Subtitles
June 19, 2026
What Is Subtitle Translation? A Complete Guide to Tools, Formats, and Best Practices
Subtitle translation is the process of converting a video's spoken dialogue and on-screen text into readable subtitles in a target language, timed to the original audio. It's how creators break language barriers and reach a global audience without recording a single extra minute — and the right tool depends on whether you're subtitling a TikTok clip, a corporate training video, or a regulated broadcast. This guide covers the formats, the methods, the real costs, and the rules most tutorials skip.
Key Takeaways
- Subtitle translation converts spoken dialogue and on-screen text from a source language into target-language subtitles, timed to the audio
- Three subtitling methods dominate: machine translation (70–85% accuracy, free), AI/neural (90%+, cheap, fast), and professional human subtitling (~99%, highest cost)
- Respect the on-screen format: 42 chars/line, two lines max, 1–6s display time — these limits aren't optional
- Choose SRT as the safest default format; VTT for streaming/HTML5; SSA for stylized anime-style work
- Free tools work for personal and internal video; paid is worth it for any brand- or compliance-facing process
What Is Subtitle Translation?
At its core, subtitle translation is the process of converting the original spoken dialogue and any relevant on-screen text from a source language into target-language subtitles that appear on screen. The translated subtitles have to do three jobs at once: stay faithful to the meaning, fit into a narrow character budget, and line up with the audio so viewers can read without losing track of the video. Good subtitling requires all three; drop one, and the audience notices immediately.
Subtitles differ from captions, though tools and teams mix them up constantly. Subtitles assume the viewer hears the audio and just needs the words in another language. Captions — and especially SDH (Subtitles for the Deaf and Hard-of-Hearing) — include non-speech cues like [laughter], [door slams], or speaker identification. Three types dominate production: open (burned into the video, can't be turned off), closed (toggled by the viewer), and SDH (closed captions with audio descriptions for accessibility).
A Verizon Media and Publicis Media study found 80% of people say they're more likely to finish a video with captions, and about half watch with sound off on mobile and desktop. Subtitles aren't a nice-to-have — they're how a large share of your audience consumes video content in the first place, and translating them into multiple target languages makes the same video accessible to audiences you otherwise couldn't reach. The experience of watching a subtitled video feels native when the subtitling is done right.
Subtitle Translation vs. Dubbing vs. Voiceover
If you're choosing between subtitling, dubbing, and voiceover, the decision usually comes down to budget, emotion, and audience expectation — not quality. Many teams choose subtitling first because the cost is lower and the subtitling workflow is faster to set up across languages. Here's how they stack up:
| Aspect | Subtitle Translation | AI / Pro Dubbing | Voiceover |
|---|---|---|---|
| Relative cost | Lowest | Mid | Highest |
| Turnaround | Hours (AI) to 1–2 days (human) | Hours (AI) to 1–2 weeks (human) | 1–2 weeks |
| Viewer experience | Reading while watching, on-screen text experience | Native-language audio experience, lip sync possible | Native voice over original audio |
| Best for | Creators, social, low budget, accessibility | Brand content, film, corporate training | Documentaries, explainers, news |
| Accessibility value | Highest (deaf/HoH viewers) | Low | Low |
Subtitles win on cost and speed; dubbing wins on emotional engagement and watch time in audio-heavy markets (Germany, France, Italy). Voiceover still has a niche for documentaries and news where you need a human voice actor's delivery without the expense of full dubbing. Most teams translating video for a global audience choose subtitling first and layer dubbing on top for the top two or three target languages where watch-time data justifies the cost. For a deeper breakdown, see our comparison of AI dubbing vs. subtitles.
How Subtitle Translation Works — The 5-Step Process
The process follows five steps regardless of method. Skip one and you end up with subtitles that flash by, overrun the screen, or drift out of sync with the audio. Each step has its own failure mode.
Transcription
The original audio from your original video is converted to text. Background noise can drop speech recognition accuracy by 15–30%, which means every mistake compounds through the rest of the process. Good tools let you review the transcript before moving on.
