[Ffmpeg-devel-irc] ffmpeg.log.20150301

burek burek021 at gmail.com
Mon Mar 2 02:05:01 CET 2015


[00:00] <shevy> is there a somewhat longer explanation of the various filters in ffmpeg? like ideally with examples?
[00:07] <pzich> shevy: read through https://trac.ffmpeg.org/wiki/FilteringGuide and https://ffmpeg.org/ffmpeg-filters.html
[00:24] <isweet32_> Hey guys. Trying to stream some video to Skype for Linux. Anyone know what pixel format / colorspace I should be using? Skype sees my v4l2 loopback device but isn't reading/rendering the frames.
[00:26] <isweet32_> Currently using: ffmpeg -vcodec png -f image2pipe -i - -f v4l2 -pix_fmt rgb24 /dev/video0
[00:26] <isweet32_> Piping a PNG stream into that
[00:27] <isweet32_> Works like a charm if I open it up with something like mplayer
[00:27] <isweet32_> But skype doesn't like it.
[00:37] <iive> isweet32_: i've never tried it, so i don't know.
[00:37] <iive> but you should also try rgb32 and bgr24 and bgr32 , as pixel formats
[00:38] <iive> maybe yuy2 or uyvy , too
[00:38] <iive> that's what my analog capture card reports as supported
[00:39] <isweet32_> Thanks, I'll give em a try
[00:39] <isweet32_> I've been using rgb24 because that's the pixel format that does'nt warp the video coming from PNG
[00:40] <isweet32_> most of the YUV formats screw with the colors
[00:40] <iive> well, it is quite likely that the codec used by skype is using yuv.
[00:41] <iive> i don't know that for sure either.
[00:42] <isweet32_> Thanks for the help, i'll try some different formats and see what happens
[00:43] <iive> tell us what works, if you find out.
[00:44] Action: iive is out
[02:02] <ben1> can anyone help me?  i want to live transcode an http mpeg2 source to http h264 destination.  i am going to open a port on my firewall to allow access to the transcoded stream over the internet.  my upload bandwitdh from my ISP is 6mb/s.  the source video is 22mb/s.
[02:03] <BtbN> http h264 destination?
[02:03] <ben1> yes the outfile
[02:04] <BtbN> ffmpeg can't act as a http server, if that's what you mean.
[02:04] <ben1> really, i thought it could?
[02:10] <Sonny_Jim> You could pipe it into netcat or something equally bizarre
[02:10] <ben1> like wowza?
[02:10] <ben1> or vlc
[02:16] <BtbN> If you want a simple streaming server, nginx-rtmp is quite nice.
[02:17] <ben1> thanks..does it accept http source?
[02:18] <BtbN> no, it's an rtmp server.
[02:20] <ben1> right, but i am talking input source..or does it only accept a file?
[02:20] <BtbN> It accepts only rtmp.
[02:21] <BtbN> It can do dash and hls re-muxing for you though. And of course stream via rtmp.
[02:22] <ben1> this is what i am trying to do:  ffmpeg -i "udp://localhost:1234" -vcodec libx264 -vb 150000 -g 60 -vprofile baseline -level 2.1 -acodec aac -ab 64000 -ar 48000 -ac 2 -vbsf h264_mp4toannexb -strict experimental -f mpegts udp://127.0.0.1:10000?pkt_size=1316
[02:22] <ben1> only http instead of udp
[02:24] <BtbN> There is no such thing as "streaming to http"
[02:24] <BtbN> ffmpeg can open a http link just fine though
[02:25] <BtbN> But it can't output to http POST or something like that.
[02:25] <ben1> can it output to udp?
[02:26] <BtbN> At least i think so, yes
[02:30] <ben1> ok what is the best protocol to use for output with real-time mpeg-ts?
[02:31] <BtbN> That entirely depends on what you need.
[02:33] <ben1> with my limited upload bandwitdh, i just want to get the best quality video possible, leaving just a little overhead.
[02:34] <BtbN> Well, if you don't even know where you want to stream stuff to.
