Programm ends with an error, but video is created

[MBadberg]

After the video is created, following error occurs:

Traceback (most recent call last): File "D:\conda\DashcamCleaner\dashcamcleaner\src\blurrer.py", line 160, in run subprocess.run( File "D:\Users\marku\miniconda3\envs\py38\lib\subprocess.py", line 493, in run with Popen(*popenargs, **kwargs) as process: File "D:\Users\marku\miniconda3\envs\py38\lib\subprocess.py", line 858, in __init__ self._execute_child(args, executable, preexec_fn, close_fds, File "D:\Users\marku\miniconda3\envs\py38\lib\subprocess.py", line 1251, in _execute_child args = list2cmdline(args) File "D:\Users\marku\miniconda3\envs\py38\lib\subprocess.py", line 553, in list2cmdline for arg in map(os.fsdecode, seq): File "D:\Users\marku\miniconda3\envs\py38\lib\os.py", line 818, in fsdecode filename = fspath(filename) # Does type-checking offilename. TypeError: expected str, bytes or os.PathLike object, not NoneType

But the video is fine

[tfaehse]

Hi!

This is just an assumption, but I think your ffmpeg-binary isn't found. The tool expects to find the path to the executable in the "FFMPEG_BINARY" path, and doesn't do any error-checking whatsoever.

This is, in general, a fairly ugly solution. I'll try to keep it short: I use OpenCV to blur objects, but it doesn't support audio tracks. So the resulting video has no sound. My current (naive) solution is to use OpenCV to create a _temp file without audio and then use ffmpeg to combine the unedited video's audio with the blurred video. Once that's done, the (muted) OpenCV result is deleted. There are a few things about this that could lead to errors, in your case it just seems to be a missing path variable for the ffmpeg executable.

It will unfortunately take a bit, but I'll improve error handling so that (hopefully) error messages actually make sense and don't crash the program.

As for your original comment: the quality is absolutely something I should work on! My own dashcam videos were already heavily compressed from the camera so I never noticed how bad it actually could get. OpenCV doesn't really allow you to control the bitrate/compression, but I found some potential workarounds. I can't really promise any timeline here, but I'll get to it sooner or later and post an update here!

[tfaehse]

@MBadberg The newest commits should address the issues, at least somewhat.

• Error messages regarding the FFMPEG_BINARY location should be expressive enough to at least show the problem
• I switched from OpenCV's Videowriter to imageio, you get direct control over ffmpeg's quality setting there. The file size varies a great deal between quality setting 0 (lowest) and 10 (highest), but since my input videos are already heavily compressed I don't notice much of a difference.