Newest 'ffmpeg' Questions - Stack Overflow

Les articles publiés sur le site

• raspberry pi, ffmpeg, libmp3lame stream to icecast2 fails with : av_interleaved_write_frame() : Connection reset by peer

  9 avril 2016, par seekwhencer

  finally, i want to merge or mix audio channels from two streams into one new audio stream. but at first i want to relay or reencode an existing stream from an icecast2 server - and send it back to the icecast as new mountpoint. this stream comes from mpd on the same raspberry pi 2. and it works fine.

  i've got lame and ffmpeg on this way:

  cd /usr/src
  git clone https://github.com/rbrito/lame
  cd lame
  sudo ./configure
  sudo make
  sudo make install
  

  then ffmpeg

  cd /usr/src
  sudo rm -R ffmpeg
  git clone git://source.ffmpeg.org/ffmpeg.git
  cd ffmpeg
  sudo ./configure --arch=armel --target-os=linux --enable-gpl --enable-libmp3lame --enable-nonfree
  sudo make
  sudo make install
  

  A simple test fails with:

  **av_interleaved_write_frame(): Connection reset by peer
  Error writing trailer of icecast://admin:hackme@127.0.0.1:8000/listen: Connection reset by peersize=       5kB time=00:00:00.28 bitrate= 144.7kbits/s speed=1.73x**
  

  (with strange spaces)

  The reencoding ffmpeg script:

  ffmpeg \
   -i http://127.0.0.1:8000/live \
   -c:a libmp3lame -b:a 128k \
   -legacy_icecast 1 \
   -content_type audio/mpeg \
   -ice_name "hack radio" \
   -f mp3 icecast://admin:hackme@127.0.0.1:8000/listen \
   -y \
   -stats \
   -report
  

  I tried legacy_icecast 1 or 0.

  The ffmpeg report:

  ffmpeg started on 2016-04-09 at 00:57:56
  Report written to "ffmpeg-20160409-005756.log"
  Command line:
  ffmpeg -i http://127.0.0.1:8000/live -muxpreload 10 -c:a libmp3lame -b:a 128k -legacy_icecast 1 -content_type audio/mpeg -ice_name "NFS One Radio" -f mp3 icecast://admin:hackme@127.0.0.1:8000/listen -y -stats -report
  ffmpeg version N-79262-gae8a13c Copyright (c) 2000-2016 the FFmpeg developers
    built with gcc 4.6 (Debian 4.6.3-14+rpi1)
    configuration: --arch=armel --target-os=linux --enable-gpl --enable-libmp3lame --enable-nonfree
    libavutil      55. 20.100 / 55. 20.100
    libavcodec     57. 34.100 / 57. 34.100
    libavformat    57. 31.100 / 57. 31.100
    libavdevice    57.  0.101 / 57.  0.101
    libavfilter     6. 41.100 /  6. 41.100
    libswscale      4.  1.100 /  4.  1.100
    libswresample   2.  0.101 /  2.  0.101
    libpostproc    54.  0.100 / 54.  0.100
  Splitting the commandline.
  Reading option '-i' ... matched as input file with argument 'http://127.0.0.1:8000/live'.
  Reading option '-muxpreload' ... matched as option 'muxpreload' (set the initial demux-decode delay) with argument '10'.
  Reading option '-c:a' ... matched as option 'c' (codec name) with argument 'libmp3lame'.
  Reading option '-b:a' ... matched as option 'b' (video bitrate (please use -b:v)) with argument '128k'.
  Reading option '-legacy_icecast' ... matched as AVOption 'legacy_icecast' with argument '1'.
  Reading option '-content_type' ... matched as AVOption 'content_type' with argument 'audio/mpeg'.
  Reading option '-ice_name' ... matched as AVOption 'ice_name' with argument 'NFS One Radio'.
  Reading option '-f' ... matched as option 'f' (force format) with argument 'mp3'.
  Reading option 'icecast://admin:hackme@127.0.0.1:8000/listen' ... matched as output file.
  Reading option '-y' ... matched as option 'y' (overwrite output files) with argument '1'.
  Reading option '-stats' ... matched as option 'stats' (print progress report during encoding) with argument '1'.
  Reading option '-report' ... matched as option 'report' (generate a report) with argument '1'.
  Finished splitting the commandline.
  Parsing a group of options: global .
  Applying option y (overwrite output files) with argument 1.
  Applying option stats (print progress report during encoding) with argument 1.
  Applying option report (generate a report) with argument 1.
  Successfully parsed a group of options.
  Parsing a group of options: input file http://127.0.0.1:8000/live.
  Successfully parsed a group of options.
  Opening an input file: http://127.0.0.1:8000/live.
  [http @ 0x343cee0] Setting default whitelist 'http,https,tls,rtp,tcp,udp,crypto,httpproxy'
  [http @ 0x343cee0] request: GET /live HTTP/1.1
  User-Agent: Lavf/57.31.100
  Accept: */*
  Range: bytes=0-
  Connection: close
  Host: 127.0.0.1:8000
  Icy-MetaData: 1
  
