Recherche avancée

Sur d’autres sites (3536)

• FFMPEG CPU encoding speed

  22 janvier 2020, par natloz

  I would like to increase the speed of encoding videos. Thou having issue with it...

  This is my CPU : Intel(R) Xeon(R) CPU E5-2603 v4 @ 1.70GHz
  OS : Windows 10 PRO 64bit

  I don’t have a graphics card.

  This is the script I’m using for current video encoding :

  ffmpeg -y -re -i o.mov -vf subtitles=SUBS.srt -s 1920x1080 -r 25 -pixel_format yuv420p -aspect 16:9 -top 1 -qp 0 -c:v libx264 -b:v 600000 -minrate 600000 -maxrate 600000 -x264opts bitrate=6000:vbv-maxrate=6000:vbv-bufsize=400:ratetol=400:bitrate=6000:vbv-init=800:force-cfr=1 -g 33 -bf 3 -muxrate 6.5M -flags cgop+ilme+ildct -c:a mp2 -ac 2 -ar 48k -b:a 192k -f mpegts out.mpeg

  The current speed I’m getting is : speed = 1x.

  How can I force my CPU to work at full speed ?

• Python : mp3 to alsaaudio through ffmpeg pipe and wave.open(f,'r')

  3 juillet 2017, par user2754098

  I’m trying to decode mp3 to wav using ffmpeg :

  import alsaaudio
  
  import wave
  
  from subprocess import Popen, PIPE
  
  
  
  with open('filename.mp3', 'rb') as infile:
  
      p=Popen(['ffmpeg', '-i', '-', '-f', 'wav', '-'], stdin=infile, stdout=PIPE)
  
      ...

  Next i want redirect data from p.stdout.read() to wave.open(file, r) to use readframes(n) and other methods. But i cannot because ’file’ in wave.open(file,’r’) can be only name of file or an open file pointer.

      ...
  
      file = wave.open(p.stdout.read(),'r')
  
      card='default'
  
      device=alsaaudio.PCM(card=card)
  
      device.setchannels(file.getnchannels())
  
      device.setrate(file.getframerate())
  
      device.setformat(alsaaudio.PCM_FORMAT_S16_LE)
  
      device.setsetperiodsize(320)
  
      data = file.readframes(320)
  
      while data:
  
          device.write(data)
  
          data = file.readframes(320)

  I got :

  TypeError: file() argument 1 must be encoded string without NULL bytes, not str

  So is it possible to handle data from p.stdout.read() by wave.open() ?
  Making temporary .wav file isn’t solution.

  Sorry for my english.
  Thanks.

  UPDATE

  Thanks to PM 2Ring for hit about io.BytesIO.

  However resulting code does not work.

  import alsaaudio
  
  import wave
  
  from subprocess import Popen, PIPE
  
  
  
  with open('sometrack.mp3', 'rb') as infile:
  
          p=Popen(['ffmpeg', '-i', '-', '-f','wav', '-'], stdin=infile , stdout=PIPE , stderr=PIPE)
  
          fobj = io.BytesIO(p.stdout.read())
  
  fwave = wave.open(fobj, 'rb')

  Trace :

  File "./script.py", line x, in <module>
  
    fwave = wave.open(fobj, 'rb')
  
  File "/usr/lib/python2.7/wave.py", line x, in open
  
    return Wave_read(f)
  
  File "/usr/lib/python2.7/wave.py", line x, in __init__
  
    self.initfp(f)
  
  File "/usr/lib/python2.7/wave.py", line x, in initfp
  
    raise Error, 'not a WAVE file'
  
  wave.Error: not a WAVE file
  
  </module>

  From /usr/lib/python2.7/wave.py :

  ...
  
  self._file = Chunk(file, bigendian = 0)
  
  if self._file.getname() != 'RIFF':
  
      raise Error, 'file does not start with RIFF id'
  
  if self._file.read(4) != 'WAVE':
  
      raise Error, 'not a WAVE file'
  
  ...

  Checking has been failed due to ’bad’ self._file object.

  Inside /usr/lib/python2.7/chunk.py i have found a source of problem :

  ...
  
  try:
  
      self.chunksize = struct.unpack(strflag+'L', file.read(4))[0]
  
  except struct.error:
  
      raise EOFError
  
  ...

  Because struct.unpack(strflag+’L’, file.read(4))[0] returns 0.
  But this function works correct.

  As specified here :

  "5-8 bytes - File size(integer)
  Size of the overall file - 8 bytes, in bytes (32-bit integer). Typically, you’d fill this in after creation."
  That’s why my script doesn’t work. wave.open and other functions cannot handle my file object because self.chunksize = 0. Looks like ffmpeg cannot insert File size when using PIPE.

  SOLUTION

  It’s simple.
  I’ve changed init function of Chunk class :

  After :

  ...
  
  try:
  
      self.chunksize = struct.unpack(strflag+'L', file.read(4))[0]
  
  except struct.error:
  
      raise EOFError
  
  ...

