#4339 videotrans.configure._except.StopRetry: No Ref Audio. Clone disabled.

This error occurs because the audio cloning feature is enabled (a clone voice role is selected) but no valid reference audio is found or the extracted speech segments are outside the required duration.

Cause: When you select a clone role (e.g., in F5-TTS, CosyVoice, GPT-SoVITS), the software attempts to clone the voice from the original video. This requires the system to extract short, clean speech segments as reference audio. If the source video has no clear, isolated speech of sufficient length (3–10 seconds) or the reference clips cannot be properly generated, the "Clone disabled" error is triggered.

Solution:

1. Verify Reference Audio Files: Check if the required reference audio files exist in the f5-tts folder within your software installation directory (F:/win-pyvideotrans-v3.99-428/f5-tts/). For a standard clone, they may be missing or incorrectly named.

2. Adjust Speech Segmentation Settings: For the clone process to work, the extracted audio clips must be between 3 and 10 seconds long. Go to Menu → Tools/Options → Advanced Options → Voice Recognition Parameters and set the following:

   • Minimum voice duration (ms): 3000
   • Maximum voice duration (seconds): 10
   • Check (enable) both Whisper pre-segmented audio and Merge short subtitles to adjacent.
3. Enable Vocal Separation: In the main interface, expand the "More Settings..." panel and check the Separate vocals and background option. This is strongly recommended for a cleaner reference audio sample.
4. Re-run the Task: Before starting again, ensure you check the Clean up generated checkbox on the main interface to force reprocessing from scratch.

请查阅相关文档：

• 如何使用原视频中音色配音
• 语音识别中的VAD参数调整