  
  [mp3 @ 0x343c7e0] Format mp3 probed with size=262144 and score=50
  [mp3 @ 0x343c7e0] Skipping 656 bytes of junk at 0.
  [mp3 @ 0x343c7e0] Before avformat_find_stream_info() pos: 656 bytes read:272000 seeks:0
  [mp3 @ 0x343c7e0] All info found
  [mp3 @ 0x343c7e0] After avformat_find_stream_info() pos: 42640 bytes read:272000 seeks:0 frames:50
  Input #0, mp3, from 'http://127.0.0.1:8000/live':
    Metadata:
      icy-br          : 256
      icy-name        : NFS One Radio
      icy-pub         : 0
      StreamTitle     : x
    Duration: N/A, start: 0.000000, bitrate: 256 kb/s
      Stream #0:0, 50, 1/14112000: Audio: mp3, 44100 Hz, stereo, s16p, 256 kb/s
  Successfully opened the file.
  Parsing a group of options: output file icecast://admin:hackme@127.0.0.1:8000/listen.
  Applying option muxpreload (set the initial demux-decode delay) with argument 10.
  Applying option c:a (codec name) with argument libmp3lame.
  Applying option b:a (video bitrate (please use -b:v)) with argument 128k.
  Applying option f (force format) with argument mp3.
  Successfully parsed a group of options.
  Opening an output file: icecast://admin:hackme@127.0.0.1:8000/listen.
  [icecast @ 0x34492c0] No default whitelist set
  [http @ 0x344b9e0] Setting default whitelist 'http,https,tls,rtp,tcp,udp,crypto,httpproxy'
  [http @ 0x344b9e0] request: SOURCE /listen HTTP/1.1
  User-Agent: Lavf/57.31.100
  Accept: */*
  Connection: close
  Host: 127.0.0.1:8000
  Content-Type: audio/mpeg
  Icy-MetaData: 1
  Ice-Name: NFS One Radio
  Ice-Public: 0
  Authorization: Basic YWRtaW46bWFyc2Jhc2U=
  

  The Output from ffmpeg:

  ffmpeg started on 2016-04-09 at 01:06:46
  Report written to "ffmpeg-20160409-010646.log"
  ffmpeg version N-79262-gae8a13c Copyright (c) 2000-2016 the FFmpeg developers
    built with gcc 4.6 (Debian 4.6.3-14+rpi1)
    configuration: --arch=armel --target-os=linux --enable-gpl --enable-libmp3lame --enable-nonfree
    libavutil      55. 20.100 / 55. 20.100
    libavcodec     57. 34.100 / 57. 34.100
    libavformat    57. 31.100 / 57. 31.100
    libavdevice    57.  0.101 / 57.  0.101
    libavfilter     6. 41.100 /  6. 41.100
    libswscale      4.  1.100 /  4.  1.100
    libswresample   2.  0.101 /  2.  0.101
    libpostproc    54.  0.100 / 54.  0.100
  [mp3 @ 0x25ec7e0] Skipping 501 bytes of junk at 0.
  Input #0, mp3, from 'http://127.0.0.1:8000/live':
    Metadata:
      icy-br          : 256
      icy-name        : NFS One Radio
      icy-pub         : 0
      StreamTitle     : x
    Duration: N/A, start: 0.000000, bitrate: 256 kb/s
      Stream #0:0: Audio: mp3, 44100 Hz, stereo, s16p, 256 kb/s
  Output #0, mp3, to 'icecast://admin:hackme@127.0.0.1:8000/listen':
    Metadata:
      icy-br          : 256
      icy-name        : NFS One Radio
      icy-pub         : 0
      StreamTitle     : x
      TSSE            : Lavf57.31.100
      Stream #0:0: Audio: mp3 (libmp3lame), 44100 Hz, stereo, s16p, 128 kb/s
      Metadata:
        encoder         : Lavc57.34.100 libmp3lame
  Stream mapping:
    Stream #0:0 -> #0:0 (mp3 (native) -> mp3 (libmp3lame))
  Press [q] to stop, [?] for help
  av_interleaved_write_frame(): Connection reset by peer
  Error writing trailer of icecast://admin:hackme@127.0.0.1:8000/listen: Connection reset by peersize=      16kB time=00:00:00.96 bitrate= 132.9kbits/s speed=3.14x
  video:0kB audio:16kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 1.183730%
  

  The icecast works fine with mpd. And icecast creates a new mountpoint when feeded.

  For some reason the connection to the icecast drops. Ideas?

