Recherche avancée

Sur d’autres sites (7032)

• Moviepy : subclip fails when iterating through CSV

  5 mai 2016, par user3316291

  I am trying to randomize the order of video clips based on timing from a CSV file and then reassemble the randomized clips into a single video. However, I am receiving an error in the loop that iterates through each clip timing to create the subclip.

  Here is what my CSV looks like :

  00:00:32.18,00:00:52.10,1
  
  00:00:52.11,00:00:56.09,2
  
  00:00:56.10,00:00:58.15,3
  
  00:00:58.16,00:01:05.16,4
  
  00:01:05.17,00:01:16.04,5

  column 1 is clip onset
  column 2 is clip offset
  column 3 is scene number that I use to randomize

  Here is my code :

  import os
  
  import csv
  
  import numpy as np
  
  from moviepy.editor import *
  
  
  
  f = open('SceneCuts.csv')
  
  csv_f = csv.reader(f)
  
  
  
  scenes = []
  
  ons = []
  
  offs = []
  
  for row in csv_f:
  
      ons.append(row[0])
  
      offs.append(row[1])
  
      scenes.append(row[2])
  
  
  
  r_scene = scenes
  
  np.random.seed(1000)
  
  np.random.shuffle(r_scene)
  
  r_scene = map(int, r_scene)
  
  
  
  clip = VideoFileClip("FullVideo.m4v")
  
  
  
  temp = []
  
  for row in r_scene:
  
      print(row)
  
      temp.append(clip.subclip(ons[row-1], offs[row-1]))
  
  
  
  catclip = concatenate_videoclips(temp)
  
  catclip_resize = catclip.resize((1024,576))
  
  catclip_resize.write_videofile("RandomVideo.mp4")

  Here is the output, error occurs at line 29 (temp.append)

  File "/Users/Dustin/anaconda/lib/python2.7/site-packages/moviepy/video/io/ffmpeg_reader.py", line 87, in initialize
  
  self.proc = sp.Popen(cmd, **popen_params)
  
  
  
  File "/Users/Dustin/anaconda/lib/python2.7/subprocess.py", line 710, in __init__
  
  errread, errwrite)
  
  
  
  File "/Users/Dustin/anaconda/lib/python2.7/subprocess.py", line 1316, in _execute_child
  
  data = _eintr_retry_call(os.read, errpipe_read, 1048576)
  
  
  
  File "/Users/Dustin/anaconda/lib/python2.7/subprocess.py", line 476, in _eintr_retry_call
  
  return func(*args)
  
  
  
  OSError: [Errno 22] Invalid argument

  Based on my research, it appears to be something regarding child processes and subprocess.Popen, but I can’t figure it out. Thanks !

  EDIT to add new information :
  I have been running the above script in Spyder (anaconda) and receiving the above errors. However, when I run from a terminal or sublime (cmd+b), the code "works". It runs and I do not get the above error, however, the resulting video file is a mess. There are multiple conflicting audio tracks that shouldn’t be there. I am not sure what is going on in Spyder, but I’d love to know. Also, I still need to fix the audio problem.

• C# FFmpeg Cannot process request because the process (11536) has exited

  30 avril 2016, par Martin Mazza Dawson

  I’m trying to convert an .ogg file to .mp3. I am following this link to do so : How to call ffmpeg.exe to convert audio files on Windows Azure ?

  I get the error message :

  Cannot process request because the process (11536) has exited.

  The ffmpeg.exe is in my root dir and it works when I convert the .ogg file to .mp3 in the command prompt, the ogg_path also has the correct bytes written to it when I check using debugging.
  Can anyone tell me what I am doing wrong please ?

      [HttpPost]
  
      public void ConvertFileToMp3(HttpPostedFileBase file)
  
      {
  
          string ogg_path = Server.MapPath("~/0477.ogg");
  
          string mp3_path = Server.MapPath("~/0477.mp3");
  
  
  
          using (var reader = new BinaryReader(file.InputStream))
  
          {
  
              System.IO.File.WriteAllBytes(ogg_path, reader.ReadBytes(file.ContentLength));
  
          }
  
  
  
          ProcessStartInfo psi = new ProcessStartInfo();
  
          psi.FileName = Server.MapPath("~/ffmpeg.exe");
  
          psi.Arguments = string.Format(@"-i ""{0}"" -y ""{1}""", ogg_path, mp3_path);
  
          psi.CreateNoWindow = true;
  
          psi.ErrorDialog = false;
  
          psi.UseShellExecute = false;
  
          psi.WindowStyle = ProcessWindowStyle.Hidden;
  
          psi.RedirectStandardOutput = true;
  
          psi.RedirectStandardInput = false;
  
          psi.RedirectStandardError = true;
  
          Process exeProcess = Process.Start(psi);
  
  
  
          try
  
          {
  
              exeProcess.PriorityClass = ProcessPriorityClass.High; //Fails here
  
              string outString = string.Empty;
  
              exeProcess.OutputDataReceived += (s, e) => {
  
                  outString += e.Data;
  
              };
  
              exeProcess.BeginOutputReadLine();
  
              string errString = exeProcess.StandardError.ReadToEnd();
  
              Trace.WriteLine(outString);
  
              Trace.TraceError(errString);
  
              exeProcess.WaitForExit();
  
          }
  
          catch (Exception e)
  
          {
  
              Trace.TraceError(e.Message);
  
          }
  
  
  
          //do other stuff with file, cut for brevity
  
      }

• python imageIO() ffmpeg output 3D ndarray

  27 juillet 2017, par Daimoj

  I’m trying to encode and than decode a collection of images using imageIO in python with the ffmpeg plugin and the HEVC codec.

  The stream I’m using is an ndarray of shape (1024,512). When I use the writer.append_data() on each image, the shape is as above (1024,512). After writer.close() is called, I create another reader on the video just made from above. When interrogating a single image of the video it’s shape is (1024,512,3). This is all grayscale, so I only expected an array of uint8’s in the shape of (1024,512). Why did ImageIO add 2 more dimensions to my video ? I only want one.