SubLunioLast updated:
Treat every episode as its own timed source
A series gives you recurring characters and settings, but each episode still has a different timeline. Recaps, opening titles, advertisements and end-credit scenes can vary between releases. A subtitle file for episode two should not be reused for episode three just because their running times look similar.
Begin with the exact episode you plan to watch. The local audio extraction follows that file’s soundtrack, while the video remains on your device. Choose one representative episode in the browser to evaluate the workflow before preparing a whole season. This makes it easier to identify naming or soundtrack problems before they affect a large queue.
Keep character and place names consistent
Recurring names are a useful reason to prepare a small glossary. Include the original term and your preferred translation, one entry per line. Focus on names that appear in the audio, especially where several spellings are plausible. A compact list is easier to understand and maintain than a large collection of unrelated background facts.
Watch how the terms appear in the preview. A glossary guides translation; it does not correct every word the speech recognizer mishears. When a character’s name changes spelling, compare the original transcription with the translation to find which stage needs attention. Keep your preferred terminology with the project so later episodes use the same starting point.
Check soundtrack order across the season
Files from a single release often share a track layout, but exceptions are possible. A special episode might place a commentary first, and a bonus video might contain a different language. Before using shared settings across a batch, inspect the tracks and listen to a sample from any unusual file.
Source language and subtitle language are separate choices. Leave source on Auto when appropriate, or specify the language you know is being spoken. Choose the reading language from 110+ targets. If a series switches languages inside an episode, check those scenes separately instead of assuming the dominant-language setting resolves every exchange.
Use a queue without losing episode identity
The desktop workflow can keep multiple local episodes in a persistent queue. Give the source files meaningful names before adding them, such as a series name followed by season and episode numbers. Output language suffixes then help distinguish translations without erasing the episode identity.
Do not move or rename source files while they still need extraction. The queue can retain upload progress, but it cannot read a file that has disappeared from the recorded location. A stable folder and enough temporary disk space make unattended processing more predictable. Review any item marked as needing attention instead of starting a duplicate of the entire batch.
Understand duration-based costs for a collection
Each complete task is priced from its own full duration at USD $1 per hour, rounded up to a cent. Two short episodes are separate tasks, so their rounding happens separately. An opening or credits section is still part of the source duration, even if it contains little speech. The free preview is for evaluation and is not deducted from the full-video price.
If you create more than one target language for an episode, each translation is a separate task. Switching between original, translated and bilingual variants within a completed task is an output choice, as is switching between SRT, VTT and ASS. Check the queue’s task count and prices before committing a whole collection.
Review across episodes before sharing results
Sample more than the first episode. Look for a fast conversation, a quiet scene, a recap and a proper name introduced later in the season. Confirm that the subtitle filename matches the actual episode and edition. This catches organizational mistakes that a language-quality check alone would miss.
Keep downloaded files and any manual corrections in your own project folder. Account records make it easier to return to completed subtitles, but they do not store your original series videos. Use and share subtitles only where you have the necessary permissions, and review automatic results for the audience and context in which they will appear.
Review each episode before batch output
Review subtitles and the playback audio track for each episode in the desktop player, then select completed tasks to export SRT, VTT or ASS subtitle files in batches or burn subtitles into the matching videos. Keep filenames clear across episodes, languages and video editions.
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
Should a whole season be joined into one video before generating subtitles?
Usually no. Separate episodes are easier to name, retry, price and match to their own timelines.
Can I reuse a glossary for recurring series characters?
Yes. Use the same concise term list for each relevant task, then check the actual output for consistency.
Why might one episode need a different audio-track setting?
Specials, alternate releases and commentary editions can have a different track order from the rest of the season.
Are separate target languages charged as separate episode tasks?
Yes. Each episode and target-language combination is its own generation task.