Practical overview
When it fits
Use it when a desktop pipeline is needed to translate supported-language audio or video into Chinese or English speech or subtitles, especially when sentence-level text editing, bilingual mixing or batch processing is useful.
Who it is for
- People preparing Chinese or English dubbed versions of Japanese, English or Chinese audio or video.
- Users who need translated subtitles without generating speech.
- Audio producers who want a bilingual mix, a separate dubbed track or batch processing for multiple tracks.
Documented uses
- Recognize and translate speech in supported source-language media.
- Generate translated speech, including voice-cloned dubbing from a source reference sentence.
- Mix translated speech with the original, with optional spatial-cue processing.
- Export dubbed media, a separate speech track, or translated subtitles.
- Process multiple audio tracks from a work folder in a batch.
Requirements & limits
The README documents Windows 10/11 and Linux x86_64 support. The Windows release can be launched without separately installing Python. Translation setup requires an API key for DeepSeek or another provider; Edge TTS requires an internet connection.
- Local-model processing is described as keeping content on the device; cloud translation sends dialogue text, and cloud speech generation sends translated text and reference audio to the provider.
- IndexTTS2 voice cloning is described as running on the user’s GPU. Edge TTS provides online fixed-voice generation without a GPU, while CPU recognition is slower.
- Recognition and translation can be skipped when the corresponding source or translated subtitles are available.
- Voice separation and replacement output are experimental and may leave original speech or damage sound effects.
- API keys are stored in plain text in the application configuration folder; the README advises not to share that folder.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Set up a Japanese-to-Chinese dubbing project on Windows
Install the packaged application, obtain its initial models, configure translation and start the documented four-stage workflow.
Before you start
- A Windows 10 or 11 computer.
- An API key for DeepSeek or another translation provider.
- An internet connection to download the release, models and configure the cloud translation service.
Steps
Download the release archive, extract it to a short-path folder and launch ASMR-Dubber.exe.
On the Models page, select the combined download for Parakeet Japanese recognition and IndexTTS2 voice cloning.
In Settings → Cloud Services, enter a translation API key.
On the Project page, add the audio, select the source and dubbing languages, then follow recognition, translation, dubbing and export.
Expected outcome
A project configured to process Japanese audio into Chinese dubbed output and proceed through the application’s recognition, translation, dubbing and export stages.
Guide limitations
- This setup follows the README’s Japanese-to-Chinese example; English or Chinese source audio uses Faster-Whisper for recognition.
- Voice cloning with IndexTTS2 is described as running on the user’s GPU. Without a GPU, the README recommends Edge TTS for fixed-voice speech generation and CPU recognition, which is slower.
- Using cloud services sends dialogue text, and cloud speech generation also sends translated text and reference audio to providers.
- Voice separation and replacement output are experimental and may leave original speech or affect sound effects.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 14 October 2026
- Original website title
- EveningStudy/asmr-dubber · Official homepage title · retained in its source language