Translation
The transcript is translated into the target language. Most creators outsource this step to Google Translate, DeepL, or full translation services, and it's where context loss happens most often — translators working from an SRT file alone miss the visual context, tone, irony, and on-screen cues that change meaning.
Timing
Each subtitle needs a start and end timecode accurate to ±200ms. Mistimed subs break comprehension before the viewer even registers it, which is why automated timing from speech recognition is such a leap over manual SRT editing.
Formatting for screen
Professional standards: maximum 42 characters per line, no more than two lines, display time between 1 and 6 seconds, reading speed around three words per second (roughly 15–20 characters per second). Break these and translated subtitles either flash by or get cut off on mobile.
Human review
Even with 90%+ neural MT accuracy, branded or public-facing content gets a human pass. Reviewing the translated text matters most where cultural nuance, legal language, or brand voice are in play. Human reviewers catch what neural MT misses: intent, cultural nuance, brand voice.
A step-by-step walkthrough of the full process: how to translate subtitles.
Methods: Machine, AI, and Professional Subtitle Translation
Four options, and they're not interchangeable. The choice comes down to accuracy you can live with versus speed and cost you can afford.
Machine translation (Google Translate, DeepL)
Classic machine translation hits 70–85% accuracy on common language pairs — fine for a personal video where viewers tolerate a clumsy phrase, not fine for a product launch or anything a compliance officer will watch. The upside: it's free or nearly free, and you can paste an SRT file in and create a translation in seconds.
Neural AI subtitle translation
Modern neural machine translation models clear 90%+ accuracy on high-resource pairs like English↔Spanish, German, or French. A purpose-built AI subtitle translator now matches mid-level human work on straightforward content in minutes instead of days, and many tools support 60 to 100+ languages out of the box. The category has matured quickly: real-time subtitle translation for live broadcasts is now a separate product class, powered by the same models.
Professional human subtitling
Professional subtitle translation services hit ~99% accuracy and price for it — €3 to €8 per minute of video content for EN→major-EU pairs, more for rare languages. For regulated content (healthcare, finance, legal), broadcast TV, or flagship brand campaigns, human services are still the default. FCC fines and ADA lawsuits cost more than translators.
Hybrid (AI draft + human review)
The subtitling workflow most translation agencies run today. AI generates editable translations, a human subtitler cleans the output, fixes cultural references, and tightens timing. Turnaround drops from weeks to hours while keeping 95%+ accuracy on branded content — and it's the one approach that scales to localized content for dozens of markets per week.
The single cheapest quality upgrade in this entire workflow is a glossary. At Dubly, we see the biggest cost in subtitle translation isn't translation itself — it's the review loop when machine output misses cultural cues. A 50-line glossary of brand terms, character names, product vocabulary, and recurring phrases, shared with the AI and the human reviewer, cuts review cycles by roughly 40% across the teams we work with. One creator we work with went from three review rounds per video down to one after setting up a glossary for their recurring cast and product names.
Build it before you start translating. It's not optional for serial output — it's the difference between a subtitling workflow that scales and one that burns hours every week re-fixing the same mistakes.
Subtitle Formats You Should Know
Once you've picked a method, the next decision is what format to export in. Which subtitle file you choose matters more than creators realize — the wrong format can break in certain players, strip your styling, or fail to upload. Here's what the main options do:
| Format | Full name | Best for | Notes |
|---|---|---|---|
| SRT | SubRip Text | YouTube, social, most players | Most popular and widely-used; plain text, no styling |
| VTT / WebVTT | Web Video Text Tracks | HTML5 video, streaming | Supports styling, positioning, metadata |
| SSA / ASS | SubStation Alpha | Anime, fan-subs, stylized | Rich formatting, fonts, colors |
| SUB | MicroDVD / SubViewer | Legacy players | Frame-based, less common today |
SRT is the safest default. It's supported by YouTube, Vimeo, Facebook, and virtually every player shipped in the last 15 years. For HTML5 video or a modern streaming platform, VTT gives you styling and positioning — it's the format the W3C standardized for browser-native video in 2014. SSA/ASS only matters if you need karaoke-style effects or heavy stylization, and it's almost entirely used in anime and fansub communities.
