Your video is doing well and a growing share of viewers don't speak its language. To reach them you have two broad options: put translated text on screen, or give the video a new voice. It's a decision about budget, audience habits and how much of the original performance you want to keep.
Quick answer
Subtitles translate the dialogue into on-screen text. They are faster and cheaper to produce, keep the original voices and work well for audiences used to reading along. Dubbing replaces the spoken audio with a new language, either as voice-over (narration over the original) or lip-sync dubbing (new dialogue fitted to mouth movements). Dubbing costs more but lets viewers watch without reading. Most teams start with subtitles and dub selectively.
| Subtitles | Voice-over | Lip-sync dubbing | |
|---|---|---|---|
| What changes | Text added on screen | Translated narration over original audio | Original dialogue replaced |
| Relative cost | Lowest | Medium | Highest |
| Turnaround | Fastest | Medium | Slowest |
| Keeps original voices | Yes | Partly (audible underneath) | No |
| Viewer effort | Reading required | Listening only | Listening only |
| Best for | Most online video, film fans, language learners | Documentaries, interviews, training | Entertainment, kids content, ads |
| Works with visuals-heavy content | Can compete for attention | Yes | Yes |
| Accessibility for deaf viewers | Can help (SDH even more) | No | No |
| Indexable text for search | Yes | No, unless you also publish a transcript | No, unless you also publish a transcript |
What subtitles do well
Subtitles translate meaning, not every word. Viewers read slower than they listen, so subtitlers condense lines to fit a comfortable reading speed and break them into short, well-timed segments. Done well, they're nearly invisible: the viewer follows the story and still hears the actor's voice, tone and timing.
Strengths:
- Authenticity. The original performance, humor and emotion stay intact.
- Speed and cost. Translation plus timing is a fraction of the work of recording new audio, so you can cover many more languages.
- Flexibility. Closed subtitle tracks can be switched on or off and corrected after publishing. See open vs closed captions for the trade-offs.
- Discoverability. Each subtitle track is text in a new language that platforms can read.
- Learning. Many people use subtitles to learn or practise a language.
Weaknesses: reading splits attention from the visuals, which matters for fast-paced, detail-heavy or instructional video. Young children and people with reading difficulties or low vision may struggle. And subtitles alone aren't a substitute for captions for deaf viewers, because they usually skip sound cues; that's the difference covered in captions vs subtitles.
What dubbing does well
Voice-over
In voice-over (sometimes called UN-style voice-over), the original audio plays at a lower volume and a translated voice speaks over it, usually starting a moment after the original speaker. The audience hears that a real person is speaking, but follows the translation. It's common for documentaries, news, interviews, corporate and training video. It's cheaper than lip-sync dubbing because lines don't have to match mouth movements.
Lip-sync dubbing
In lip-sync dubbing, the original dialogue track is replaced. Adapters rewrite the translation so it fits the length, rhythm and visible mouth shapes of each line, voice actors perform it in a studio, and engineers mix it with the original music and effects. It's the most immersive option and the most demanding: adaptation is a creative task, not just translation, and casting matters.
Strengths of dubbing in general:
- No reading. Viewers can watch the visuals, cook dinner or follow along on a small screen.
- Reach. It serves children, people with reading difficulties and audiences who simply prefer it.
- Emotional connection. A good dub in your own language can feel more personal than text.
Weaknesses: higher cost and longer turnaround per language, a new voice that some viewers dislike, and translation compromises forced by lip-sync timing.
Audience preferences
Habits matter as much as budget. Broadly:
- Dubbing traditions. Several large European markets, including Germany, France, Italy and Spain, have long histories of dubbing foreign film and television, and many viewers there expect it.
- Subtitle traditions. Smaller language markets, such as the Netherlands, Scandinavia and parts of Central Europe, have long relied on subtitles, partly because dubbing for a smaller audience was harder to justify.
- Voice-over traditions. In some countries, such as Poland, single-voice voice-over has been a common way to translate TV.
- Shifting habits. Streaming has put both options in front of everyone, and preferences now vary by age, genre and device as much as by country. Kids' content is dubbed almost everywhere.
The best evidence is your own: look at your analytics for where viewers come from, test a subtitled and a dubbed version of a video in one market, and compare watch time.
