#6930 语音识别阶段出错[faster-whisper(内置)] No transcription results returned. Please check the original audio/video [info.duration_aft

171.91.**4 Posted at: 1 hour ago

语音识别阶段出错[faster-whisper(内置)] No transcription results returned. Please check the original audio/video [info.duration_after_vad=0.0s].
info=TranscriptionInfo(language='ja', language_probability=1, duration=3118.4946875, duration_after_vad=0.0, all_language_probs=None, transcription_options=TranscriptionOptions(beam_size=5, best_of=5, patience=1, length_penalty=1, repetition_penalty=1.5, no_repeat_ngram_size=0, log_prob_threshold=-1.0, no_speech_threshold=0.6, compression_ratio_threshold=1.0, condition_on_previous_text=True, prompt_reset_on_temperature=0.5, temperatures=[0.0], initial_prompt=None, prefix=None, suppress_blank=True, suppress_tokens=(1, 2, 7, 8, 9, 10, 14, 25, 26, 27, 28, 29, 31, 58, 59, 60, 61, 62, 63, 90, 91, 92, 93, 359, 503, 522, 542, 873, 893, 902, 918, 922, 931, 1350, 1853, 1982, 2460, 2627, 3246, 3253, 3268, 3536, 3846, 3961, 4183, 4667, 6585, 6647, 7273, 9061, 9383, 10428, 10929, 11938, 12033, 12331, 12562, 13793, 14157, 14635, 15265, 15618, 16553, 16604, 18362, 18
......
ned. Please check the original audio/video [info.duration_after_vad=0.0s].
info=TranscriptionInfo(language='ja', language_probability=1, duration=3118.4946875, duration_after_vad=0.0, all_language_probs=None, transcription_options=TranscriptionOptions(beam_size=5, best_of=5, patience=1, length_penalty=1, repetition_penalty=1.5, no_repeat_ngram_size=0, log_prob_threshold=-1.0, no_speech_threshold=0.6, compression_ratio_threshold=1.0, condition_on_previous_text=True, prompt_reset_on_temperature=0.5, temperatures=[0.0], initial_prompt=None, prefix=None, suppress_blank=True, suppress_tokens=(1, 2, 7, 8, 9, 10, 14, 25, 26, 27, 28, 29, 31, 58, 59, 60, 61, 62, 63, 90, 91, 92, 93, 359, 503, 522, 542, 873, 893, 902, 918, 922, 931, 1350, 1853, 1982, 2460, 2627, 3246, 3253, 3268, 3536, 3846, 3961, 4183, 4667, 6585, 6647, 7273, 9061, 9383, 10428, 10929, 11938, 12033, 12331, 12562, 13793, 14157, 14635, 15265, 15618, 16553, 16604, 18362, 18956, 20075, 21675, 22520, 26130, 26161, 26435, 28279, 29464, 31650, 32302, 32470, 36865, 42863, 47425, 49870, 50254, 50258, 50359, 50360, 50361, 50362, 50363), without_timestamps=False, max_initial_timestamp=1.0, word_timestamps=True, prepend_punctuations='"\'“¿([{-', append_punctuations='"\'.。,,!!??::”)]}、', multilingual=False, max_new_tokens=None, clip_timestamps=[0.0], hallucination_silence_threshold=None, hotwords=''), vad_options=VadOptions(threshold=0.5, neg_threshold=None, min_speech_duration_ms=0, max_speech_duration_s=inf, min_silence_duration_ms=140, speech_pad_ms=400))
[SpeechToText]语音转录: self.out_format='srt',self.copysrt_rawvideo=True,self.spk_insert=False
[TaskCfgSTT]
原始输入文件名: F:/D/2/honnamatv-399.mp4,
输出结果保存到文件夹: D:/BaiduNetdiskDownload/win-pyvideotrans-v4.11/output/recogn,
临时文件夹: D:/win-pyvideotrans-v4.11/tmp/12720/5bbbe6e475
已启用CUDA加速
未选 降噪
语音识别:faster-whisper(内置), large-v3 断句方式:默认断句
发音语言:ja
未选 降噪


=
system:Windows-10-10.0.19045-SP0
version:v4.11
frozen:True
language:zh
root_dir:D:/win-pyvideotrans-v4.11

1 Replies

#1 DeepSeek reply at 1 hour ago

这是 VAD(语音活动检测)将整段音频判定为“无人声” 导致的错误。日志中 duration_after_vad=0.0 表示:原始音频时长约 3118 秒,但经过 VAD 过滤后,被判定为语音的有效时长是 0 秒,所有片段都被当作静音/噪声剔除,因此 Whisper 识别不到任何内容。

请按以下顺序排查

  1. 先确认原视频能正常听到人声
    用播放器打开 honnamatv-399.mp4,确认不是无音轨、静音或音量极低(特别是背景噪声大时)。
  2. 降低“语音阈值”
    进入 菜单 → 工具/选项 → 高级选项 → 语音识别参数,找到 “语音阈值”,默认值为 0.5,建议改为 0.3 甚至更低的 0.2。阈值越小越灵敏,能避免将正常语音误判为静音。
  3. 勾选“分离人声背景声”
    在主界面 设置更多参数.. 中,勾选 分离人声背景声,让背景噪声被剔除后再送识别。
  4. 检查音频格式是否损坏
    用 FFmpeg 或其他工具将原视频音频提取为 wav 测试是否能正常识别,若提取后仍为 0 秒有效音频,则原视频音轨可能损坏,需要重新获取文件。
  5. 尝试其他识别渠道
    若上述调整后仍失败,可在语音识别渠道中切换为 openai-whisper(内置) 或降低模型大小(如从 large-v3 改为 medium),排除模型兼容性问题。

请查阅相关文档:

Thinking...
Post Your Reply
If you have more than 50 messages, you will no longer receive an automatic reply. Please ask a new question!
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