SubLunioLast updated:
Begin with the voice track you want to understand
An anime file may contain Japanese dialogue, a localized dub and extra commentary. These tracks can differ in phrasing, jokes and character names. Choose the track you actually intend to hear, then listen briefly before beginning extraction. A file labelled with a language does not guarantee that its first audio track uses it.
SubLunio works from sound. The browser extracts compressed audio locally and uploads that audio for recognition and translation. The video picture stays on your device. Text printed on signs, title cards or a character’s phone is not transcribed simply because it appears in the frame; it needs a separate visual-text workflow if it matters to your viewing.
Use a glossary for names and invented vocabulary
Fantasy terms, attack names and unfamiliar character names are common sources of inconsistent wording. A short glossary can express the spellings or translations you prefer. Keep entries concrete, and avoid conflicting instructions for the same source word. An existing official localization may help you choose terminology when you have permission to use it as a reference.
The glossary guides translation after speech recognition. If the recognizer hears a name as an ordinary word, the preferred entry may not match. Compare the original line with the translated one in the preview to understand the problem. Important names deserve a second look near their first occurrence and when a different character says them.
Evaluate quick exchanges in context
Short exclamations, overlapping dialogue and music under speech can make an otherwise clear scene difficult. The preview lets you check a continuous section rather than judging isolated text. It begins at the first subtitle and covers up to 5 minutes, or the remaining video when it is shorter.
Watch whether each cue stays long enough to read and whether a joke remains understandable in the target language. A literal translation of a wordplay may not convey its purpose. Automatic output can provide a useful starting point, but creative dialogue often benefits from editorial judgment before it is shared with other viewers.
Choose bilingual output for closer comparison
Bilingual subtitles put the recognized source and the translation in the same timed cue. This can help with listening practice or checking a line without repeatedly switching tracks. Enable bilingual generation before starting if that is your plan, so the dialogue is split with the extra reading load in mind.
For compatible desktop players, ASS gives the two languages distinct visual treatment. SRT keeps the structure simpler and is useful when device compatibility matters more than styling. You can also download separate original and translated files. Two independent files are useful for language selection, but showing both at once depends on the player’s capabilities.
Plan a series workflow without confusing versions
Choose the target language from 110+ subtitle languages; the spoken-language menu has a separate recognition list and defaults to Auto. If the audio is consistently Japanese, selecting Japanese explicitly can remove ambiguity. A dub should instead use the language spoken in that dub.
For several episodes, keep episode numbers in the filenames and inspect exceptions such as specials or alternate cuts. The desktop queue can organize repeated extraction and uploading. Do not assume every file has the same soundtrack order. A quiet first episode also does not prove that later, faster dialogue will have the same recognition quality.
Continue, download and make a final viewing check
You can try the web preview as a guest. Sign in to retain the task and continue the full video using the audio already uploaded. Full generation costs USD $1 per hour of the whole video, rounded up to a cent; changing an available output format does not create another charge for that task.
After downloading, check the subtitle against the exact episode edition and inspect a few later scenes. Confirm character names, timing and line wrapping on the screen you will use. These are dialogue subtitles, not a guarantee of complete sign translation or accessibility sound descriptions. Processing or distributing an anime file also requires the appropriate rights; generated subtitles do not supply those rights.
Check bilingual layout before burn-in
On desktop, export separate subtitle files or burn subtitles into the video. Before burn-in, check bilingual line order, long-line wrapping, font coverage and safe margins in the player so two lines do not obscure important parts of the picture.
Desktop checks each file before offering video export methods. Successful subtitle generation alone does not guarantee burn-in or embedded-track support.
A few things worth knowing
Will anime subtitles include Japanese text visible on signs?
This workflow transcribes audio, not on-screen writing. Signs that are not spoken need separate transcription or translation.
Can I keep a preferred spelling for an anime character name?
Add it to a concise glossary before generation, then check both the recognized source and translation.
Should I choose Japanese when listening to an English anime dub?
Choose the language actually spoken in the selected dub, or use Auto.
Which output helps compare anime dialogue with its translation?
Bilingual output places both in one timed cue; ASS offers additional styling in compatible desktop players.