[02:35] <ben1> i will stream it to over the internet to myself when i am on the road.
[02:37] <BtbN> Just setup an nginx-rtmp server then, it's exactly what you want.
[02:37] <ben1> the http source is something i already have on my intranet.  i just want to transcode the stream and make it available over the internet
[02:38] <ben1> so does nginx-rtmp transcoded also?
[02:38] <BtbN> no, it's a pure rtmp server with additional support for re-muxing to hls/dash.
[02:39] <BtbN> You stream to it via rtmp, it streams to clients via rtmp.
[02:39] <BtbN> It can do on-demand stuff though, where it launches some command which feeds it an rtmp stream.
[02:39] <BtbN> Just read the documentation on how to do that.
[02:41] <ben1> i dont understand why adding another piece to the puzzle.  if i am changing the protocal, can do rtmp output with ffmeg, correct?
[02:41] <BtbN> yes, ffmpeg can stream to an rtmp server.
[02:41] <BtbN> ffmpeg cannot by itself act as a server.
[02:42] <ben1> ok makes sense..thanks
[02:45] <ben1> think i will try to use my raspberrypi for the rtmp server
[02:46] <BtbN> An RPi won't be able to encode h264, it's way too slow.
[02:48] <ben1> i cant encode on another computer and just use the raspberry pi for the server?
[02:48] <BtbN> Why? You just complained about adding another pice of software, now you want to throw in more hardware.
[02:48] <BtbN> Also, on-demand won't be possible that way.
[02:49] <BtbN> just configure nginx-rtmp to fire up ffmpeg on-demand, that then does the transcoding you want.
[02:49] <BtbN> Just read the documentation on that.
[02:52] <ben1> ok ill check it out, but raspberrypi does have hardware accelerated h264 encoding i think.
[02:52] <BtbN> There is no support for that in ffmpeg.
[04:32] <BtbN> Anyone ever used the dashenc? When using it with -c:a copy, it seems to put an invalid codec name in the mpd file. When re-encoding the aac, the audio works.
[04:32] <BtbN> It's putting mp4a.40 there in copy-mode, and mp4a.40.2 when re-encoding.
[06:01] <smo_> hi, is it possible to use the -ss option with -re ?
[06:01] <smo_> whn i do it my video start and play normally but if i try seeking the video won t restart
[06:01] <smo_> :/
[06:02] <smo_> works well without -re
[08:31] <YaMoonSun> What cmd did I do wrong? http://pastebin.com/4pQTUxRk
[09:06] <Shells> is there any way of using ffmpeg in a similar way to -vcodec copy -acodec copy however it will fix any errors in the stream (such as adding blank frames (audio or video) if there is corruption in a frame) but without otherwise changing the streams ...?
[09:17] <relaxed> Shells: no
[09:26] <YaMoonSun> relaxed, dont suppose you checked out muh link?
[09:59] <Shells> thanks relaxed
[09:59] <relaxed> YaMoonSun: -vf scale=-2:480
[10:00] <relaxed> YaMoonSun: "[libx264 @ 0000000002c6a640] width not divisible by 2 (853x480)"
[10:01] <YaMoonSun> Well darn - It worked with webm
[10:02] <YaMoonSun> Thanks, I appeciate it man
[10:03] <relaxed> you're welcome
[10:20] <YaMoonSun> This may seem like a dumb question, but can I make an mp4 "web optimized" manually with ffmpeg?
[10:54] <YaMoonSun> Why are my 640x480p files.. different? http://webmshare.com/play/eX4DP
[11:09] <zenny> Hi, facing a problem with sound capturing in static build of ffmpeg (both git and stable), arecord -l shows all the devices, but ffmpeg is executed, get an error "[alsa @ 0x2bccb80] cannot open audio device hw:2,0 (No such file or directory)
[11:09] <zenny> hw:2,0: Input/output error".
[11:09] <zenny> any input?
[12:05] <_genuser_> hello guys.
