What is Dubbing 2.0: Meaning-first translation approach

Dubbing 2.0 is Rask's dubbing engine built on a meaning-first approach. Rask translates with the context of your full script, adapts each line to its timing and voices it with a new speech model that keeps the speaker's emotion and natural pace. The result sounds like the speaker actually said it in another language.

Dubbing 2.0 is selected by default for new projects. Dubbing 1.0 stays available, so you can choose the engine project by project. Projects created through the Rask API use Dubbing 2.0 in every language.

Dubbing 1.0 vs Dubbing 2.0

	Dubbing 1.0	Dubbing 2.0
Translation	Line by line, shaped early to fit the original timing	Follows the meaning of the whole script, so phrasing sounds natural to native speakers
Voice	Clear and consistent	Keeps the speaker's emotion, intonation and energy
Pacing	Each line fits its original slot, often by speeding the voice up or slowing it down	Lines can extend into natural pauses by up to 0.5 seconds, so speech flows at a human pace
Auto-dub	Fits the text to the timing	Adapts the wording to the timing while keeping the meaning
Translation options in the editor	Several variants to choose from	One translation that you edit directly
Generation time	Faster	A little longer, since Rask analyzes the full script and builds a more expressive voice
Languages	All supported languages	29 target languages in the web app (see the list below), every language in the API

How Dubbing 2.0 works

Meaning-first translation. Rask translates each line with the whole project text in mind, so names, terms and tone stay consistent and sentences read the way people actually talk.
Timing. Rask adapts the phrasing and placement of each line to the length of the original segment. The final track matches the video length while keeping a natural pace.
Expressive voice. Voiceovers come with clearer accents, natural emotion and a voice that stays stable from the first line to the last. Voice clones pick up the delivery of the original speaker as well as its sound.

How to use Dubbing 2.0

Upload a video as usual.
In the upload window, check that "Dubbing 2.0" is selected. It is marked "Recommended".
Keep "Auto-dub" on. Dubbing 2.0 then tunes the wording and the timing together. With a manual dub, Rask keeps your text as written and fits its length to the timing by adjusting the audio.
Review the translation as usual and listen to the result at 1× before you export.

To use Dubbing 1.0 for a project, select it in the same upload window.

Supported languages

Dubbing 2.0 supports 29 target languages: Arabic, Bulgarian, Chinese (Mandarin), Croatian, Czech, Danish, Dutch, English (US), Filipino, Finnish, French (France), German, Greek, Hindi, Indonesian, Italian, Japanese, Korean, Malay, Polish, Portuguese (Portugal), Romanian, Russian, Slovak, Spanish (Spain), Swedish, Tamil, Turkish and Ukrainian.

When your target language supports Dubbing 2.0, you'll see the option in the upload window. In the web app, regional variants such as Brazilian Portuguese, Latin American Spanish, British English and Canadian French use Dubbing 1.0 for now. Through the API, every language, including these regional variants, is dubbed with Dubbing 2.0.

Who can use it

Dubbing 2.0 is available to every Rask user, including the free trial. Prices are now lower on every plan, and one minute of video still uses one minute of your balance. In a Teamspace, everyone in the workspace can use it under the owner's plan. Dubbing 2.0 works in the Rask web app and through the Rask API.

Dubbing 2.0 in the API

Every project created through the API is processed with Dubbing 2.0, in all supported languages, including regional variants such as Brazilian Portuguese.
Dubbing 2.0 is the default model, so you don't need to add or change any parameters.
A response field that shows which engine was used for a project is coming to the API soon.
Creating a project from an existing video_id and transcript_id starts dubbing right away. To use a different voice, change it on the project and regenerate.

FAQ

What happens to my existing projects?
They keep the engine they were dubbed with, and files you exported earlier stay unchanged. A full re-dub moves an existing project to Dubbing 2.0, while re-dubbing a single segment keeps the project on its current engine.

Does it work with my glossary and custom instructions?
Yes. Your translation dictionary and translation instructions are applied to the translation, and the voice is generated with Dubbing 2.0.

Can I go back to Dubbing 1.0?
Anytime. Choose Dubbing 1.0 in the upload window for any new project.

Why does generation take longer?
Dubbing 2.0 analyzes the full script before it translates and builds a more expressive voice, so the voiceover takes a little more time to generate.

What happens when I edit the translation?
Rask voices your wording as written. After you edit a line, apply the changes or redub to hear the new audio.

If you have any questions, please contact us via the support chat or support@rask.ai.