If your transcript repeats the same words over and over, it is a known speech recognition limitation, not a device defect. This guide gives three prevention tips. Background music or noise is the most common trigger.
Applies to: v4.0.x or later. Still using v3.30.x or earlier Plaud Mobile App? See Repeated output in transcription.
Steps
Repetition is a known limitation of speech recognition models (often called hallucination), typically triggered by background music or noise rather than a device defect.
- Record in a single language: mixing languages in one recording raises the error rate.
- Avoid recording music: ringtones and background music are common triggers.
- Reduce background noise: cleaner audio in, cleaner transcript out.
If the issue persists
Repetition in clean, single-language recordings is worth investigating. Submit a request and include the following so our support team can investigate:
- Your device model and Plaud Mobile App version
- The affected recording or a screenshot of the repeated transcription
- The recording conditions: language, background noise, and any music
Disclaimer: The information in this article is for reference only and may change as Plaud product features continue to update. Please refer to the latest version of the product for the most current experience.