Recherche avancée

Sur d’autres sites (9071)

• Difficulties saving an animation (matplotlib)

  14 avril 2020, par Ben C.

  I have a simple animation that I want to save. I followed the example : https://matplotlib.org/examples/animation/basic_example_writer.html

  



  But I get the following error : RuntimeError : Requested MovieWriter (ffmpeg) not available

  



  I installed ffmpeg and checked via ffmpeg -version that it really is installed and the path correct.

  



  Here is my code :

  



  import numpy as np
import matplotlib.animation as animation
import matplotlib.pyplot as plt

# Set up formatting for the movie files
Writer = animation.writers['ffmpeg']
writer = Writer(fps=15, metadata=dict(artist='Me'), bitrate=1800)

fig, ax = plt.subplots()

x = np.random.uniform(0,1)
y = np.random.uniform(0,1)


scat = ax.scatter(x,y, color= 'blue')
circle = plt.Circle((x,y), radius=0.1, color='blue', fill=False, lw=0.5)
ax.add_patch(circle)

def init():
    scat = ax.scatter(x, y, color = 'blue')
    circle = plt.Circle((x,y), radius=0.1, color='blue', fill=False, lw=0.5)

def animate(i):
    random = np.random.uniform(0,1)
    if random < 0.5:
        scat.set_color('red')
        circle.set_edgecolor('blue')
    else:
        scat.set_color('blue')
        circle.set_edgecolor('red')
    return circle, scat,


ani = animation.FuncAnimation(fig, animate, init_func=init(), interval=1000, blit=True)
ani.save('test.mp4', writer=writer)


  



  None of the proposed solution in RuntimeError : No MovieWriters available in Matplotlib animation worked for me. Any ideas ?

  



  Edit : I am using Windows (10)

  


• FileNotFoundError on aws Lambda when concatenating videos with ffmpeg

  2 juillet 2021, par Shibu Menon

  Goal :

  



  

  • Concat 2 videos (both are in an s3 bucket) via aws Lambda using ffmpeg

  


  • Upload the resultant output.mp4 to another S3 bucket

  


  • Python 3+

  


  



  I've already created a layer containing a static ffmpeg

  



  The Error :

  



  {
  "errorMessage": "[Errno 2] No such file or directory: '/tmp/output.mp4'",
  "errorType": "FileNotFoundError",
  "stackTrace": [
    [
      "/var/task/lambda_function.py",
      19,
      "lambda_handler",
      "s3.Object(bucketLowRes, mp4OutputFileName).put(Body=open(new_file_key, 'rb'))"
    ]
  ]
}


  



  My Lambda function :

  



  import json
import os
import subprocess
import boto3

s3 = boto3.resource('s3')
bucketLowRes = s3.Bucket("bucket-conc-lowres")

def lambda_handler(event, context):
    # TODO implement

    mp4OutputFileName = 'output.mp4'

    new_file_key = os.path.abspath(os.path.join(os.sep, 'tmp', mp4OutputFileName))
    subprocess.call( ['/opt/ffmpeg', '-i', 'concat:s3://bucket-word-clips/00th76kqwfs915hbixycb77y9v3riwsj30.mp4|s3://bucket-word-clips/00uoakp6jyafbu13ycvl6w2i9tj42eux30.mp4', new_file_key ] )

    s3.Object(bucketLowRes, mp4OutputFileName).put(Body=open(new_file_key, 'rb'))

    return {
        'statusCode': 200,
        'body': json.dumps('Hello from Lambda!')
    }


  



  Question :

  



  

  • FileNotFoundError : Where is the output mp4 file of my ffmpeg concat being saved ??

  


  • And if it is being saved to /tmp/output.mp4 , then why the FileNotFoundError ??

  


  



  thanks

  


• Using ffprobe/ffmpeg to extract individual frames and types in playback sequence

  24 janvier 2018, par Hélder

  I am able to extract keyframes using ffmpeg. Something like this that I have been using :

  ffmpeg -i input.mp4 -vf "select='eq(pict_type\,I)" -vsync vfr -qscale:v 2 I-thumbnails-%02d.png -vf "select='eq(pict_type\,B)" -vsync vfr -qscale:v 2 B-thumbnails-%02d.png -vf "select='eq(pict_type\,P)" -vsync vfr -qscale:v 2 P-thumbnails-%02d.png

  Now the issue is, I would like these extracted frames to be in playback sequence, if possible, the way they are extracted should have a timestamp or any way to know that they start in a certain sequence, example, from start to end :

  IBBBIPPB......BI

  but in a way that I can sort the frames in the playback sequence.

  I want to use this to load in python to extract motion vectors but they should all follow certain playback sequence. Any help ?