  Before :

  ...
  
  try:
  
      self.chunksize = struct.unpack(strflag+'L', file.read(4))[0]
  
      currtell = file.tell()
  
      if self.chunksize == 0:
  
          file.seek(0)
  
          file.read(currtell)
  
          self.chunksize = len(file.read())-4
  
          file.seek(0)
  
          file.read(currtell)
  
  except struct.error:
  
      raise EOFError
  
  ...

  Of course editing of original module is bad idia. So I’ve create custom forks for 2 classes Chunk and Wave_read.

  Working but unstable full code you can find here.

  Sorry for my awful english.

  Thanks.

• Python : mp3 to alsaaudio through ffmpeg pipe and wave.open(f,'r')

  16 avril 2015, par user2754098

  I’m trying to decode mp3 to wav using ffmpeg :

  import alsaaudio
  
  import wave
  
  from subprocess import Popen, PIPE
  
  
  
  with open('filename.mp3', 'rb') as infile:
  
      p=Popen(['ffmpeg', '-i', '-', '-f', 'wav', '-'], stdin=infile, stdout=PIPE)
  
      ...

  Next i want redirect data from p.stdout.read() to wave.open(file, r) to use readframes(n) and other methods. But i cannot because ’file’ in wave.open(file,’r’) can be only name of file or an open file pointer.

      ...
  
      file = wave.open(p.stdout.read(),'r')
  
      card='default'
  
      device=alsaaudio.PCM(card=card)
  
      device.setchannels(file.getnchannels())
  
      device.setrate(file.getframerate())
  
      device.setformat(alsaaudio.PCM_FORMAT_S16_LE)
  
      device.setsetperiodsize(320)
  
      data = file.readframes(320)
  
      while data:
  
          device.write(data)
  
          data = file.readframes(320)

  I got :

  TypeError: file() argument 1 must be encoded string without NULL bytes, not str

  So is it possible to handle data from p.stdout.read() by wave.open() ?
  Making temporary .wav file isn’t solution.

  Sorry for my english.
  Thanks.

  UPDATE

  Thanks to PM 2Ring for hit about io.BytesIO.

  However resulting code does not work.

  import alsaaudio
  
  import wave
  
  from subprocess import Popen, PIPE
  
  
  
  with open('sometrack.mp3', 'rb') as infile:
  
          p=Popen(['ffmpeg', '-i', '-', '-f','wav', '-'], stdin=infile , stdout=PIPE , stderr=PIPE)
  
          fobj = io.BytesIO(p.stdout.read())
  
  fwave = wave.open(fobj, 'rb')

  Trace :

  File "./script.py", line x, in <module>
  
    fwave = wave.open(fobj, 'rb')
  
  File "/usr/lib/python2.7/wave.py", line x, in open
  
    return Wave_read(f)
  
  File "/usr/lib/python2.7/wave.py", line x, in __init__
  
    self.initfp(f)
  
  File "/usr/lib/python2.7/wave.py", line x, in initfp
  
    raise Error, 'not a WAVE file'
  
  wave.Error: not a WAVE file
  
  </module>

  From /usr/lib/python2.7/wave.py :

  ...
  
  self._file = Chunk(file, bigendian = 0)
  
  if self._file.getname() != 'RIFF':
  
      raise Error, 'file does not start with RIFF id'
  
  if self._file.read(4) != 'WAVE':
  
      raise Error, 'not a WAVE file'
  
  ...

  Checking has been failed due to ’bad’ self._file object.

  Inside /usr/lib/python2.7/chunk.py i have found a source of problem :

  ...
  
  try:
  
      self.chunksize = struct.unpack(strflag+'L', file.read(4))[0]
  
  except struct.error:
  
      raise EOFError
  
  ...

  Because struct.unpack(strflag+’L’, file.read(4))[0] returns 0.
  But this function works correct.

  As specified here :

  "5-8 bytes - File size(integer)
  Size of the overall file - 8 bytes, in bytes (32-bit integer). Typically, you’d fill this in after creation."
  That’s why my script doesn’t work. wave.open and other functions cannot handle my file object because self.chunksize = 0. Looks like ffmpeg cannot insert File size when using PIPE.

  SOLUTION

  It’s simple.
  I’ve changed init function of Chunk class :

  After :

  ...
  
  try:
  
      self.chunksize = struct.unpack(strflag+'L', file.read(4))[0]
  
  except struct.error:
  
      raise EOFError
  
  ...

  Before :

  ...
  
  try:
  
      self.chunksize = struct.unpack(strflag+'L', file.read(4))[0]
  
      currtell = file.tell()
  
      if self.chunksize == 0:
  
          file.seek(0)
  
          file.read(currtell)
  
          self.chunksize = len(file.read())-4
  
          file.seek(0)
  
          file.read(currtell)
  
  except struct.error:
  
      raise EOFError
  
  ...

  Of course editing of original module is bad idia. So I’ve create custom forks for 2 classes Chunk and Wave_read.

  Working but unstable full code you can find here.

  Sorry for my awful english.

  Thanks.