Recherche avancée

Sur d’autres sites (10394)

• Can not set period/timeout option for ffmepg

  14 novembre 2016, par newBike

  I tried to capture streaming via rtsp and limit the clip duration under 3 sec

  But the option doesn’t work. The ffmpeg won’t be terminated anymore.

  Is there any workaround to fix the problem.

  Because I have to run hundreds of similar commands in a batch with Python script.

  ffmpeg -loglevel verbose   -i rtsp://172.19.1.42/live.sdp -acodec copy -vcodec copy  c0_s1_h264_640x480_30_vbr_500_99_40000000.mp4 -timeout 3 -y

  $ ffmpeg -h ffmpeg version 1.2.4 Copyright (c) 2000-2013 the FFmpeg
  developers built on Nov 22 2013 11:59:59 with Apple LLVM version 5.0
  (clang-500.2.79) (based on LLVM 3.3svn)

  The detailed log at https://gist.github.com/poc7667/8234701

• ffmpeg hls video with piped audio stuttering after period of time

  31 mai 2023, par urico12

  I have an ffmpeg command generating an hls video stream with audio that is piped using pyaudio.

  


  The ffmpeg command...

  


  ffmpeg -f v4l2 -i /dev/video2 -f s16le -c:a pcm_s16le -ac 1 -ar 16000 -i pipe:0 -hls_time 6 
-hls_list_size 100 -hls_flags append_list+delete_segments -strftime 1 -strftime_mkdir 1
-hls_segment_filename recordings/'%s.ts' -force_key_frames expr:'gte(t,n_forced*2)' -vcodec libx264 master.m3u8


  


  Run as a python subprocess with...

  


  from subprocess import run, Popen, PIPE, STDOUT, DEVNULL
record_command = Popen(command, 
                       shell=True, 
                       stdin=PIPE, 
                       stdout=DEVNULL, 
                       stderr=STDOUT)


  


  And then I use pyaudio to pipe audio segments...

  


  audio_stream = PyAudio().open(
               rate=16000,
               format=pyaudio.paInt16,
               channels=1,
               input=True,
               input_device_index=DEVICE_INDEX,
               frames_per_buffer=320,
               stream_callback=callback)


  


  Where my callback function is...

  


  def callback(self, in_data, frame_count, time_info, status):
    data_arr = np.frombuffer(in_data, dtype=np.int16)
    if self.data_arr_full is None:
        self.data_arr_full = data_arr
    else:
        self.data_arr_full = np.concatenate((self.data_arr_full, data_arr))
    if len(self.data_arr_full) == 16000:
        frame_data = librosa.util.buf_to_float(self.data_arr_full, n_bytes=2, dtype=np.int16)
        # performing audio classification on frame_data
        ...
        record_command.stdin.write(self.data_arr_full.tobytes())
        self.data_arr_full = None
    self.frame = data_arr.tobytes()
    
    return (None, pyaudio.paContinue)


  


  This captures video from a camera device as hls and pipes audio segments in from a microphone device (after it has been run through an image classification algorithm). It works quite well for a period of time, but after running for a while the recorded video begins to stutter while the audio remains smooth. It may be important to note that I am running this from a jetson nano. I'm far from an expert in ffmpeg so any help is appreciated !

  


• avdevice/v4l2 : only use frame period from v4l2 if valid

  2 février 2014, par Michael Niedermayer

  avdevice/v4l2 : only use frame period from v4l2 if valid
  There is evidence that some drivers do not set a valid value
  See : http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=718805
  
  Signed-off-by : Michael Niedermayer <michaelni@gmx.at>
  

  • [DH] libavdevice/v4l2.c