Subtitle Templates: One Timed Master File for Every Language

What a subtitle template is

In subtitling, timing a file (also called spotting) is a separate skill from translating it. A template splits the two. One person times and segments the source-language subtitles carefully against the final cut, the file is checked and locked, and translators then work cue by cue, writing their language into slots that are already timed.

Templates are common in film and series localization, where one master feeds many languages, but the idea scales down. A company releasing a training video in six languages, or a creator sending a talk to three volunteer translators, benefits in the same way. The template is usually an SRT or another timed format, sometimes accompanied by notes for translators.

A template is not a pivot. It defines timing, and translators may still translate from the source language directly. Routing a translation through an intermediate language is a different practice with its own risks, explained in pivot language translation.

Why teams translate from a template

• Timing is done once. Spotting a file well takes time and care; with a template, each translator skips it.
• Versions stay aligned. Every language has the same cue count and boundaries, which simplifies side-by-side review and bilingual displays.
• QC is faster. Timing checks done on the master largely carry over, so per-language QC can focus on text and reading speed.
• Changes propagate cleanly. When a cue's timing changes, you know exactly which cue to update in every language.
• Translators focus on language. Freelancers without timing software or experience can still deliver usable files.
• Consistent annotations. Notes on names, jokes and on-screen text reach every translator in the same form.

What goes into a good template

A template multiplies its own flaws across every language, so it deserves more care than a single-language file.

• Final picture. Build it from the locked cut; any later edit shifts times in every version.
• Accurate source text. Misheard words become mistranslations in every language.
• Sensible timing. Cues start with speech, end soon after it, respect shot changes where possible, and leave small gaps between consecutive cues. The rules professionals use are covered in subtitle timing guidelines.
• Generous segmentation. Break cues at sense boundaries and avoid packing too much into one cue, since many languages need more characters than the source.
• Annotations. Notes for translators on proper names, idioms, wordplay, cultural references, deliberately untranslated speech, and on-screen text that needs a cue.
• A matching style guide and glossary. Translators need the same terminology and conventions; creating a translation style guide shows how to write one.

How translators work from a template

The rules for what a translator may change should be written down before work starts. Common arrangements:

Strict

Text only. Every cue keeps its timing and the cue count stays fixed. Simplest to manage, hardest on languages that run long

Merge allowed

Translators may merge two adjacent cues when the target language reads better as one, keeping the outer start and end times

Limited retiming

Small adjustments to in and out times are allowed within stated limits, for example to keep reading speed acceptable

Free

Template used for reference only; translators retime as needed. Flexible, but versions no longer align

Translators usually work in a subtitle editor that shows the source text alongside an empty or editable target column, so they overwrite text without touching timestamps by accident. If a translator works in a plain text editor instead, the editing advice in how to edit an SRT file applies.

Building and distributing a template: step by step

1. Lock the picture and note its version, duration and frame rate.
2. Produce the source-language subtitle file and correct the text against the audio.
3. Time and segment it to your rules, leaving room for longer languages.
4. Add annotations for anything a translator might misread, either as a separate notes document keyed to cue numbers or as comments in a format that supports them.
5. Run a full QC pass on the template, as if it were going to publication.
6. Freeze it with a version number and send it with the style guide, glossary, a viewing copy of the video and the written rules on merging and retiming.
7. QC each returned language for text, reading speed and file validity.

Limits and trade-offs of template-based translation

• Text expansion. Many translations run longer than the source, and a cue timed for short English may be too fast to read in another language. The effect is explained in text expansion in translation.
• Word order. A cue break that suits English can force an awkward split in languages that put the verb at the end or order clauses differently.
• Errors multiply. A misheard name or a badly placed break in the template appears in every language unless translators catch it.
• Late picture changes. Each edit to the cut means updating every language file, and the more languages, the more costly late changes become.
• Not a substitute for a subtitler's judgement. Strict templates can push translators into literal, cramped lines; allowing merges or limited retiming often produces better subtitles at the cost of perfect alignment.
• Dubbing scripts are different. A template designed for reading speed is not a dubbing script, which follows mouth movements and spoken length.

Change control once translations are out

Treat the template like source code. Keep it under a version number, log every change with the cue numbers affected, and send translators the changes rather than a whole new file when possible. A simple table with cue number, old text, new text and reason is usually enough. Without this, it becomes impossible to say which language files reflect which version of the video.

Using a mydubly source-language SRT as a template

A mydubly transcript job in the spoken language gives you a timed SRT and VTT, numbered in order, with one single-line cue per recognized speech segment, plus a plain and a timestamped transcript. That is a useful first draft for a template: the timing comes from the audio, and the text is a machine transcript you can correct. A 30-minute video costs 30 credits (3¢) for the transcript job, which includes the subtitle files.

It is a draft, not a finished template. Cue boundaries follow recognized segments rather than sense units, cues are single lines that may be long, and there are no annotations, italics or speaker labels. Open the SRT in a subtitle editor, correct the text, merge or split cues for sense and reading speed, apply your timing rules, add translator notes, then run QC before distributing it to human translators.

Two limits matter for planning. mydubly cannot import an SRT, so once you have edited the template it cannot be fed back into mydubly for translation. And a translated mydubly job works from the audio again: its target-language cues follow recognized segments and will not reflect your template edits, so treat its output as a separate machine draft rather than a translation of your template. The subtitle generator and the SRT generator page show what the files contain, and video localization covers the wider workflow around multiple languages.

Next step: build one template before you need ten

Choose your next video that will go into more than one language, build and QC a source-language template from it, and write down the merge and retiming rules in a few lines. Sending that package to the first translator will show you what to change before the approach scales to every language.