One practical note: most AI tools let you upload a video and export subtitles in multiple formats from a single pass. If yours doesn't, look elsewhere — having to convert between formats manually is a signal the tool isn't built for production use.
Best Subtitle Translation Tools
I've tested most of the tools on this list, some for clients and some just to see how they hold up against what we build at Dubly. Here's what's actually worth using, with one gotcha per option worth knowing.
- Dubly — Full video translation across 40+ languages. Creates subtitles, native dubbing, and lip sync in one workflow. Good fit for creators and social media teams who need subtitles plus translated audio from the same upload. Exports SRT, VTT, WebVTT.
- HeyGen — Avatar-first, ~40 languages. Best if you're producing AI presenter videos where the avatar is the main event. Subtitle output is solid but secondary to the avatar rendering — which you're paying for either way.
- Rask AI — 130+ languages on paper, but quality drops noticeably past the top 20 pairs. Good for high-volume creator work where the language breadth matters more than the last 5% of accuracy.
- Subtitle Edit — Free, open-source desktop editor. Best tool on the list for manually editing SRT files and format conversion. Professional subtitlers use this daily. The UI is utilitarian — you're here for the engine, not the design.
- Aegisub — Free, cross-platform. Powerful for stylized subtitles (karaoke, positioning, fonts), but the interface is stuck in 2008. Skip unless you're in the anime or fansub world.
- OpenL / Smartcat — SRT-focused AI translators. Fast for simple files. Limited editing features compared to a full subtitle editor, so expect to clean up output elsewhere.
- Google Translate / DeepL — Baseline machine translation in 100+ languages. DeepL is noticeably better than Google on European pairs; Google wins on coverage. Fine for drafts, not finals.
For AI-specific tool depth, jump to our AI subtitle translator deep-dive.
How Much Does Subtitle Translation Cost?
Subtitle translation is priced three ways, and they are not comparable: software charges per minute of video, human subtitlers charge per word or per minute at rates that vary widely by language pair, and free tools charge you in hours. Which model wins depends on two things — how many languages you need, and how public the video is.
The full breakdown, including what the flat-per-language behaviour of software does to a four-language budget, is in what subtitle translation costs. For at-volume AI pricing, see Dubly pricing.
One pattern worth naming here: the buyer who insists on a free workflow usually pays more in review time than a paid tool would have cost. The real decision is not free against paid — it is whether your own hours are worth the saving.
Free vs. Paid Subtitle Translation — When Free Actually Works
"Free" shows up in almost every subtitle translation search because there's real demand to avoid paying for simple subtitling jobs. Sometimes the free stack is fine. Sometimes it's the most expensive option in disguise.
Free works when: the video is personal, internal, or user-generated content; translation quality isn't brand-critical; the language pair is common (EN↔ES/DE/FR/IT); and you have time to review and fix the machine output. Google Translate plus Subtitle Edit is a legitimate free stack for a creator making one video a month.
Free falls short when: it's public-facing brand content; you need compliance (FCC, ADA, WCAG); you're creating subtitles for more than a few videos per month; the target language is low-resource (MT drops below 70% on those pairs); or timing needs to be precise — most free tools skip sync entirely. For anyone doing subtitle translation as part of a job, the time cost alone makes paid subtitling tools worth it within the first week.
Subtitle Translation Best Practices
The rules below aren't theoretical — they're the ones I see teams break first, and the ones that quietly cost the most time in rework.
Stay inside 42 characters per line, two lines maximum. The rule most often broken. Mobile viewports clip anything longer, which is where most of your audience actually watches.
Hold each subtitle on screen for 1 to 6 seconds. Under one second reads as a flash; over six breaks the reading rhythm and viewers lose focus on the video itself.
