#6548 No dubbing audio generate, view logs:{'text': 'Привет, мой дорогой друг. Желаю, чтобы каждый твой день был прекрасен и п

This error indicates that the TTS provider selected (tts_type=1) failed to generate audio, likely because the role cosy.wav is not a valid voice for that provider, or the provider/service is not correctly configured.

The log shows the queue item uses role: 'cosy.wav'. If you intended to use a local reference audio file for cloning, note these requirements:

1. The audio file must be a WAV file, 3–10 seconds, with clear single voice and no background noise.
2. For most built-in clone channels, place the file in the software root directory under f5-tts/ (e.g., D:/win-pyvideotrans-v4.07/f5-tts/cosy.wav).
3. For GPT-SoVITS, the reference audio must be placed in the GPT-SoVITS root directory, not in f5-tts/.
4. If cosy.wav is meant to be a voice role, verify you selected the correct TTS provider and that the provider supports this role.

If the file is present and correctly named, then check the selected provider:

• If it is an online API (e.g., CosyVoice, Azure, etc.), confirm the API address/key is correctly filled in Settings -> Translation/TTS settings.
• If it is a local API (e.g., CosyVoice), ensure the local service is actually running and the API URL uses 127.0.0.1 instead of 0.0.0.0.

Also, inspect the detailed errors in logs/ folder (the file named by date, e.g., 20260614.log). Search for the timestamp around the failure to see the underlying API/provider message.

Please check related documentation:

• CosyVoice
• Using reference audio for cloning / TTS errors
• TTS channels overview