AI narration has a flaw you can't consciously name — a word loses its ending, an S goes dull or too hot.

The exact clicks

1. Open a terminal (a text window where you type a command and press enter) in the folder that has your narration file and voice_doctor.py.
2. Type: python3 voice_doctor.py your-narration-file.wav — and press enter.
3. Read the report it prints: pacing, rate, sibilance (how harsh the S sounds are), and word-ending strength, each measured against that same file's own average line, not a fixed outside standard.
4. If a line is flagged, run it again with the repair option: python3 voice_doctor.py your-narration-file.wav --repair fixed-file.wav — this levels just that line to match the rest.
5. Run the same check again on the repaired file, to confirm the flag is actually gone, not just quieter.
6. Only ship once a run comes back with nothing left flagged.

What changes

• A dulled word ending or a too-hot S is something most listeners can't name, but hear as "something's a little off" — measuring it directly catches what conscious listening misses.
• The repair step checks its own output again rather than assuming one pass fixed it.

What it will not do

• This measures against the narration's own median, not a fixed external standard — a whole file recorded badly could pass its own bad baseline.
• It catches what it's built to measure — pacing, rate, sibilance, endings. It is not a general "does this sound good" check.