FAQ

Currently, dubbing is available from any language into Latvian, Russian, and Ukrainian. In development are Lithuanian, Estonian, English, Chinese, Japanese, German, French, Spanish, Korean, Arabic, Danish, Italian, Polish, and Portuguese.

If you know the language you want to dub your movie into, you can create a high-quality dubbing yourself in 3-4 days using AI Studio. You’ll need to check and correct every line in the movie yourself. The AI ​​will automatically produce the translation, but you’ll still need to manually check it. Currently, AI isn’t capable of dubbing movies with 100% quality; human review is required.

Yes, it’s possible. However, the quality of automatic AI dubbing will not be 100%. Depending on the language, the quality of automatic dubbing will range from 80 to 95%. If 100% quality is required, a human reviewer will be required. In our AI Studio program, you can dub yourself with 100% accuracy or entrust the dubbing (with 100% accuracy) to our team. In this case, MultiVoice.AI will review the AI ​​work by a human (native speaker) and provide a professional translation of all phrases in the film. The production time for a 90-minute film dubbing (with 100% quality) by our team is approximately 1 week. We currently offer this service for Latvian, Russian, and Ukrainian. In the future, we will offer dubbing for Lithuanian, Estonian, English, Chinese, Japanese, German, French, Spanish, Korean, Arabic, Danish, Italian, Polish, and Portuguese.

The main difference between dubbing and voice-over is that with dubbing, the original audio track is completely replaced with a new, translated one, while with voice-over, the original audio track is heard together with the translation, which is usually heard over the original.

• The original audio remains audible, but the translation is overlaid on top. Therefore, the original audio may be distracting to some viewers due to the overlapping sounds. • These days, consumers are accustomed to big-budget films being fully dubbed.

Voiceovers are now being used for documentaries, streams, and podcasts. Viewers can hear the presenter’s original voice and simultaneously hear a translation into a language they understand. Voiceovers (also known as voiceovers) are easier and cheaper to produce than full dubbing.

Yes, it’s possible. Through our API, developers can order automatic subtitles, as well as human-verified subtitles.

The quality of automatic subtitle creation depends on many factors. Currently, there’s no way to create automatic subtitles with 100% accuracy. If you want 100% accurate subtitles, you’ll need to have them reviewed and edited by a human. We offer this option, and then the accuracy will be 100%.

For a 90-minute duration, automatic subtitles are created in approximately 40 minutes. We use large AI models to create subtitles, so subtitle processing takes time and still delivers the highest quality. Subtitles can be created in a shorter time, but accuracy will be compromised. If you use our service—subtitle creation and review by a native speaker—subtitle creation will take 1-3 days. You can also adjust the subtitles yourself using our AI Studio program.

Yes, you can. In the API section, you’ll find all the technical specifications for integrating our transcription service into your project.

Speech recognition quality varies depending on the language and audio source. If the audio contains background noise or speech is unclear, recognition accuracy will decrease. If you want 100% speech recognition accuracy, use our optional verification service with a professional native speaker. Our expert will then verify the AI’s results, ensuring 100% accuracy.

No. We can recognize speech in existing (recorded) audio files. You upload the file to our service via the API (for developers) or through AI Studio (for Windows and Mac), and after processing, you receive the finished text. File processing occurs on our servers. Processing takes some time.

Yes, you can. See the API section of our website.

We independently develop speech synthesis models for various languages. Our first model is a speech synthesis model for Latvian, Russian, and Ukrainian, using 24 different voices (10 male, 10 female, and 4 child voices). Our team independently creates and updates the dataset for the speech synthesis model monthly. We have hired Latvian, Russian, and Ukrainian philologists to validate and refine our language model. We believe that our speech synthesis model for Latvian, Russian, and Ukrainian is as good as AI speech synthesis from Google, Amazon, Event Labs, and Tilde (for Latvian).

You can test speech synthesis in Latvian, Russian and Ukrainian (using 24 voices) using this link.

Yes. Translations can be certified with the translator’s digital signature. In certain cases, the translator’s authenticity can be certified by a Latvian notary.

We specialize in literary translation of films for dubbing and voiceover. We also specialize in translating and transcribing audio recordings of court proceedings into other languages. We can also provide other types of translation.

We can currently offer this service for Latvian, Lithuanian, Estonian, English, Russian, German and Polish.

Legislation varies across countries. For example, in the EU, broadcasting films (and other video content) on television during the daytime is prohibited, especially those containing scenes of violence and cruelty that could negatively impact the psyche of children. Also, some countries prohibit the promotion of smoking (tobacco).

Our service checks a 90-minute film within 2-3 days.

In the event of a censorship audit, our company is prepared to enter into a contract and guarantee the compliance of the reviewed content with local legislation for five years (or until legislation changes in the country for which the censorship audit is being conducted). We are prepared to cover potential fines from TV regulators of up to €10,000. We are currently prepared to conduct censorship audits for the following countries: Latvia, Lithuania, Estonia, Poland, Ukraine, and Russia.

Yes, you can. You can download courtroom transcripts and audio recordings through our API.

If only one language is spoken in a courtroom (for example, only Latvian or only Russian), then speech recognition is very good. It also depends on the quality of the court’s audio recording. If the audio recording is barely audible, speech recognition will be poor.

Only you will have access to the files you upload. Once the files are processed, you can delete them from our system.