when called split_video_ffmpeg throw an error

[pengyouak]

Some videos will throw this error halfway through the process

bool' object has no attribute 'shape'

[Breakthrough]

Can you provide more context? A backtrace, etc? What OS, Python, and package versions you're using? Have you tried updating OpenCV or using a different backend?

Run scenedetect version and report that. Can you also share more about your workflow?