I want to stream me playing dwarf fortress :P
[22:38] <JEEB> I'd say the most luser-usable thing for that
[22:39] <JEEB> is VLC
[22:39] <burek> Mattias, if your audio card doesn't support recording of "what you hear" then try with jack audio
[22:39] <o_0oo> does anyone know anything about filter graphs?
[22:39] <burek> o_0oo, yes, what do you need
[22:40] <Mattias> burek: yeah, shitty usbsoundcard, got nothing but a "Speaker" and "Capture" for mic..  I'll take a look at jack audio after I get at least video running
[22:40] <burek> use -f x11grab if you are on linux
[22:41] <Mattias> burek: http://pastebin.com/aqdFAgTy  <-- I get this error
[22:41] <Mattias> [libx264 @ 0x205f400] broken ffmpeg default settings detected  -.-
[22:41] <o_0oo> burek, i need to overlay a colour source onto a stream, but with a select applied to the chain with the colour, so that it's only displayed for the period that the select is valid for
[22:42] <burek> broken ffmpeg default settings detected
[22:42] <burek> that means your ffmpeg is old
[22:42] <Mattias> eeh
[22:42] <burek> o_0oo, I see.. let me think
[22:42] <o_0oo> thanks dude
[22:43] <Mattias> burek: 20120127-3 <-- this is the version I'm using, how old can that be?
[22:43] <Mattias> I'll try dev version
[22:43] <burek> Mattias, it doesn't matter when has it been compiled, it matters from when the source code is
[22:44] <Mattias> getting git versions
[22:45] <burek> o_0oo, first read this please: http://ffmpeg.org/ffmpeg.html#Filtergraph-syntax
[22:45] <Mattias> both for x264 and ffmpeg
[22:45] <burek> so I can explain it to you more easy
[22:46] <o_0oo> ok
[22:46] <burek> o_0oo, also this: http://ffmpeg.org/libavfilter.html#Tutorial
[22:46] <burek> it's not too much, but it'll explain the basics
[22:46] <o_0oo> i've read that one :-)
[22:46] <burek> Mattias, that's the smartest idea :)
[22:47] <burek> o_0oo, oh, so you are familiar with chains
[22:47] <o_0oo> to an extent, yes
[22:49] <andrees> jeeb another question, is it posible to encode using variable frame rate? and if it is, does it reduce file size?
[22:50] <o_0oo> ive tried a number of approaches, firstly by creating a colour source, applying the select and then overlaying with the original - -vf "color=#8EABB8 at 0.9:480x208,select='gte(t,10)*lte(t,20)' [over]; [in][over] overlay [out]"
[22:50] <burek> o_0oo, so, use the split, then in one branch, use just the fifo and overlay, just like in that libavfilter example image
[22:50] <burek> in another branch use select
[22:51] <o_0oo> i tried that too - but it seemed like the select was applied to the primary chain
[22:51] <burek> if I get you correctly, you need the normal video and overlayed video only when select allows it
[22:51] <o_0oo> or, to both chains
[22:51] <o_0oo> burek - exactly
[22:51] <burek> ill paint you a diagram :)
[22:51] <o_0oo> :)
[22:54] <andrees> pasteeater, when using -q:a 100, how do i change quality, what happen if i use less than 100 or grater than 100?
[22:54] <burek> o_0oo, http://pastebin.com/3DRedLJf
[22:55] <o_0oo> awesome, hold on
[22:56] <Mattias> burek: latest git and x264 git, exact same error
[22:57] <burek> o_0oo, this is a more complete version: http://pastebin.com/FE6tyFuP
[22:59] <o_0oo> ok, cool, this is what I have - it looks the same, but the stream seems to freeze after 10 seconds - http://pastebin.com/bznGMUCF
[23:00] <Mattias> burek: ah, just found on google... apparently x264 thinks ffmpegs encoding settings are broken and refuses to run just like that O.o x264 and ffmpeg aren't friends?
[23:07] <pasteeater> andrees: see 'man faac'
[23:08] <o_0oo> well, to be clear, it doesn't freeze, the output is just the 10 seconds of the select
[23:09] <pasteeater> Mattias: ffmpeg and x264 work fine together. you just have to get thr right syntax for your version(s).
