Broken accents, replacement boxes or sequences of strange characters can indicate a text-encoding or font-rendering problem. The visible symptom matters because those two stages need different checks.
Keep the original subtitle file unchanged while identifying where readable text becomes damaged.
Distinguish decoded text from missing glyphs
An encoding tells software how bytes represent characters. If the wrong interpretation is used, an accented name can turn into an unexpected sequence. A font problem can instead leave the intended character without a suitable visible glyph.
A box or question mark alone does not prove one cause. Inspect a non-private example in a trusted editor or another supported text view if available. Record whether the underlying characters are already wrong there or only become wrong during playback.
Changing the interface language does not necessarily change the subtitle file’s encoding or the player’s font coverage. Keep those settings separate. A source can also contain an actual spelling error that no encoding choice will repair.
Use a short representative sample
Choose a line containing the affected accent, script or punctuation. Keep its cue time and intended wording if known. Compare it with ordinary letters in the same file. This is more useful than declaring every subtitle unreadable.
For a file you are entitled to edit, save any encoding conversion as a separate copy through a trusted workflow. Repeatedly opening and resaving damaged text can preserve the wrong interpretation. The original bytes are valuable for diagnosis.
| Symptom | Question |
|---|---|
| Unexpected character sequence | Were bytes decoded with the wrong encoding? |
| Empty boxes for one script | Does the rendering font cover those characters? |
| Correct editor text, wrong playback | What does the player interpret or render? |
| Same misspelling everywhere | Is the source text itself wrong? |
Verify more than one line after a change
Check names, punctuation, non-Latin text and line breaks near the beginning, middle and end. A conversion that fixes one accent may still mishandle another part of the file. Keep timing unchanged unless that is a separate authorized task.
If the subtitle is supplied directly by a service, report the title, language, cue time and device rather than attempting to rewrite the source. Remove account details from captures. A precise example gives support enough information without sharing an entire subtitle track unnecessarily.
- Preserve the original file or source identity.
- Record one representative damaged line.
- Compare text interpretation and rendering separately.
- Use a trusted conversion only for media you may edit.
- Verify several character types after any correction.
Useful answers
Does changing menu language fix subtitle encoding?
Not necessarily. Interface language, subtitle encoding and font coverage are separate.
Should I overwrite the original after the first conversion?
Keep the original until the result is verified. Its bytes help identify whether the interpretation was correct.