Jumpcutting 12 hour video "jumpcutter.py in line 108"...

Hi, So I tried this on a 12-hour video, and it failed on audio. It throws me an error saying:

Traceback (most recent call last):
  File "jumpcutter.py", line 108, in <module>
    sampleRate, audioData = wavfile.read(TEMP_FOLDER+"/audio.wav")
  File "C:\Python37\lib\site-packages\scipy\io\wavfile.py", line 264, in read
    is_big_endian, mmap)
  File "C:\Python37\lib\site-packages\scipy\io\wavfile.py", line 129, in _read_data_chunk
    data = numpy.frombuffer(fid.read(size), dtype=dtype)
ValueError: buffer size must be a multiple of element size.

Here is the full log: https://pastebin.com/5Re462TW. Now, I still have the cmd open and TEMP folder with ALL the 1.8 million frames. I need to edit your code, to just start extracting the audio, and continue. Can somebody help me with that? I started learning python only this year at my school.

[torsrex]

May I suggest splitting the video up into 1h cunks, running the script on those 1h chunks and then combining them together again?

Splitting can be done using:

ffmpeg -i input.mp4 -c copy -map 0 -segment_time 01:00:00 -f segment output%03d.mp4

Then run the jumpcutter script on each individual file. Then move the files to a separate folder. Run this

(for %i in (*.mp4) do @echo file '%i') > mylist.txt

Which will create a list of all the files. You might want to check the mylist.txt file to ensure the order is correct. Then run

ffmpeg -f concat -safe 0 -i mylist.txt -c copy output.mp4

to combine the files back together, and you should be good!

[kubo6472]

Hi, thanks for everything. The combination of re-encoding thru HandBrake, and obtaining a new jumpcutter.py along with all the extensions, pip, and python updates made it eventually working. I am so happy this project has got this active community.