Target ~3 words per second. Half your audience can't keep up past that, especially in German and other languages where compound words take longer to parse.
Give the translator the video, not just the SRT. This is the one everyone skips, and it's the biggest accuracy drain I've watched happen. A line about "the bar" reads very differently if the visual is a courtroom versus a pub. Visual context is half the meaning.
Adapt idioms and culture-specific references. Machine translation alone won't nail this. If a joke depends on a cultural reference the target audience won't get, rewrite it for the local market rather than translating literally.
Review anything public-facing. AI drafts are fine internally. For branded or regulated content, human review from professional subtitle translation services isn't optional — and the cost of skipping it compounds fast across languages.
Accessibility: FCC, ADA & WCAG Compliance
Subtitle translation and accessibility overlap more than creators realize. Making video content accessible to deaf and hard-of-hearing viewers in multiple languages isn't a feature — it's a legal requirement in several markets, and it's what turns a video into truly accessible media. If you're distributing video in the US or building for a regulated industry, these rules apply regardless of language:
- FCC closed captioning rules require video programming distributors on US television to caption nearly all programming. Quality rules cover accuracy, synchronization, completeness, and placement.
- ADA has been applied to web video in a growing series of lawsuits — public-accommodation websites and apps need captions on their video content.
- WCAG 2.1 Level AA requires captions for prerecorded audio content and audio descriptions for video. Most enterprise procurement now requires WCAG AA conformance.
For SDH specifically, you need to include speaker identification, sound effects, and music cues in brackets. "Closed captions" and "SDH" aren't interchangeable — US closed captioning rules assume SDH-style non-speech descriptions, while subtitles in most other markets don't. If you're translating subtitles for enterprise delivery, talk to procurement about the target standard before starting. Professional subtitling services in regulated industries will handle this automatically, but DIY workflows need to understand the distinction.
One underrated benefit: search engines read subtitle tracks. Translated subtitles give your videos localized content that YouTube and Google index in the target language, which means the same upload can rank in multiple markets and reach audiences in languages you aren't actively creating for. For social media teams, this is often the easiest way to scale reach without scaling production.
AI Subtitle Translation: What's Actually Changing
A few things are genuinely different about AI subtitle translation in 2026:
Neural models match humans on common pairs. For EN→ES/DE/FR/IT, a modern AI subtitle translator is indistinguishable from a mid-level human subtitler on neutral content. Where AI still lags: cultural nuance, regulated vocabulary, brand voice, and low-resource languages.
Turnaround collapsed. Work that took a week in 2022 now takes an hour — often under 30 minutes for a 10-minute video. Per-video cost dropped by roughly 90% compared to a traditional human-only workflow. That changes what you can do: same-day global launches, weekly uploads in eight languages, ad variants tested across markets in a single sprint.
Lip sync and subtitles are converging. For creators producing both subtitled and dubbed versions, tools like Dubly handle transcription, translation, subtitle export, voice cloning, and lip sync in one run. You're not running three workflows for the same video — you're running one.
We see creators at Dubly who start with one language pair — say EN→ES for their YouTube channel — scale to three or more within three months once the subtitle and dubbing workflows collapse into a single run. Reaching new audiences in new markets gets cheaper the second and third time you do it.
Where AI still falls short, honestly: cultural nuance in idiom-heavy content, low-resource languages (anything below the top 30 pairs), and any content where a wrong word has real legal or medical consequence. For those, human-first is still the only responsible call. When you have to choose between speed and certainty, choose an editor in the loop.
Conclusion
Subtitle translation is the lowest-cost, fastest way to make video content reach audiences in their own language. Pick SRT as your default format, respect the 42-character and 1–6-second limits, and keep a human in the loop on anything public-facing or regulated. If you also want dubbed audio and lip sync from the same upload, that's where a full platform earns its place over a subtitles-only tool.
Back to the complete guide: Subtitles — the complete guide
Translate Your First Video
Results in just a few minutes
No credit card required
Best translation quality worldwide
About the author
Leon Bach
Growth Marketing Manager