In Quven, translating a subtitle is done from the player, and the decision that matters is which source track you translate from. That choice decides whether the timing survives, whether a signs-only track gets translated as if it were the script, and whether the result is a subtitle you’ll want to use again. This guide covers how to pick the source, what happens with image-based subtitles from discs, what the fallback is when a file has no subtitle at all, and who on a shared server is allowed to ask for any of it.
Start with the release and account you are using
Open a title, start playback and open the subtitle controls. The translation action appears on a Max account and on a Lifetime one, which carries the same capabilities without a renewal, and the server has to be on Quven 1.1.0 or later. Translation is the one step in this pipeline that leaves your machine: it sends derived text through Quven Cloud, so it asks for your consent and it’s metered, while everything around it stays local.
The allowance is weekly, and it belongs to the account. Two servers linked to the same Quven Account will draw on one budget, and when that budget is gone the player tells you in as many words: “This week’s AI subtitle translation budget is spent. The allowance is shared across every server linked to your Quven Account and renews weekly.” A budget the server couldn’t read counts as available, which is deliberate, so an unreachable Cloud won’t grey the action out and you should hear about it from the request itself.
What if the action isn’t there? There’s no server-side switch that forces it on, so don’t go looking for one. You should check that the server and the native client are both on a current compatible release, that the client is signed in to the Quven Account you mean to use, and that you accepted the text-transfer consent when Quven asked for it. If all three are right and the action still doesn’t appear, that account or that release doesn’t carry it, and nothing in the media-library settings will change that.
Who can ask for it on a shared server
Subtitle work is held to the person who owns the server. A guest profile opens the same panel and finds a line where the buttons belong: “Only the person who owns this server can create subtitles for it.” Recognition and translation both sit behind that boundary, and so does the subtitle editor, because the job would spend the owner’s hardware and the owner’s metered allowance.
Watching is a different matter. A guest can see the queue, so a job somebody else started is visible while it runs, and once the track lands it’s on the title for everyone who opens it. If you’re the guest here and you want a subtitle translated, you should ask whoever runs the server, since there’s nothing on your side to switch on first.
Pick the subtitle that contains the film
Choose Translate subtitles in the player and set the target language. Above that sits a Subtitle source row, which opens on Automatic (best available): Quven picks the original it rates highest and will read the audio only when nothing usable is there. Open the row and the tracks themselves appear, each one labelled with its language and with what it’s made of, so you might read text, or image, needs recognition, or sidecar file, or generated by Quven, or edited by you. That second half of the label is the part you should look at, because it tells you whether the translation starts from a sentence somebody wrote or from a picture the server has to recognise first.
A full dialogue subtitle in an ordinary text format is the best starting point you can have. It already carries the cue timing, the line breaks and the editorial judgement about what to leave out, which speech recognition can only approximate.
The language label on its own tells you very little. A forced track may contain foreign dialogue and on-screen signs and nothing more. A signs-and-songs track is a handful of cues across two hours. Commentary captions and duplicate streams look just like the real thing in a menu. When the file has a genuine dialogue track, choose that one. Quven will write the translation as a new timed track and leave the original where it was.
What happens to PGS, VobSub and DVB subtitles
Blu-ray and DVD subtitles sit in the menu looking like any other track, but a PGS, VobSub or DVB track is made of pictures. Each cue is a bitmap with display timing attached, which is why a player draws them without complaint and a text extractor comes back with nothing. There is no sentence in the file to read.
Quven turns those images into timed text on your own hardware before any translation happens. It resolves the subtitle stream and its cue timings, drops the transparent state frames, recognises the characters, and writes a valid WebVTT source. The translated result then joins the same player workflow an authored text track uses.
This needs Quven 1.1.0 or later. An earlier release won’t have the path, however the track is labelled. You don’t install an OCR engine and you don’t point Quven at a folder of extracted images; supported builds carry the recognition path themselves.
When the file has no usable subtitle track
If there’s no eligible text track, no usable bitmap output and no sidecar file, Quven can fall back to local speech-to-text for the media file you selected. The route is conservative on purpose. It takes the audio stream you chose, derives timestamp-preserving PCM on the home server, recognises it there, and normalises the transcript into WebVTT before any optional translation.
Treat that as the last resort. An authored subtitle will almost always read better, knows who is speaking, and breaks its lines where a person decided they should break. This needs 1.1.0 or later as well, so an older installed server won’t offer it, whatever this guide says about the workflow.
Let the job finish, then keep the useful track
Translation produces a real timed asset, and it will still be there when you reopen the title, whenever that might be. The jobs behind it recover after a restart, merge duplicate requests into one piece of work, and honour cancellation, so a late result can’t turn up half an hour after you changed your mind.
On a long film, watch the progress the player reports; starting the same request again won’t make it faster. When the translated track appears, select it as you would any subtitle. Keep the original enabled only if you genuinely want both on screen, because two full dialogue tracks at once will make a good translation look worse than it is.
Keep the privacy boundary in view
OCR and speech recognition both run on the self-hosted server. Temporary audio, decoded frames and recogniser output stay on that machine. The optional Cloud step receives derived subtitle text only once the consent and account checks have passed, and that is why this is a metered feature and not a button that quietly steps around the rest of Quven’s privacy model.
A title that plays directly on one client may well take a different route on another, and the subtitle outcome changes with it. The native playback guide explains why the container, the client and the kind of subtitle still matter. For a library that matches cleanly before playback is even a question, read how to organise films and TV shows.
Frequently asked questions
What does translating a subtitle cost?
It needs a Max or Lifetime account and a server on Quven 1.1.0 or later. The allowance is weekly, and one account shares it across every server linked to it. Only the finished text will be sent for translation; your audio and your files stay where they are.
Can it translate the subtitles on a Blu-ray?
Yes, though not in one step. PGS and VobSub tracks are pictures, so the server reads them into text first, on your own hardware, and the translation starts from that text.
Can a guest on my server translate a subtitle?
No. Creating a subtitle belongs to the person who owns the server, and so does editing one. A guest reads “Only the person who owns this server can create subtitles for it.” where the buttons would be, and they can still watch the queue and use the finished track.
The translation action is not there. What is wrong?
Check the release and the account before anything else: server and native client on a current compatible release, the client signed in to the Quven Account you mean to use, and the text-transfer consent accepted when Quven asks. Changing media-library settings won’t move that boundary.