[12:06] <_genuser_> I see online some people use -vcodec libx264, and some example show -vcodec h264. I'm using h264. what's the difference between the two?
[12:21] <dave88> Hi all - getting an error on 'make' - any ideas?
[12:21] <dave88> libavcodec/libavcodec.a(libx264.o): In function `X264_init':
[12:21] <dave88> /home/media/src/ffmpeg/libavcodec/libx264.c:552: undefined reference to `x264_encoder_open_146'
[12:21] <dave88> collect2: error: ld returned 1 exit status
[12:21] <dave88> make: *** [ffmpeg_g] Error 1
[14:10] <relaxed> dave88: is that x264 git?
[14:12] <relaxed> if so, run ./configure --extra-cflags="-I/path/to/x264/prefix/include" --extra-ldflags="-L/path/to/x264/prefix/lib"
[14:26] <_genuser_> so I'm encoding a video with -bufsize 1400 and it's making the output 1G. with handbrake and bitrate of 1400 kbps it's about 200MB. and it's deinterlaced so smoother.
[14:27] <ChocolateArmpits> _genuser_: you can deinterlacer in command line too via the yadif filter
[14:27] <ChocolateArmpits> deinterlace*
[14:27] <JEEB> uhh, maxrate and bufsize (those two always go together) and rate control are only related. you limit the latter with the first
[14:27] <_genuser_> any recommendations on what else to do to a) maintain 1400 kbps bitrate, b) cfr, c) deinterlace (going to tyring using vf/yadif), d) and keep the file size to < 300MB (20min)
[14:27] <ChocolateArmpits> You need a 2 pass encode then
[14:28] <JEEB> so you either use 2pass and bit rate, or CRF if you don't want an exact size. and then limit with maxrate/bufsize if you have specific VBV limitations :P
[14:28] <_genuser_> ChocolateArmpits: yeah, I'm looking up the yadiff documentation right now to see how to put that in there. current command looks as follows
[14:28] <_genuser_> "%FFMPEG%" -i %1 -bufsize 1400 -vcodec h264 -acodec aac -strict -2 %FINALTARGET% -async 1 -aspect 720:480
[14:29] <ChocolateArmpits> if the file has interlace information set then simply "-vf yadif"
[14:29] <ChocolateArmpits> If however there is none you will have to specific field order
[14:29] <ChocolateArmpits> specify*
[14:29] <_genuser_> it's tv recording. so I'm assuming it will always have interlacing.
[14:29] <_genuser_> altho, I'm not the most educated about video formats.
[14:29] <ChocolateArmpits> Yeah it might have but the stream can either be defined as progressive or interlaced
[14:30] <ChocolateArmpits> stream metadata in the file
[14:30] <_genuser_> ok, so if I put in -vf yadiff blindly, it might hurt quality?
[14:30] <ChocolateArmpits> it might not deinterlace at all
[14:30] <_genuser_> which wouldn't be a problem it's not interlaced to begin with, correct?
[14:31] <ChocolateArmpits> If the flags are set incorrectly, i.e. video is interlaced but the stream flag is set as progressive, then it will hurt
[14:31] <_genuser_> I see because ffmpeg will see the stream flag as progressive and not do anything?
