Voiceover in my project is inconsistent (distorted). What should I do?

Rarely the voiceover might have inconsistent voice, including fluctuations or a creaking voice. Such problem is associated with ML models that we use and can be solved by simple regeneration of the segments with issues.


You can choose one of the following options to fix the voiceover problem:

  1. Make a small adjustment to the segment, like deleting and re-adding a punctuation mark or inserting a line break using the "Enter" key or making other changes to translated text.
    This way, our system will recognize it as a modification. Then, save the segment & click the "re-dub" button and voiceover for the modified segment will be regenerated.

  2. Merge the distorted segment with the previous one that has better quality, and then redo the voiceover for this merged segment.

    Screenshot 2025-04-21 at 18.36.32

  3. Split the segments with distorted voice into smaller parts, adjust the timing, and redo the voiceover for each of these smaller segments.

    Screenshot 2025-04-21 at 18.37.33

  4. Switch to one of our preset voices by changing the voice assigned to speaker.

    Screenshot 2025-04-21 at 18.38.24

    Screenshot 2025-04-21 at 18.38.53

We hope these tips will help you achieve the desired result and enhance your content localization. If you have any questions, feel free to reach out to us through the support chat or via email at support@rask.ai.