[23:11] <Mattias> I noticed x264 wouldn't run on git version, reverted back to stable arch linux package
[23:12] <Mattias> hm, I wonder what I'm missing -.-
[23:12] <pasteeater> ioni does a good job at keeping ffmpeg updated. probably better than any other sane distro.
[23:13] <burek> Mattias, -vpre is deprecated
[23:14] <burek> ffmpeg now uses native x264 profiles/presets/tunes
[23:14] <Mattias> can't find any presets for x264 either :/
[23:14] <burek> using -preset -profile -tune
[23:14] <burek> -vpre used some sort of files, containing a lot of x264 specific options
[23:14] <burek> that's now deprecated, that's why you get that error
[23:14] <Mattias> but I haven't used -vpre, it just tells me to use it
[23:15] <relaxed> jamiel: output all the frames and process them based on their sequence number(s)
[23:15] <pasteeater> i'd assume 37208-g01fcbdf is new enough to use -preset instead of -vre
[23:15] <pasteeater> not sure why is suggests vpre
[23:15] <Mattias> ah, tried profile, not preset... -preset ultrafast works
[23:16] <Mattias> getting a lot of VBV underflows though
[23:23] <o_0oo> burek, any further ideas?
[23:23] <burek> just a sec please
[23:23] <burek> im on the phone
[23:29] <rainmaker1> Does anyone know why what's the purpose of pkt_size? Yes, I know set the size of udp packets in bytes, but is only because of network MTU or there is something else?
[23:29] <rainmaker1> me and my fat fingers, sorry for typos :)
[23:29] <pasteeater> you need a dialing wand.
[23:30] <rainmaker1> :D:D:D
[23:32] <pasteeater> rainmaker1: did you see 'man ffmpeg'? Under PROTOCOLS there is a udp section.
[23:34] <o_0oo> i think filtergraphs limit the output to the length of the shortest chain
[23:35] <o_0oo> so if one chain has a select, that will determine the length of the whole output
[23:35] <Mattias> woho, I streamed :D
[23:35] <Mattias> got it to work :)
[23:35] <nextime> lucky you.
[23:37] <Mattias> just followed this guide and changed vpre to preset = http://michaelchinen.com/2011/06/27/justin-tv-streaming-in-linux/  ^.^
[23:37] <Mattias> works like a charm ^.^
[23:38] <rainmaker1> pasteeater: yes, but I don't understand the point? I know what pkt_size option do, but I dont know WHEN I should use it... I guess Ill need a little help from tcpdump
[23:40] <burek> im here
[23:40] <burek> sorry for keeping you waiting
[23:40] <burek> o_0oo, well, you need to tell select
[23:41] <burek> to pass empty frames or something, just to continue the output
[23:41] <burek> im not sure, but i think it stops because it waits for the 2nd branch to get the input from
[23:43] <nextime> Mattias : i insist, lucky you that you can solve your issue with some doc
[23:43] <nextime> my options is to write code or to use gst instead of ffmpeg ( and i hate gst and love ffmpeg )
[23:44] Action: nextime sad
[23:45] <burek> nextime, what's your issue
[23:46] <burek> rainmaker1, the packet size is usually prioritized on routers
[23:46] <nextime> burek : the lack of support for tidspbridge on ffmpe
[23:46] <nextime> ffmpeg
[23:46] <burek> the smaller the packet, the bigger priority will be
[23:46] <burek> it's just a tweaking option, the stream should work even without touching it
[23:46] <nextime> there are some patches in some ml posts of 2 years ago
[23:46] <burek> what is tidspbridge :)
[23:46] <nextime> and nothing more apparently
[23:47] <nextime> burek : the support for encoding and decoding using the texas instruments DSP that exists in a lot of cortex a8 and a9 based soc
[23:47] <rainmaker1> burek: I am not quite sure, the packet is prioritized based on CoS/QoS values not the size AFAIK
[23:47] <pasteeater> nextime: we are interested in orphaned patches. link?
[23:48] <nextime> pasteeater i found some about that here
[23:48] <nextime> http://ffmpeg.org/pipermail/ffmpeg-devel/2010-September/100622.html
[23:49] <nextime> pasteeater : at least in debian from debian-multimedia they appears to be not present
[23:49] <nextime> i'm looking at the latest sources if the support exists
[00:00] --- Fri Feb 24 2012


More information about the Ffmpeg-devel-irc mailing list