[14:32] <ChocolateArmpits> Well that's how yadif works, I think other available deinterlace filters will do the work regardless of the flag
[14:32] <ChocolateArmpits> That's why you have to specify the field order in the filter
[14:32] <ChocolateArmpits> to be absolutely sure
[14:33] <_genuser_> ok, let me see what I can find about field order in docs
[14:33] <ChocolateArmpits> so "-vf yadif=parity=tff" if you're in NTSC land
[14:33] <ChocolateArmpits> and bff if PAL
[14:34] <ChocolateArmpits> oh my mistake
[14:34] <ChocolateArmpits> "The default value is auto. If the interlacing is unknown or the decoder does not export this information, top field first will be assumed. "
[14:34] <_genuser_> should I put this in withotu quotes
[14:34] <ChocolateArmpits> ok so yadif will deinterlace regardless
[14:34] <ChocolateArmpits> yeah without
[14:35] <ChocolateArmpits> Also your finaltarget variable should be at the very end of your command line
[14:35] <ChocolateArmpits> unless you have multiple outputs
[14:35] <_genuser_> thanks. also I noticed that I had found online -s 720x480, however resulting video is 1920x1080
[14:35] <ChocolateArmpits> well that's trange
[14:37] <ChocolateArmpits> hmm aspect should define the aspect ratio rather than the resolution for video who's sample aspect isn't 1:1
[14:37] <ChocolateArmpits> Otherwise you will get distorted video on playback
[14:37] <_genuser_> "%FFMPEG%" -i %1 -bufsize 1400 -vf yadif=parity=tff -vcodec h264 -acodec aac -strict -2 -pass 2 -async 1 -s 720x480 %FINALTARGET%
[14:37] <ChocolateArmpits> cool
[14:37] <_genuser_> my understanding was the -2 would do a 2 pass. I think I'll remove that and leave -pass 2.
[14:37] <ChocolateArmpits> oh wait
[14:38] <ChocolateArmpits> you define video bitrate with -b:v, not -bufsize
[14:38] <ChocolateArmpits> bufsize defines the size of variable bitrate buffer and works along with -maxrate
[14:39] <ChocolateArmpits> To performa a 2 pass encoding you first have to do a 1st pass encoding with "-pass 1"
[14:39] <ChocolateArmpits> and then a second pass with "-pass 2"
[14:39] <ChocolateArmpits> optionally you can define where the pass log file should be stored or else it will be saved in your current directory
[14:39] <_genuser_> I see. I was hoping it'd be like handbrake, where I just "hey do 2pass". lol
[14:39] <ChocolateArmpits> Well you can do a for loop in windows cmd
[14:40] <ChocolateArmpits> for /L %i in (1,1,2) do (...)
[14:40] <ChocolateArmpits> this will output 1 and 2
[14:40] <ChocolateArmpits> but ok
[14:40] <_genuser_> it's just two lines, guess I don't mind.
[14:41] <_genuser_> so I'd need a 2pass encoding to get it to a specific filesize. is that true? I was thinking of just going with crf 23 or something.
[14:41] <ChocolateArmpits> crf produces a variable bitrate output and adheres to no specific bitrate limit
[14:42] <ChocolateArmpits> It's harder to control and if the source has more motion you can expect the filesize to be bigger
[14:42] <ChocolateArmpits> You can do a 1 pass encoding by just defining the bitrate but it won't conserve the bits as well as a 2 pass encoding where it will give more bitrate to more motion active areas
[14:43] <ChocolateArmpits> Basically a 1 pass encode will give 1,4 Mbit for every second
[14:44] <_genuser_> I see, then I'd stick with 2pass encode targetting 1400 kpbs bitrate. I found it's just about 200MB (roughly) for 20min episodes. doesn't lose quality much and easy to store.
[14:45] <_genuser_> I don't watch these on 90inch wall projected system comparing it to blu-ray. heh. just recording tv shows.
[14:46] <ChocolateArmpits> you should also define audio bitrate with -b:a
[14:46] <ChocolateArmpits> something like 192k
[14:46] <ChocolateArmpits> maybe slightly more
[14:47] <ChocolateArmpits> defining pixel format is necessary also as filter output will be in 4:2:2
[14:48] <_genuser_> heh, I just did. :)
[14:48] <ChocolateArmpits> so "-pix_fmt yuv420p"
[14:48] <_genuser_> do I need pixel format?
[14:48] <ChocolateArmpits> well 422 will take more space
[14:48] <ChocolateArmpits> 4:2:0 is a happy optimum for most use cases
[14:48] <ChocolateArmpits> And very likely your source isn't more than that to begin with
[14:49] <ChocolateArmpits> (while that can be increased without downscaling)
[14:49] <ChocolateArmpits> with*
[14:49] <ChocolateArmpits> anyways just set the pix fmt
[14:50] <_genuser_> alright, setting that too.
