'utf-8' codec can't decode byte 0xc1 in position 2079: invalid start byte

[ChristianMalazarte]

I was getting this error on a few of my files. So I did some digging and read that if you add 'ignore' onto the decode part, it fixes this problem: Zulko/moviepy#959

So if anyone comes across this problem, just change err.decode('utf-8') to err.decode('utf-8', 'ignore') on both instances. I'm still not sure if this breaks anything, but apparently, it just tells it to ignore the extra byte.

    if proc.returncode != 0:
        err_lines = err.decode('utf-8', 'ignore').split('\n')[-5:]
        logger.error(err_lines)
        raise Exception('Problem trying to ffmpeg images for {}'.format(video_file))

    # Speed
    end = time.time()
    seconds = round(end - start, 1)
    speed = re.findall('speed= ?([0-9]+\.?[0-9]*|\.[0-9]+)x', err.decode('utf-8', 'ignore'))
    if speed:
        speed = speed[-1]
    logger.info('Generated Video Preview for {} HW={} TIME={}seconds SPEED={}x '.format(video_file, hw, seconds, speed))

[stevezau]

@ChristianMalazarte do you mind openig a PR so i can merge it in?