Katarzyna Petru
YouTube is developing an automatic dubbing feature to facilitate film viewing for people who do not know the original language. The tool uses artificial intelligence to translate dialogue and replace the original soundtrack with a voiceover in another language. Currently, automatic dubbing supports 27 languages, and users can specify their preferred language in their account settings. If a particular language version is available, YouTube will play it automatically so that content from around the world is understandable immediately after clicking "play".
Less "robotically". YouTube improves auto-dubbing
YouTube admits that automatic dubbing can be annoying if it sounds too artificial or does not match the visuals. That is why the platform has introduced a new feature, Expressive Speech, which aims to better convey the tone of voice, emotions, and speech rhythm in the translated audio track. The goal is to make the dubbing sound as natural as possible and closely resemble the original. The feature is already available for channels in several languages, including Polish, English, French, German, Spanish, and Portuguese, and YouTube plans to gradually expand the list. At the same time, a pilot Lip Sync system is also being tested, which gently adjusts the lip movements of the speaking person to the translated sound, aiming to reduce the effect of "mismatched" visuals and audio. Although the dubbing is created automatically, creators retain full control — they can completely disable this feature or replace it with their own language version. YouTube also employs intelligent filters that prevent dubbing of content where translation does not make sense, such as films without dialogue or materials based entirely on music.
YouTube acknowledges that automatic dubbing is still not without its flaws. The problems most often arise from inaccurate speech recognition or poor sound quality in the original material. However, the platform emphasises that the system will be gradually improved as more data and user feedback come in. Meanwhile, YouTube is increasingly focusing on personalisation based on artificial intelligence. An example is the Recap feature, which assigns users a specific "profile" based on their viewing history, helping to better tailor the recommended content and the way it is presented. This is another step towards a more intelligent and personalised viewing experience.
Source: digitaltrends.com
