Recherche avancée

Sur d’autres sites (12912)

• ffmpeg failing to add png mask to video : Requested planes not available

  23 août 2022, par Alexandr Sugak

  I am trying to add png mask to make webm video round (cut off its corners).

  


  The command I am using :

  


  video="./dist/tmp/19_2.webm"
mask="./dist/tmp/mask.png"
output="./dist/tmp/circle.webm"

ffmpeg -report -c:v libvpx-vp9 -i "${video}" -loop 1 -i "${mask}" -filter_complex " \
[1:v]alphaextract[alf];\
[0:v][alf]alphamerge" \
-c:a copy -c:v libvpx-vp9 "${output}"


  


  The command output :

  


  sh ./scripts/video_mask.sh 
ffmpeg started on 2022-08-23 at 17:27:48
Report written to "ffmpeg-20220823-172748.log"
Log level: 48
ffmpeg version 5.1-tessus Copyright (c) 2000-2022 the FFmpeg developers
  built with Apple clang version 11.0.0 (clang-1100.0.33.17)
  configuration: --cc=/usr/bin/clang --prefix=/opt/ffmpeg --extra-version=tessus --enable-avisynth --enable-fontconfig --enable-gpl --enable-libaom --enable-libass --enable-libbluray --enable-libdav1d --enable-libfreetype --enable-libgsm --enable-libmodplug --enable-libmp3lame --enable-libmysofa --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenh264 --enable-libopenjpeg --enable-libopus --enable-librubberband --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvmaf --enable-libvo-amrwbenc --enable-libvorbis --enable-libvpx --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxavs --enable-libxvid --enable-libzimg --enable-libzmq --enable-libzvbi --enable-version3 --pkg-config-flags=--static --disable-ffplay
  libavutil      57. 28.100 / 57. 28.100
  libavcodec     59. 37.100 / 59. 37.100
  libavformat    59. 27.100 / 59. 27.100
  libavdevice    59.  7.100 / 59.  7.100
  libavfilter     8. 44.100 /  8. 44.100
  libswscale      6.  7.100 /  6.  7.100
  libswresample   4.  7.100 /  4.  7.100
  libpostproc    56.  6.100 / 56.  6.100
[libvpx-vp9 @ 0x7fa072f05140] v1.11.0-30-g888bafc78
    Last message repeated 1 times
Input #0, matroska,webm, from './dist/tmp/19_2.webm':
  Metadata:
    ENCODER         : Lavf59.27.100
  Duration: 00:00:02.77, start: -0.007000, bitrate: 308 kb/s
  Stream #0:0(eng): Video: vp9 (Profile 0), yuva420p(tv, unknown/bt709/iec61966-2-1, progressive), 640x480, SAR 1:1 DAR 4:3, 1k tbr, 1k tbn (default)
    Metadata:
      ALPHA_MODE      : 1
      ENCODER         : Lavc59.37.100 libvpx-vp9
      DURATION        : 00:00:02.744000000
  Stream #0:1(eng): Audio: opus, 48000 Hz, mono, fltp (default)
    Metadata:
      ENCODER         : Lavc59.37.100 libopus
      DURATION        : 00:00:02.767000000
Input #1, png_pipe, from './dist/tmp/mask.png':
  Duration: N/A, bitrate: N/A
  Stream #1:0: Video: png, pal8(pc), 640x480 [SAR 2835:2835 DAR 4:3], 25 fps, 25 tbr, 25 tbn
[libvpx-vp9 @ 0x7fa082f04880] v1.11.0-30-g888bafc78
Stream mapping:
  Stream #0:0 (libvpx-vp9) -> alphamerge
  Stream #1:0 (png) -> alphaextract:default
  alphamerge:default -> Stream #0:0 (libvpx-vp9)
  Stream #0:1 -> #0:1 (copy)
Press [q] to stop, [?] for help
[libvpx-vp9 @ 0x7fa082f04880] v1.11.0-30-g888bafc78
[Parsed_alphaextract_0 @ 0x7fa083906e80] Requested planes not available.
[Parsed_alphaextract_0 @ 0x7fa083906e80] Failed to configure input pad on Parsed_alphaextract_0
Error reinitializing filters!
Failed to inject frame into filter network: Invalid argument
Error while processing the decoded data for stream #0:0
Conversion failed!


  


  I've tried different combination of codecs and pixel formats but I still get the same error. My initial understanding was that ffmpeg fails to find the alpha channel in the input video. By setting -c:v libvpx-vp9 option it looks like ffmpeg correctly picks up yuva420p pixel format but it still gives the same error.

  


  What I am doing wrong ?

