When I run this it is showing ffmpeg has no attribute input. How to solve this?

Question

def extract_audio_from_video(file_path: str) -> np.ndarray:
    inputfile = ffmpeg.input(file_path)
    out = inputfile.output('-', format='f32le', acodec='pcm_f32le', ac=1, ar='44100')
    raw = out.run(capture_stdout=True)
    del inputfile, out
    return np.frombuffer(raw[0],np.float32)

Answer 1

If you'd like to explore a different option, check out [my ffmpegio package:

pip install ffmpegio

import ffmpegio

fs,x = ffmpegio.audio.read(file_path, sample_fmt='f32le', ac=1, ar=44100)
# fs - actual sampling rate
# x - [nx1] numpy array