  Thanks

• How to add libmp3lame, libx264, libvo_aacenc and libvorbis codecs to existing ffmpeg install on Redhat 6.7

  9 avril 2016, par rjn

  Adobe cq uses ffmpeg to transcode videos. There is a workflow that is triggered when a user uploads any video to the digital asset management (DAM) section. This workflow internally uses ffmpeg to transcode the uploaded video and spits out additional video renditions in formats like ogg, m4v, mp3.

  The server (Red Hat Enterprise Linux Server release 6.5) on which adobe cq runs needs to have ffmpeg installed for the DAM workflow to function. ffmpeg has already installed been installed, but it is missing certain codecs: libmp3lame, libx264, libvo_aacenc and libvorbis.

  How to add and enable these codecs ?

• Decoding EVRC speech codec

  8 avril 2016, par NitinG

  I want to decode the EVRC speech codec. I have checked the ffmpeg library but it seems like, EVRC is not being currently supported by ffmpeg.

  Is there any alternate library which can be used to decode EVRC data?

  Or if anybody have some algorithm or decoding mechanism or code snippet to do so, please let me know.

  Thanks Nitin

• Recursively convert media directory from HEVC to h.264 with ffmpeg

  8 avril 2016, par chuckcastle

  I have media server with two directories: Movies and TV Shows. Within each of those directories, each entry exists in a sub-directory which contains the video file and subtitle files.

  I've scoured the web and have found an excellent perl script from Michelle Sullivan, posted here:

      #!/usr/bin/perl
  
  use strict;
  use warnings;
  
  open DIR, "ls -1 |";
  while ()
  {
          chomp;
          next if ( -d "$_"); # skip directories
          next unless ( -r "$_"); # if it's not readable skip it!
          my $file = $_;
          open PROBE, "ffprobe -show_streams -of csv '$file' 2>/dev/null|" or die ("Unable to launch ffmpeg for $file! ($!)");
          my ($v, $a, $s, @c) = (0,0,0);
          while ()
          {
                  my @streaminfo = split(/,/, $_);
                  push(@c, $streaminfo[2]) if ($streaminfo[5] eq "video");
                  $a++ if ($streaminfo[5] eq "audio");
                  $s++ if ($streaminfo[5] eq "subtitle");
          }
          close PROBE;
          $v = scalar @c;
          if (scalar @c eq 1 and $c[0] eq "ansi")
          {
                  warn("Text file detected, skipping...\n");
                  next;
          }
          warn("$file: Video Streams: $v, Audio Streams: $a, Subtitle Streams: $s, Video Codec(s): " . join (", ", @c) . "\n");
          if (scalar @c > 1)
          {
                  warn("$file has more than one video stream, bailing!\n");
                  next;
          }
          if ($c[0] eq "hevc")
          {
                  warn("HEVC detected for $file ...converting to AVC...\n");
                  system("mkdir -p h265");
                  my @params = ("-hide_banner", "-threads 2");
                  push(@params, "-map 0") if ($a > 1 or $s > 1 or $v > 1);
                  push(@params, "-c:a copy") if ($a);
                  push(@params, "-c:s copy") if ($s);
                  push(@params, "-c:v libx264 -pix_fmt yuv420p") if ($v);
                  if (system("mv '$file' 'h265/$file'"))
                  {
                          warn("Error moving $file -> h265/$file\n");
                          next;
                  }
                  if (system("ffmpeg -xerror -i 'h265/$file' " . join(" ", @params) . " '$file' 2>/dev/null"))
                  {
                          warn("FFMPEG ERROR.  Cannot convert $file restoring original...\n");
                          system("mv 'h265/$file' '$file'");
                          next;
                  }
          } else {
                  warn("$file doesn't appear to need converting... Skipping...\n");
          }
  }
  close DIR;
  

  The script performs perfectly - as long as it is run from within the directory containing the media.

  My question: Can this script be modified to run recursively from the root directory? How?

  Thanks in advance.

  (Michelle's script can be seen here: http://www.michellesullivan.org/blog/1636)

• How can I limit the file size when doing a write() while writing data from a video stream in Python

  8 avril 2016, par Michael Bailey

  I am pulling in a video stream of an unknown length and writing that file to disc using .write() in Python. My basic need is to have a rolling 2-3 minutes of video recorded to that file at all times VS having that file grow to a huge size. On the other end I'm using FFMPEG to grab clips from the .dat file being written.

  with open("./site/files/tmp/" + hashid + ".dat", 'wb') as f:
          while True:
              data = stream.read(1024)
              f.write(data)
  

  Alternately if you know a way I could store that same data in memory probably only 20-50mb (2-3 mins) and send that to FFMPEG that might be an even better option as well. I'm calling FFMPEG via Popen subprocess.