  


  Update : if I remove the alphaextract step as suggested in comments, the ffmpeg starts processing video indefinitely (the video I use to test is only 2 sec long). If I specify the number of frames manually, then the output is generated but the mask does not seem to have any effect :

  


  ffmpeg -c:v libvpx-vp9 -i "${video}" -loop 1 -i "${mask}" -filter_complex " \
[0:v][1:v]alphamerge" \
-c:a copy -b:v 2000k -vframes 60 "${output}"


  


   sh ./scripts/video_mask.sh 
ffmpeg version 5.1-tessus Copyright (c) 2000-2022 the FFmpeg developers
  built with Apple clang version 11.0.0 (clang-1100.0.33.17)
  configuration: --cc=/usr/bin/clang --prefix=/opt/ffmpeg --extra-version=tessus --enable-avisynth --enable-fontconfig --enable-gpl --enable-libaom --enable-libass --enable-libbluray --enable-libdav1d --enable-libfreetype --enable-libgsm --enable-libmodplug --enable-libmp3lame --enable-libmysofa --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenh264 --enable-libopenjpeg --enable-libopus --enable-librubberband --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvmaf --enable-libvo-amrwbenc --enable-libvorbis --enable-libvpx --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxavs --enable-libxvid --enable-libzimg --enable-libzmq --enable-libzvbi --enable-version3 --pkg-config-flags=--static --disable-ffplay
  libavutil      57. 28.100 / 57. 28.100
  libavcodec     59. 37.100 / 59. 37.100
  libavformat    59. 27.100 / 59. 27.100
  libavdevice    59.  7.100 / 59.  7.100
  libavfilter     8. 44.100 /  8. 44.100
  libswscale      6.  7.100 /  6.  7.100
  libswresample   4.  7.100 /  4.  7.100
  libpostproc    56.  6.100 / 56.  6.100
[libvpx-vp9 @ 0x7fdd6b005f00] v1.11.0-30-g888bafc78
    Last message repeated 1 times
Input #0, matroska,webm, from './dist/tmp/19_2.webm':
  Metadata:
    ENCODER         : Lavf59.27.100
  Duration: 00:00:02.77, start: -0.007000, bitrate: 308 kb/s
  Stream #0:0(eng): Video: vp9 (Profile 0), yuva420p(tv, unknown/bt709/iec61966-2-1, progressive), 640x480, SAR 1:1 DAR 4:3, 1k tbr, 1k tbn (default)
    Metadata:
      ALPHA_MODE      : 1
      ENCODER         : Lavc59.37.100 libvpx-vp9
      DURATION        : 00:00:02.744000000
  Stream #0:1(eng): Audio: opus, 48000 Hz, mono, fltp (default)
    Metadata:
      ENCODER         : Lavc59.37.100 libopus
      DURATION        : 00:00:02.767000000
Input #1, png_pipe, from './dist/tmp/mask.png':
  Duration: N/A, bitrate: N/A
  Stream #1:0: Video: png, pal8(pc), 640x480 [SAR 2835:2835 DAR 4:3], 25 fps, 25 tbr, 25 tbn
File './dist/tmp/circle.webm' already exists. Overwrite? [y/N] y
[libvpx-vp9 @ 0x7fdd6b007ec0] v1.11.0-30-g888bafc78
Stream mapping:
  Stream #0:0 (libvpx-vp9) -> alphamerge
  Stream #1:0 (png) -> alphamerge
  alphamerge:default -> Stream #0:0 (libvpx-vp9)
  Stream #0:1 -> #0:1 (copy)
Press [q] to stop, [?] for help
[libvpx-vp9 @ 0x7fdd6b007ec0] v1.11.0-30-g888bafc78
[libvpx-vp9 @ 0x7fdd6b024580] v1.11.0-30-g888bafc78
Output #0, webm, to './dist/tmp/circle.webm':
  Metadata:
    encoder         : Lavf59.27.100
  Stream #0:0: Video: vp9, yuva420p(tv, unknown/bt709/iec61966-2-1, progressive), 640x480 [SAR 1:1 DAR 4:3], q=2-31, 2000 kb/s, 1k fps, 1k tbn
    Metadata:
      encoder         : Lavc59.37.100 libvpx-vp9
    Side data:
      cpb: bitrate max/min/avg: 0/0/0 buffer size: 0 vbv_delay: N/A
  Stream #0:1(eng): Audio: opus, 48000 Hz, mono, fltp (default)
    Metadata:
      ENCODER         : Lavc59.37.100 libopus
      DURATION        : 00:00:02.767000000
frame=   60 fps= 16 q=2.0 Lsize=     285kB time=00:00:01.98 bitrate=1175.5kbits/s speed=0.526x    
video:270kB audio:11kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 1.529399%


  


• FFMPEG Loudnorm reading JSON data

  9 juillet 2022, par NineCattoRules

  I tried to normalize some audio files using FFMPEG Loudnorm as described here.

  


  However, in Python, I don't understand how to read data info from 1st pass.

