#2928 Dubbing Failed不存在参考音频，无法克隆:role='clone',wavfile=''

This error occurs because you selected the "clone" role for TTS (voice cloning) but did not provide a reference audio file for the model to clone.

Solution:

1. Check your TTS settings: Go to the "TTS Settings" tab and select your TTS channel (e.g., F5-TTS, Qwen-TTS(local), CosyVoice, GPT-SoVITS).

2. For "clone" role: If you want to use the "clone" role to mimic the original video's voice, you must provide a reference audio file.

   • Method A (Recommended for video translation): In the main translation task interface, under "Advanced Options" -> "TTS Parameters", ensure the "Use original video audio as reference for cloning" option is checked. This will automatically use a segment of the original video's audio.
   • Method B (Manual): In the TTS channel settings, manually select a reference audio file (WAV format, 3-10 seconds of clear speech) in the designated field.
3. If you don't have a reference audio: Do not select the clone role. Instead, choose a specific built-in voice provided by your TTS service (e.g., "Vivian" for Qwen-TTS, or any other named voice for Edge-TTS/F5-TTS).

Important Note for Qwen-TTS(local):

• The clone role requires you to have started the Voice Clone model (Qwen3-TTS-12Hz-0.6B-Base or Qwen3-TTS-12Hz-1.7B-Base). Starting the Custom Voice model or Voice Design model will cause this error when using clone.
• If you are testing in the TTS settings panel and get this error, try clearing the "Reference Audio" text box first, then test. It will use the built-in voice "Vivian" to connect to the Custom Voice model.

Please refer to the relevant documentation:

• 如何使用原视频中音色配音
• qwen-tts(本地)配音渠道遇到的问题 (See the section on Cannot find a function with api_name: /run_voice_clone)

win预打包版请下载这个sp.exe,覆盖旧的sp.exe, sp.exe下载地址