[14:50] <_genuser_> so -pass 1 output can be NUL, correct? that's what the documentation is saying.
[14:50] <ChocolateArmpits> yes
[14:50] <ChocolateArmpits> the point is to get a pass log file
[14:50] <ChocolateArmpits> that will then get used by the second pass
[14:50] <_genuser_> I also think I forgot the k in -b:v 1400k . I had set to 1400
[14:51] <ChocolateArmpits> yeah else it will be only 1400 bits
[14:51] <_genuser_> nice, glad I caught it.
[14:52] <_genuser_> now only if I could just auto convert them instead of manual intervention. sigh
[14:52] <ChocolateArmpits> that's what batch scripting is for
[14:56] <_genuser_> sure, I got the scripts. I got them schedules in windows to run at 3am. it's the commercial stripping that require manual editing.
[14:56] <_genuser_> so one script runs at 3am converts from wtv to dvr-ms. later when I get a chance, I edit out commercials and dump them into another folder. another script picks them up and runs overnight to convert them to mp4s.
[14:59] <_genuser_> hmmm unable to find suitable output format for 720x480
[15:06] <_genuser_> "%FFMPEG%" -i %1 -b:v 1400k -vf yadif=parity=tff -vcodec h264 -pass 1 -acodec aac -b:a 160k -pix_fmt yuv420p -strict -async 1 -s 720x480 NUL
[15:06] <_genuser_> says unable to find suitable format for '1'.
[15:16] <ChocolateArmpits> you need to define the wrapper
[15:16] <ChocolateArmpits> "-f mp4"
[15:16] <ChocolateArmpits> _genuser_:  when outputting to NUL
[15:31] <_genuser_> ChocolateArmpits: ah, I see. thatnks :)
[15:32] <c_14> You can also use -f null
[15:52] <_genuser_> took some tries but I think I have a working pass 1 and pass 2 script. running thru first pass now. let's see how it does.
[16:04] <alimj> Hello. Good morning.
[16:04] <alimj> Could anyone please inform me, what is the current preferred method to use FFMPEG or LibAV in Python?
[16:05] <alimj> I would like to generate a video from some sequenced PNGs from Python
[16:23] <Fyr> guys, how do I use NVIDIA encoder with ffmpeg?
[16:23] <Fyr> I've installed the latest driver.
[16:43] <BtbN> Extract the nvEncodeApi.h from the NVENC sdk, put it somewhere configure will find it, and build with --enable-nvenc
[16:55] <Fyr> is there a prebuilt binary?
[16:56] <Fyr> libfdk_aac enabled would be cool as well. =)
[17:02] <c_14> I know libfdk_aac with ffmpg is not redistributable, I think nvenc also has licensing issues
[17:03] <Fyr> =(
[17:05] <Fyr> FOUND!
[17:05] <Fyr> glad that there are the guys who take crap onto the license issues.
[17:20] <Fyr> why the developers haven't ported libfdk-aac functionality to libvoaac_enc?
[17:21] <c_14> Because if they looked at the code they'd be "tainted" by the license.
[17:22] <c_14> Also, it's a lot of work.
[17:22] <Fyr> =(
[17:22] <Fyr> I hoped it can be rewritten without copying the code.
[17:24] <c_14> It can, but it's a lot of work.
[17:48] <dusted> I read somewhere that when using the -ss / -t with -c copy, one can risk video/audio desync, however, copy seems to me to be the only way to guarantee that there is no change in quality, I've done a few tests which have been inconclusive, do anyone any insights on this?
[17:52] <dusted> I mean, I risk losing quality, even if I reencode with the same codec/parameters as the initial source, or did I misunderstand that?
[17:52] <c_14> If you use a lossy codec you always lose quality
[17:54] <dusted> So, any ideas on how to create clips without losing quality, without desync and without destination filesize becoming larger than the source ?
[17:59] <c_14> I'd try copy. If that doesn't work you might have to live with recoding at least the audio.