  


  My code :

  


  getLoud =  subprocess.Popen(f"ffmpeg -i {file_path} -filter:a loudnorm=print_format=json -f null NULL", shell=True, stdout=subprocess.PIPE).stdout
getLoud =  getLoud.read().decode()
# parse json_str:
jsonstr_loud = json.loads(getLoud)


  


  This gives me "errorMessage": "Expecting value: line 1 column 1 (char 0)"

  


  I tried also this :

  


  os.system(f"ffmpeg -i {file_path} -filter:a loudnorm=print_format=json -f null NULL")


  


  and it outputs :

  


  ffmpeg version N-60236-gffb000fff8-static https://johnvansickle.com/ffmpeg/  Copyright (c) 2000-2022 the FFmpeg developers...
...
[Parsed_loudnorm_0 @ 0x5921940] 
{
    "input_i" : "-9.33",
    "input_tp" : "-0.63",
    "input_lra" : "0.60",
    "input_thresh" : "-19.33",
    "output_i" : "-24.08",
    "output_tp" : "-15.40",
    "output_lra" : "0.60",
    "output_thresh" : "-34.08",
    "normalization_type" : "dynamic",
    "target_offset" : "0.08"
}


  


  In Python, how can I use those parameters, such as input_i, input_tp etc. that I need for the 2nd pass ?

  


  I can't use ffmpeg-normalize because I'm using FFMPEG as a Layer in Lambda.

  


• OSError : [Errno 9] Bad file descriptor when downloading using pytube and playing with discord.py

  15 septembre 2022, par Trevor Mathisen

  When using pytube to download a YouTube video and discord.py to play it, I am getting a OSError : [Errno 9] Bad file descriptor error. I've had this working at one point and can't seem to figure out what I changed which broke it.

  


  Full traceback :

  


  2022-09-15 20:20:44 INFO       discord.voice_client The voice handshake is being terminated for Channel ID 902294184994693224 (Guild ID 902294184994693220)
2022-09-15T20:20:44.010142763Z 2022-09-15 20:20:44 INFO     discord.voice_client Disconnecting from voice normally, close code 1000.
2022-09-15T20:20:44.058592513Z 2022-09-15 20:20:44 ERROR    discord.player Exception in voice thread Thread-5
2022-09-15T20:20:44.058623864Z Traceback (most recent call last):
2022-09-15T20:20:44.058629130Z   File "/usr/local/lib/python3.10/dist-packages/discord/player.py", line 698, in run
2022-09-15T20:20:44.058632003Z     self._do_run()
2022-09-15T20:20:44.058634500Z   File "/usr/local/lib/python3.10/dist-packages/discord/player.py", line 691, in _do_run
2022-09-15T20:20:44.058637013Z     play_audio(data, encode=not self.source.is_opus())
2022-09-15T20:20:44.058639334Z   File "/usr/local/lib/python3.10/dist-packages/discord/voice_client.py", line 683, in send_audio_packet
2022-09-15T20:20:44.058648759Z     self.socket.sendto(packet, (self.endpoint_ip, self.voice_port))
2022-09-15T20:20:44.058653057Z OSError: [Errno 9] Bad file descriptor
2022-09-15T20:20:44.058673762Z 2022-09-15 20:20:44 INFO     discord.player ffmpeg process 12 has not terminated. Waiting to terminate...
2022-09-15T20:20:44.062083854Z 2022-09-15 20:20:44 INFO     discord.player ffmpeg process 12 should have terminated with a return code of -9.


  


  Dockerfile :

  


  ENV MEDIA_DIR='/media/'
RUN mkdir -p /media


  


  Download function (check_path returns True, showing the file is there) :

  


  def download_videos(stream, filename):
    print(f'downloading {filename}')
    filepath = stream.download(output_path=config_db.environ_path, filename=f'{filename}.mp4')
    print(f'completed download of {filepath}')
    check_path = Path(filepath)
    if check_path.is_file():
        print(check_path)
    return filepath


  


  Play function :

  


  async def play(voice_channel, message, control = None):
    vc = get_vc()
    if not vc:
        vc = await voice_channel.connect()
    next_yt = YouTube(next_song)
    next_file = sub(r'\W+', '', next_yt.title.replace(' ', '_').replace('__', '_')).lower()
    next_song_path = download_videos(next_yt.streams.filter(only_audio=True, file_extension='mp4')[0], next_file)

    await message.channel.send(f'Done getting ready, I\'ll be there in a moment.')

while next_song:
    while vc.is_playing():
        await asynciosleep(0.5)
        continue

    try:
        vc.play(FFmpegPCMAudio(next_song_path))
        print(f'Playing {next_song_path} with latency of {vc.average_latency}')
        vc.source = PCMVolumeTransformer(vc.source, volume=0.15)
    except Exception as e:
        print(e)
        await vc.disconnect()
        next_song = None
        return
    next_song = sounds_db.get_next_song()
    next_yt = YouTube(next_song) if next_song else None
    next_file = sub(r'\W+', '', next_yt.title.replace(' ', '_').replace('__', '_')).lower() if next_song else None
    next_song_path = download_videos(next_yt.streams.filter(only_audio=True, file_extension='mp4')[0], next_file) if next_song else None


  


  I've mounted the /media/ dir during docker -v and can see the file is getting downloaded and when I copy the file to my local machine I can play it in an audio player.

  


  The program can access the SQLite database right next to the files in question just fine.

  


  I've deployed the container locally and to two different VPS's with the same file structure with the same behavior. I'm ripping my hair out.