Instructions for Adding Voiceovers to a Production
Kapwing, a popular video editing platform, now offers a dubbing feature that allows users to add voiceovers to their videos in a variety of ways. Here's a breakdown of what you can expect when using Kapwing's dubbing tool.
Default Dubbing and Voice Options
By default, Kapwing uses voice clones taken from the source video if voice cloning is supported for that language. Users can choose the target language and dialect when starting a new dubbed project. However, it's important to note that the platform does not support bulk import or export, with each video needing to be uploaded and exported individually.
Stock Synthetic Voices and Voice Cloning
Kapwing offers two main options for dubbing: stock synthetic voices and voice cloning (custom voices). The platform provides realistic AI-generated text-to-speech voices in over 40 languages for the former. For Business and Enterprise customers, Kapwing can clone the original speaker's voice from the video, creating a highly realistic custom voiceover that matches the original speaker.
Managing Multiple Speakers
Kapwing detects speaker changes in videos with multiple speakers and automatically creates different voices for each to improve dubbed audio quality. Users can also upload their own SRT files and use custom pronunciation guides or glossaries to tweak the dubbing output.
How to Use the Dubbing Feature
To use the dubbing feature on Kapwing, follow these steps:
- Upload your video.
- Use the "Dub Audio" feature to generate a transcription.
- Select the language and the voice type (stock synthetic or cloned if eligible).
- Customize timing and lip sync options if desired.
- Download your dubbed video or audio after the AI generates the dubbed voiceover.
Additional Features
Kapwing's dubbing tool offers several other features, including automatic speech recognition (ASR) and captioning, machine translation, background sound preservation, lip sync, support for uploaded SRT files and Slavic, Arabic, and RTL languages, realistic synthetic voices, import from YouTube and Google Drive, speaker labels and dubbing for videos with multiple speakers, translate text embedded in the video, advanced timing and speed adjustments technology, real-time collaboration and team comments, custom spelling glossary, translation rules available across the brand workspace, regeneration of text-to-speech layers, download SRT of translated captions or embed captions into the video, download dubbed MP3 or Mp4, video uploads up to 6GB and 2 hrs long, dialect-specific language selection, and a custom pronunciation guide.
Limitations and Availability
Expressive emotional control is limited—the voice cloning uses the same voice variant throughout, with punctuation-based inflection tweaks possible but no deep emotional variance. Voice cloning is not supported in certain languages, as marked with an asterisk in the list of supported languages. Business plan customers can save up to 2 voice clones in their Brand Kit.
Kapwing's Dubbing platform supports over 40 languages for dubbing, similar to its text-to-speech capabilities. The list of supported languages includes English (US, UK, AUS), Arabic, Bulgarian, Chinese (Mandarin), Croatian, Czech, Danish, Dutch, Finnish, Filipino (Tagolog), French, German, Greek, Gujarati, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Kannada, Korean, Lithuanian, Malay, Malayalam, Norwegian, Polish, Portuguese (Brazil, Portugal), Romanian, Russian, Slovak, Spanish (Spain, Mexico), Swedish, Tamil, Telugu, Turkish, Ukrainian, and Vietnamese.
To add a voice clone, users must be a Business customer and upload an example of the speaker whose voice they want to clone. Users can also upload an SRT file for the dubbing process.
In summary, Kapwing supports realistic stock voices by default, offers voice cloning for higher-tier users, and automatically manages multi-speaker dubbing with options for customization via glossaries and timing adjustments.
Technology plays a significant role in Kapwing's dubbing feature, as it uses artificial intelligence (AI) for text-to-speech voices, automatic speech recognition (ASR), machine translation, and advanced timing adjustments. Users can expect high-quality dubbing due to technology like voice cloning that replicates the original speaker's voice. However, there's a limitation in terms of expressive emotional control, where the same voice variant is used throughout the video with punctuation-based inflection tweaks as the only available option.