[18:00] <dusted> c_14, ah, I didn't think about re-encoding only the audio, that would be acceptable
[18:37] <smo_> hi, is it possible to use the -ss option with -re ?
[18:37] <smo_> when i do it my video start and play normally but if i try seeking the video won t restart or i have to wait the time i seek to...
[18:38] <smo_> works well without -re
[18:40] <c_14> put -ss before -i
[18:41] <smo_> this is what i do c_14
[18:41] <smo_> args = ['-ss' , start,'-re','-i' ...........
[18:44] <c_14> Works fine for me
[18:44] <c_14> What version are you running?
[18:47] <smo_>  ill verify
[18:48] <smo_> ffmpeg version 2.5.2-   http://johnvansickle.com/ffmpeg/
[18:49] <c_14> I'm using recent git head, but that version should be new enough.
[18:49] <smo_> i ll try to git
[18:49] <smo_> the
[18:56] <smo_> same thing c_14
[18:56] <smo_> i seek then i have
[18:56] <smo_> grep stderr: frame=    1 fps=0.1 q=-1.0 size=       5kB time=00:12:57.60 bitrate=   0.1kbits/s    many times
[18:57] <smo_> if i wait 12 min 57 i ll see my video lol
[18:59] <c_14> `ffmpeg -ss 20:00 -re -t 60 -i foo.mkv out.mkv' starts encoding immediately and gives me 60 secods of output starting at minute 20
[19:02] <smo_> ffmpeg -ss 00:48 -re -i '/media/partage/audio/sons finis/truepiano_smo.mp3' -filter_complex [0:a]showwaves=mode=cline:rate=25,format=yuv420p[vid] -map [vid] -map 0:a -codec:v libx264 -crf 18 -preset ultrafast -codec:a libvorbis -threads 0 -copyts -sn -f matroska /tmp/test.mkv
[19:02] <smo_> wrong ?
[19:03] <smo_> humm don t link the copyts
[19:05] <c_14> works for me
[19:05] <c_14> (without the copyts)
[19:05] <smo_> Oo
[19:05] <smo_> ah yes
[19:05] <c_14> With the copyts it hangs
[19:05] <smo_> works in command line but not in my app it restart from begining
[19:05] <smo_> (i use it thru nodejs pipes)
[19:07] <smo_> humm no ...
[19:07] <smo_> ffmpeg -ss 00:48 -re -i '/media/partage/audio/sons finis/truepiano_smo.mp3' -filter_complex [0:a]showwaves=mode=cline:rate=25,format=yuv420p[vid] -map [vid] -map 0:a -codec:v libx264 -crf 18 -preset ultrafast -codec:a libvorbis -threads 0 -sn -f matroska /tmp/truc.mkv
[19:07] <dave88> Hi, I dont suppose anyone can help with this
[19:07] <dave88> media at media-desktop:~/serviio-1.5.1/bin$ ./serviio.sh
[19:07] <dave88> log4j:ERROR setFile(null,true) call failed.
