#6568 Reference audio is empty after silence removal. Try setting preprocess_prompt=False.,Traceback (most recent call last):

70.44.**3 Posted at: 1 day ago

Reference audio is empty after silence removal. Try setting preprocess_prompt=False.,Traceback (most recent call last):
File "videotrans\process\omnivoice_tts.py", line 74, in omnivoice_fun
File "C:\Users\User\Downloads\win-pyvideotrans-v4.09\_internal\torch\utils\_contextlib.py", line 116, in decorate_context

return func(*args, **kwargs)

File "omnivoice\models\omnivoice.py", line 567, in generate

voice_clone_prompt=voice_clone_prompt,

File "omnivoice\models\omnivoice.py", line 959, in _preprocess_all
File "omnivoice\models\omnivoice.py", line 675, in create_voice_clone_prompt

logger.warning(

ValueError: Reference audio is empty after silence removal. Try setting preprocess_prompt=False.

Traceback (most recent call last):
File "videotrans\task\only_one.py", line 106, in run
File "videotrans\task\_stage_dubbing.py", line 24, in dubbing
File "videotrans\task\_stage_dubbing.py", line 100, in _tts
File "videotrans\tts\__init__.py", line 228, in run
File "videotrans\tts\_base.py", line 105, in run
File "videotrans\tts\_omnivoice.py", line 41, in _exec
File "videotrans\configure\base.py", line 282, in _new_process
videotrans.configure.excepts.VideoTransError: Reference audio is empty after silence removal. Try setting preprocess_prompt=False.,Traceback (most recent call last):
File "videotrans\process\omnivoice_tts.py", line 74, in omnivoice_fun
File "C:\Users\User\Downloads\win-pyvideotrans-v4.09\_internal\torch\utils\_contextlib.py", line 116, in decorate_context

return func(*args, **kwargs)

File "omnivoice\models\omnivoice.py", line 567, in generate

voice_clone_prompt=voice_clone_prompt,

File "omnivoice\models\omnivoice.py", line 959, in _preprocess_all
File "omnivoice\models\omnivoice.py", line 675, in create_voice_clone_prompt

logger.warning(

ValueError: Reference audio is empty after silence removal. Try setting preprocess_prompt=False.

cfg=[TaskCfgVTT]当前工作模式: 翻译视频或音频 单视频模式
原始输入文件名: C:/Users/User/Videos/TV Series/TV Series Season 1/TV Series - E01.mkv,
输出结果保存到文件夹: C:/Users/User/Videos/TV Series/_video_out/TV Series Season 1/TV Series - E01-mkv,
临时文件夹: C:/Users/User/Downloads/win-pyvideotrans-v4.09/tmp/12344/c654119a53
未选 清理已存在
已启用CUDA加速
未选 降噪
STT Service:Whisper.cpp(WinBuilt-in), ggml-large-v3-turbo.bin, 发音语言: Portuguese, 断句方式:默认断句
TranslateSrt:Google Translate (Free),原始语言:Portuguese,目标语言:English, 已选 SendSRT
TTS Service:OmniVoice(Built-in), 角色:clone, 配音语言:English, 未选 二次语音识别
音量:+0%, 语速:+0%, 已选 音频加速, 未选 视频慢速
字幕: Soft Subtitles

=
system:Windows-10-10.0.26200-SP0
version:v4.09
frozen:True
language:en
root_dir:C:/Users/User/Downloads/win-pyvideotrans-v4.09

3 Replies

#1 DeepSeek reply at 1 day ago

This error occurs because OmniVoice, in clone mode, extracts a short reference-audio clip from the original video for each subtitle line. After OmniVoice internally removes silence/noise from that clip, the clip became completely empty — the extracted segment was too short, contained only background sound, or was composed mostly of silence.

How to fix

  1. Enable Whisper audio pre-segmentation (required for clone mode)
    Menu → Tools/Options → Advanced Options → Speech Recognition Parameters:

    • Enable Whisper pre-segment audio?
    • Set Longest speech duration (seconds) to 6–10
    • Set Shortest speech duration (milliseconds) to 3000–4000
    • Enable Merge too-short subtitles into adjacent

  2. Improve reference-audio quality
    In the main interface, click More Settings, then enable:

    • Separate vocals and background
    • Optionally also enable Noise reduction

    This gives OmniVoice a clean vocal-only reference clip, which drastically reduces the chance of an empty reference after silence removal.

  3. Verify the source audio
    Make sure the original video actually contains clear, continuous human speech at the subtitle timestamps. If the segment contains only music/noise, it will be filtered out as "silence" and produce this exact error.

Note

The error text "Try setting preprocess_prompt=False" is an OmniVoice-internal option. In the built-in OmniVoice channel you cannot set it directly; the correct path is to adjust the reference clip parameters above. If the problem persists, send the latest log from the logs/ folder (last ~30 lines) for further analysis.

Please also consult:

#2 70.44.**3 reply at 7 hours ago

Separate vocals and background
Optionally also enable Noise reduction

These two options fixed it for me. Please note that advanced options need to be enabled to see them. Please close the ticket!

#3 DeepSeek reply at 7 hours ago

感谢反馈。解决方案确定为：勾选「分离人声背景声」和「降噪」后即可规避该报错。补充说明：若主界面未显示这些选项，需先在 菜单 → 工具/选项 → 高级选项 → 通用设置 中勾选「主界面显示所有参数？」，返回主界面即可看到。问题已关闭。

Thinking...
Post Your Reply
Open source and free maintenance is not easy. If this project is helpful to you, please consider making a small donation to help the project continue to maintain and update.

Dondate via ko-fi.com

Related resource