#2961 RuntimeError: No speech detected. Ensure audio/video contains speech in the selected language.

This error occurs because the faster-whisper model did not detect any speech in your audio/video file. The most common reasons are incorrect language settings, low audio volume, or background noise interference.

Solution Steps:

1. Check the "Source Language" setting: Do not rely on "Auto". Manually select the correct language spoken in your video (e.g., "Chinese" or "English").
2. Verify the audio content: Ensure your video file actually contains clear speech. You can try playing it in another media player.

3. Adjust VAD (Voice Activity Detection) settings:

   • Go to Menu -> Tools -> Advanced Options -> Speech Recognition Parameters.
   • Lower the "Speech Threshold" value (e.g., from 0.5 to 0.3). This makes the model more sensitive to potential speech.
   • Ensure "VAD Model" is set to ten-vad or silero.

4. Pre-process the audio (if needed):

   • If there is loud background music or noise, use the "Separate" tool (Menu -> Tools -> Separate) to isolate the vocals before recognition.
   • Alternatively, enable "Enable Noise Reduction" in the Advanced Options.

If the issue persists after these steps, please provide the log file from the logs/ folder in your software directory (copy the bottom ~30 lines of the error log) for further analysis.

Please refer to the relevant documentation:

• 语音识别中的VAD参数调整
• 语音降噪方法