[19:07] <dave88> java.io.FileNotFoundException: /home/media/serviio-1.5.1/log/serviio.log (Permission denied)
[19:07] <dave88> 	at java.io.FileOutputStream.open(Native Method)
[19:07] <dave88> 	at java.io.FileOutputStream.<init>(FileOutputStream.java:213)
[19:07] <dave88> 	at java.io.FileOutputStream.<init>(FileOutputStream.java:133)
[19:07] <smo_> the stram start at 00:00
[19:07] <dave88> 	at org.apache.log4j.FileAppender.setFile(FileAppender.java:294)
[19:07] <dave88> 	at org.apache.log4j.RollingFileAppender.setFile(RollingFileAppender.java:207)
[19:07] <dave88> 	at org.apache.log4j.FileAppender.activateOptions(FileAppender.java:165)
[19:07] <dave88> 	at org.apache.log4j.config.PropertySetter.activate(PropertySetter.java:307)
[19:08] <dave88> 	at org.apache.log4j.xml.DOMConfigurator.parseAppender(DOMConfigurator.java:295)
[19:08] <dave88> 	at org.apache.log4j.xml.DOMConfigurator.findAppenderByName(DOMConfigurator.java:176)
[19:08] <dave88> 	at org.apache.log4j.xml.DOMConfigurator.findAppenderByReference(DOMConfigurator.java:191)
[19:08] <dave88> 	at org.apache.log4j.xml.DOMConfigurator.parseChildrenOfLoggerElement(DOMConfigurator.java:523)
[19:08] <dave88> 	at org.apache.log4j.xml.DOMConfigurator.parseRoot(DOMConfigurator.java:492)
[19:08] <dave88> 	at org.apache.log4j.xml.DOMConfigurator.parse(DOMConfigurator.java:1001)
[19:08] <dave88> 	at org.apache.log4j.xml.DOMConfigurator.doConfigure(DOMConfigurator.java:867)
[19:08] <smo_> use pastebin dave88
[19:08] <dave88> 	at org.apache.log4j.xml.DOMConfigurator.doConfigure(DOMConfigurator.java:773)
[19:08] <dave88> 	at org.apache.log4j.helpers.OptionConverter.selectAndConfigure(OptionConverter.java:483)
[19:08] <dave88> 	at org.apache.log4j.LogManager.<clinit>(LogManager.java:127)
[19:08] <dave88> 	at org.slf4j.impl.Log4jLoggerFactory.getLogger(Log4jLoggerFactory.java:66)
[19:08] <dave88> 	at org.slf4j.LoggerFactory.getLogger(LoggerFactory.java:277)
[19:08] <dave88> 	at org.slf4j.LoggerFactory.getLogger(LoggerFactory.java:288)
[19:08] <dave88> 	at org.serviio.MediaServer.<clinit>(MediaServer.java:77)
[19:08] <dave88> media at media-desktop:~/serviio-1.5.1/bin$
[19:08] <dave88> ok mate
[19:08] <smo_> /home/media/serviio-1.5.1/log/serviio.log (Permission denied)
[19:08] <smo_> verify your permissions
[19:09] <dave88> It says owner is root - I can't change it in GUI, I'm a bit of a newbie on linux
[19:09] <dave88> Is there a terminal option
[19:11] <smo_> need to use chown
[19:11] <smo_> with the right user, and not sure your issue is ffmpeg retaled
[19:11] <dave88> Ah - I installed it as root, that's why. SImple reinstall - all is well, cheers smo_!
[19:12] <smo_> np :p
[19:12] <smo_> no idears why it start at 00:00 for me c_14 ?
[19:13] <c_14> no clue
[19:17] <smo_> strange :/
[19:17] <dave88> smo_ - just a linux question. I have 2 HDD's /dev/sda and /dev/sdb (SDB is a 2TB that I have mounted and I plan to put my media in). Any idea how I link to it in Serviio - I can't navigate to /dev/sdb
[19:17] <smo_> don t know serviio look at the conf files of it
[19:18] <dave88> CAncel that - sorted it
[19:18] <smo_> :)
[19:21] <dave88> Ha - found the HDD but owner is root. So, this chmod, the help file is in a different language! Any idea on the command to use to change the permissions to local user (media) ?
[19:25] <smo_> not normal look mount options of your drive
[19:25] <smo_> mount command and /etc/fstab
[19:35] <smo_> what s your version c_14
[19:35] <smo_> crazy i don t have same resultst than you Oo
[19:56] <c_14> smo_: N-70145-g721c916
[19:57] <smo_>  N-44838-g2b40416-
[19:57] <smo_> linux ? win ?
[19:57] <smo_> (thanks)
[19:58] <smo_> i don t understand  :/
[19:59] <c_14> linux
[20:00] <smo_> me too
[20:06] <smo_> so it works in my app but i need the copyts :/ c_14
[20:06] <smo_> if i don t have it my player progress bar do not work
[20:07] <c_14> What about -copyts -start_at_zero ?
