I just tried exporting using the Export: “Subtitles” option under Netflix, and noticed some issues.
- Header column order. For Japanese, it appears the column headings are mismatched, with the Romaji and Hiragana headers switched:
|18s|しかし…|However…|しかし …|shikashi …|
|20s|彼らは全員 偽名|They only know each other
by their false names.|かれら は ぜんいん ぎめい|karera ha zen’in gimei|
Time resolution. It appears the time stamp is rounded to seconds, and an ‘s’ is appended. Why? The rounding kind breaks use of the file for loading the content into an external tool that wants to match the subtitles with the audio.
Hiragana spaces. It appears the hiragana column (heading Romaji) has spaces between parsed words. Not necessarily a bug, and this may be useful for some people, but some, like me, might prefer them raw, as I do my own automated word parsing. Easy enough to filter out, but not without breaking the language generality.