Recherche avancée

Sur d’autres sites (8244)

• Video - ffmpeg wrong frame number when merging mkv files

  9 juin 2014, par communications

  im trying to join several mkv files together with ffmpeg.

  they all have the same encoding settings.

  i think ffmpeg joins the files, because the filesize of the resulting file has the size of all others combined

  BUT when i open the file, only the number of frames of the last part are shown, so i guess its a problem in the header, where is says how long the video is...

  How can i fix that ?

  Heres the command i use :

  ffmpeg -i "concat:100frames_part0.mkv|100frames_part1.mkv|100frames_part3.mkv" -c copy 100frames.mkv

• Kivy does not load ff_mpeg

  25 mai 2016, par Edv Beq

  I have been trying to use a simple code to play videos with Kivy and Python. The shortest code I found is the one shown below.

  1. At first, Kivy was showing an error on Pyglet and ff_mpeg. Therefore, I installed Pyglet - which consequently require AVbin, - none of which would load automatically.

  2. I found the following solution on another post - which fixed the sound :

     import pyglet
     
     pyglet.lib.load_library('avbin')
     
     pyglet.have_avbin=True

  3. Can I do the same with ff_mpeg ? Can it be loaded manually ?

  4. I have already installed ffmpeg :

     • When I type ffmpeg -version on cmd - it displays ffmpeg version N-.... built with gcc 4.9.2(GCC). Also, I have added all the paths in WIndows.

  5. Also my pip list :

     Cython (0.21.2)
     
     docutils (0.12)
     
     ffmpegwrapper (0.1.dev0)
     
     Kivy (1.8.0)
     
     Kivy-Garden (0.1.1)
     
     Pillow (2.1.0)
     
     pip (6.0.8)
     
     pygame (1.9.2a0)
     
     pyglet (1.2.1)
     
     requests (2.5.1)
     
     setuptools (12.0.5)

  6. Error on Kivy :

     [DEBUG      ] [Video     ] Ignored <ffmpeg> (import error)
     
     [INFO       ] [Video     ] Provider: pyglet(['video_ffmpeg'] ignored)
     
     </ffmpeg>

     and

     File "C:\Python33\lib\site-packages\kivy\core\video\video_pyglet.py", line   67, in _update
     
        self._player.dispatch_events(dt)
     
     AttributeError: 'Player' object has no attribute 'dispatch_events'

  7. And, finally here is the code - I am working with :

     import kivy
     
     import pyglet
     
     pyglet.lib.load_library('avbin')
     
     pyglet.have_avbin=True
     
     
     
     kivy.require('1.8.0')
     
     
     
     from sys import argv
     
     from os.path import dirname, join
     
     from kivy.app import App
     
     from kivy.uix.videoplayer import VideoPlayer
     
     
     
     class VideoPlayerApp(App):
     
     
     
         def build(self):
     
             if len(argv) > 1:
     
                 filename = argv[1]
     
             else:
     
                 curdir = dirname(__file__)
     
                 filename = join(curdir, 'project.mp4')
     
             return VideoPlayer(source=filename, state='play')
     
     
     
     if __name__ == '__main__':
     
         VideoPlayerApp().run()

  I have searched a lot of other threads, and installed ffmpeg several times with no luck. Any help would be greatly appreciated. Thank you !

• I want to use my AMD GPU in Python FFMPEG [duplicate]

  1er août 2023, par Filipe Braga

  I'm using this code to take videos that range from 15 to 30 seconds long, and looping them to a target duration of 1:05 :

  


  import os
import glob
import torch
from moviepy.editor import VideoFileClip, concatenate_videoclips

def estender_duracao_gpu(video_path, duracao_desejada):
    video = VideoFileClip(video_path)
    duracao_atual = video.duration

    if duracao_atual < duracao_desejada:
        repeticoes = int(duracao_desejada / duracao_atual)
        video_clips = [video] * repeticoes
        video_clips.append(video.subclip(0, duracao_desejada % duracao_atual))

        # Concatena os clips utilizando a GPU
        video_final = concatenate_videoclips(video_clips, method="compose")

        return video_final
    else:
        return None

def estender_videos_em_massa(pasta_origem, pasta_destino, duracao_desejada):
    if not os.path.exists(pasta_destino):
        os.makedirs(pasta_destino)

    videos = glob.glob(os.path.join(pasta_origem, "*.mp4"))

    for video_path in videos:
        nome_video_sem_extensao = os.path.splitext(os.path.basename(video_path))[0]
        
        try:
            video_estendido = estender_duracao_gpu(video_path, duracao_desejada)

            if video_estendido is not None:
                nome_video_final = os.path.join(pasta_destino, f"{nome_video_sem_extensao}_estendido.mp4")
                video_estendido.write_videofile(nome_video_final, codec="libx264", write_logfile=False, threads=torch.cuda.device_count())
                print(f"Duração do vídeo {nome_video_sem_extensao} estendida com sucesso.")
            else:
                print(f"O vídeo {nome_video_sem_extensao} já possui a duração desejada ou é mais longo.")
        
        except Exception as e:
            print(f"Ocorreu um erro ao estender o vídeo {nome_video_sem_extensao}: {str(e)}")
            continue

# Example of use:
pasta_origem = "pasta_com_os_videos"  # Replace with the path of the folder containing the input videos
pasta_destino = "pasta_com_videos_estendidos"  # Replace with the path of the folder to save the extended videos
duracao_desejada = 65  # 65 seconds (1 minute and 5 seconds)

estender_videos_em_massa(pasta_origem, pasta_destino, duracao_desejada)


  


  The code is working, but it does not use my AMD GPU (RX 5700XT). How can I make it do so ?