[20:08] <smo_> i ll try
[20:08] <smo_> '-copyts','-start_at_zero','-ss' , start,'-re','-i', ''+link+'',
[20:08] <smo_> this order is ok ? c_14
[20:09] <c_14> should be
[20:09] <smo_> ok i try
[20:10] <smo_> same stuck ... :/
[20:10] <smo_> c_14,
[20:12] <c_14> But does it work with your app if you wait?
[20:12] <c_14> I'm not entirely sure why you need the copyts
[20:13] <smo_> yup it works in my player without copyts, but it s a html5 player if i don t have copyts the porgress bar do not work
[20:13] <smo_> can t detect the currentTime
[20:14] <c_14> hmm?
[20:14] <c_14> It should still get a time, it should just be offset by the seek time
[20:15] <smo_> yes but it will show 00:10:00 for exemple but in reality im at 01:10:00 or don t know what
[20:17] <c_14> Can't you just take the current time and add the seek time to it?
[20:18] <smo_> now i have -re and copyts, im at 01:00 i seek to 01:30, my player seek and i have to wait 01:30 before seeing the image and everything is ok
[20:18] <smo_> i retry without copyts, i think i can 'force' it yes, maybe
[20:20] <smo_> ok, without copyts i m a 01:00 i seek to 01:30 it works fine but my player restart to 00:00 in the progressbar
[20:20] <smo_> will look if i can add the seektime to currentTime ...
[20:41] <dave88> Hi folks - struggling with serviio console - hope you can help. Whenever I add a folder I get error 500 and it won't save
[20:42] Last message repeated 1 time(s).
[20:43] <dave88> Theres nothing in the log
[20:46] <smo_> dave88, why do you ask it in #ffmpeg  :) ??
[20:47] <dave88> I found someone saying it was a place of expertise on a form :P
[20:47] <dave88> forum*
[20:48] <smo_> ah.. don t think so maybe for transcoding problems but not serviio config/access :p
[20:48] <dave88> :(
[20:48] <smo_> look you debug console in chrome or don t know what you use
[20:48] <smo_> see why you  have arr500
[20:48] <smo_> err
[20:49] <smo_> look like permissions problems too :p
[20:51] <dave88> I thought that but it's owner read/write?
[20:52] <dave88> I've even tried adding the desktop folder - same issue
[21:05] <smo_> ok got it working 100% c_14 i modified mu player , thanks !
[21:05] <smo_> my*
[22:38] <lg188> Hi, I've tried to convert an mkv file to mp4 or ogv. But it says it's not supported. I'm using 2.5.4 is this still accurate ?
[22:38] <c_14> it's probably a codec or somethnig
[22:38] <c_14> *ing
[22:39] <lg188> https://bpaste.net/show/21082d79f89c
[22:40] <c_14> The input is either corrupt or doing weird things. (probably)
[22:40] <c_14> can you try with a static build from git?
[22:40] <c_14> http://johnvansickle.com/ffmpeg/
[22:41] <lg188> It feels kind of insulting to throw a pastebin explanation at me as the first thing. Just saying
[22:41] <lg188> I'll see if I've compiled it with the right useflags brb
[22:41] <shevy> ffmpeg version 2.5.1
[22:42] <shevy> what are the flags?
[22:44] <lg188> shevy: https://bpaste.net/show/5e1d7fb1c91a
[22:45] <lg188> I'll have to add opus for ogv I think
[22:45] <shevy> no sorry
[22:45] <shevy> I meant from ffmpeg -v or something
[22:45] <shevy> my build has this right now "--prefix=/usr --enable-gpl --enable-libmp3lame --enable-libopus --enable-libtheora --enable-libvorbis --enable-libvpx --enable-libxvid --enable-nonfree --enable-postproc --enable-pthreads --enable-shared --enable-version3 --enable-x11grab --extra-libs=-ldl --disable-libopenjpeg --enable-libx264"
[22:46] <lg188> I'm talking about gentoo compilation flags, work a bit different, I assume
[00:00] --- Mon Mar  2 2015


More information about the Ffmpeg-devel-irc mailing list