AI dubbing: what to watch for
Synthetic voices, voice cloning and automated translation have made dubbing far cheaper to try. Some platforms now support multiple audio tracks on one video, so viewers can pick a language. Before you publish an AI dub, check for:
- Translation errors. Idioms, jokes, product names and technical terms often come out wrong. Review the translated script before generating audio.
- Pronunciation. Names, brands and acronyms are frequent failures.
- Flat delivery. Emotion, emphasis and pacing can sound off, which matters more for storytelling than for tutorials.
- Timing drift. Some languages take longer to say the same thing. Lines can overrun scenes or clash with on-screen text.
- Consent and disclosure. Cloning someone's voice needs their permission. Be transparent with your audience about synthetic voices, and check the policies of the platforms you publish to.
AI dubbing is a reasonable choice for tutorials, explainers and internal training where clarity matters more than performance. For brand campaigns and high-stakes content, budget for human review at minimum.
How to decide
| If... | Lean toward |
|---|---|
| You need many languages on a small budget | Subtitles |
| Your audience is mostly adults on desktop or TV | Subtitles, test dubbing later |
| Your content is for children | Dubbing |
| Viewers need to watch your hands or the screen (tutorials, cooking, software) | Voice-over or dubbing |
| The speaker's personality is the product (interviews, comedy, vlogs) | Subtitles |
| You're entering a market with a strong dubbing tradition | Dubbing for flagship content |
| You want search and AI tools to find your content in other languages | Subtitles (plus published transcripts) |
A common sequence: subtitle everything into your top languages, then dub the handful of videos that perform best in each market.
The workflow for each, starting from one transcript
Both paths start with the same asset: an accurate, timed transcript of the original audio.
Subtitles
- Get a clean transcript with timestamps and review names and terms.
- Translate, condensing lines to a comfortable reading speed.
- Re-time segments if the translation changes length, then export SRT or VTT. Our SRT vs VTT guide covers the formats.
- Have a native speaker review in context, watching the video, not just reading the file.
Dubbing
- Same clean transcript, plus notes on speakers, tone and on-screen text.
- Translate and adapt for speech: natural spoken phrasing, timing that fits each scene, lip-sync where needed.
- Record with voice talent, or generate with a synthetic voice.
- Mix with the original music and effects track (ask for an M&E track if you're working with a production partner).
- Review the final mix with a native speaker.
- Publish as an alternate audio track where supported, and keep subtitles available too.
Getting the transcript and translation
The first step in either workflow is a reliable transcript. For YouTube videos, YouTubeTranscript.dev pulls existing captions or transcribes the audio with AI, and can translate the transcript into 100+ languages and export SRT or VTT. Try a single video with the free translator, or use the API for a whole channel. Machine translation is a strong first draft for subtitles or a dubbing script, and it still deserves a native speaker's review before it goes out under your name.
Frequently asked questions
Is it better to watch with subtitles or dubbing?+
Neither is objectively better. Subtitles keep the original performance and are cheaper to produce; dubbing lets viewers watch without reading and suits children, multitaskers and markets that grew up with dubbed TV. Many viewers have a strong preference shaped by what they grew up with.
Is dubbing more expensive than subtitles?+
Almost always. Subtitles need translation and timing. Dubbing needs translation adapted for speech, voice talent or a synthetic voice, recording, mixing and review for every language. AI dubbing narrows the gap but still needs human review for quality.
What is the difference between voice-over and dubbing?+
In voice-over, a translated narration plays over the original audio, which is often still faintly audible. In lip-sync dubbing, the original dialogue is replaced and the new lines are written and timed to match the speakers' mouth movements. Voice-over is common for documentaries and interviews.
Is AI dubbing good enough to publish?+
For some content, yes, after review. Synthetic voices and voice cloning have improved, but they can mispronounce names, flatten emotion, mistime lines and mistranslate idioms. Voice cloning also raises consent questions, so get permission from the original speaker.
Which countries prefer dubbing over subtitles?+
Several large European markets, such as Germany, France, Italy and Spain, have long traditions of dubbing foreign films and TV. Smaller language markets, such as the Netherlands and the Nordic countries, have long favored subtitles. Preferences are shifting with streaming, especially among younger viewers.
Do subtitles or dubbing help YouTube videos reach more viewers?+
Both can. Translated subtitles are the low-cost first step and also give platforms indexable text in each language. Dubbed audio tracks can reach viewers who won't read subtitles. Many creators start with subtitles and add dubbing for their best-performing videos.