You may occasionally see words replaced with asterisks (***) in DeepL Voice transcripts and translations. This is intentional. DeepL Voice automatically masks profanity in the text it produces.
What is masked
Profanity is masked in text output only. This includes:
- Live transcripts
- Translated captions
- Transcripts you download or share
Masking applies to both the original transcription and the translation. In some cases, words next to the masked term are replaced as well. This is more likely in Japanese, Chinese and Thai, where words are not separated by spaces and DeepL Voice cannot always determine exactly where a term begins and ends.
Speech recognition can occasionally mishear a word, including names and technical terms, as something offensive. When that happens, the word is masked. Asterisks in a transcript do not necessarily mean a participant said anything inappropriate.
Spoken audio is not masked
Can I turn masking off?
No. Profanity masking is always on. It cannot be changed by users or by team admins.