#2443 self.proxy_str='http://127.0.0.1:7897',self.uuid=None2026-01-03 12:27:24,932 - INFO - 最终配置信息：self.cfg=TaskCfg(cache_fold

64.120* Posted at: 2 hours ago 👁5

self.proxy_str='http://127.0.0.1:7897',self.uuid=None
2026-01-03 12:27:24,932 - INFO - 最终配置信息：self.cfg=TaskCfg(cache_folder='D:/py/win-pyvideotrans-v3.92/tmp/3940/fe6a7875ee', target_dir='C:/Users/16226/Desktop/video/How to Create Melee Combat in Godot 4, #2： Destroy enemies ｜ ARPG in Godot 4 ｜ tutorial ｜ zeldalike-mp4/#2-mp4', remove_noise=False, is_separate=False, detect_language='en', subtitle_language=None, source_language='英语', target_language='简体中文', source_language_code='en', target_language_code='zh-cn', source_sub='C:/Users/16226/Desktop/video/How to Create Melee Combat in Godot 4, #2： Destroy enemies ｜ ARPG in Godot 4 ｜ tutorial ｜ zeldalike-mp4/#2-mp4/en.srt', target_sub='C:/Users/16226/Desktop/video/How to Create Melee Combat in Godot 4, #2： Destroy enemies ｜ ARPG in Godot 4 ｜ tutorial ｜ zeldalike-mp4/#2-mp4/zh-cn.srt', source_wav='D:/py/win-pyvideotrans-v3.92/tmp/3940/fe6a7875ee/en.wav', source_wav_output='C:/Users/16226/Desktop/video/How to Create Melee Combat in Godot 4,
......
, [09:28.352 -> 09:30.944], [09:32.704 -> 09:34.272], [09:34.624 -> 09:37.376], [09:39.488 -> 09:40.800], [09:41.152 -> 09:43.200], [09:47.552 -> 09:51.264], [09:53.120 -> 09:54.592], [09:55.200 -> 09:56.896], [09:57.536 -> 09:59.744], [10:00.896 -> 10:02.784], [10:03.264 -> 10:05.408], [10:05.696 -> 10:07.840], [10:09.152 -> 10:10.496], [10:11.200 -> 10:14.112], [10:20.512 -> 10:21.440], [10:21.888 -> 10:26.784], [10:27.680 -> 10:31.584], [10:34.112 -> 10:36.512], [10:37.216 -> 10:37.504], [10:38.112 -> 10:38.720], [10:39.104 -> 10:40.448], [10:41.024 -> 10:41.472], [10:41.856 -> 10:45.120], [10:45.856 -> 10:47.200], [10:47.584 -> 10:49.280], [10:50.496 -> 10:51.008], [10:51.424 -> 10:53.184], [10:54.816 -> 10:59.200], [11:00.768 -> 11:03.232], [11:03.776 -> 11:04.320], [11:04.928 -> 11:06.592], [11:07.168 -> 11:09.248], [11:10.720 -> 11:11.968], [11:15.776 -> 11:16.384], [11:16.960 -> 11:18.944], [11:21.024 -> 11:23.136], [11:23.616 -> 11:24.960], [11:25.696 -> 11:28.192], [11:28.608 -> 11:31.040], [11:32.224 -> 11:34.752], [11:35.328 -> 11:38.240], [11:39.552 -> 11:42.656], [11:47.264 -> 11:48.640], [11:49.024 -> 11:51.488], [11:51.776 -> 11:53.088], [11:55.200 -> 11:58.400], [11:59.712 -> 12:01.984], [12:02.944 -> 12:04.864], [12:06.048 -> 12:07.168], [12:07.520 -> 12:09.664], [12:10.240 -> 12:12.736], [12:13.760 -> 12:15.200], [12:16.672 -> 12:19.264], [12:20.128 -> 12:23.584], [12:24.800 -> 12:25.216], [12:25.600 -> 12:27.072], [12:27.584 -> 12:28.384], [12:28.672 -> 12:30.848], [12:31.328 -> 12:32.768], [12:36.448 -> 12:39.680], [12:40.512 -> 12:41.216], [12:42.336 -> 12:44.256], [12:44.992 -> 12:46.368], [12:46.784 -> 12:48.576], [12:49.952 -> 12:55.584], [12:57.504 -> 12:59.232], [12:59.776 -> 13:02.848], [13:03.360 -> 13:08.384], [13:08.576 -> 13:11.072], [13:12.160 -> 13:16.736], [13:18.176 -> 13:19.840], [13:20.992 -> 13:21.248]
2026-01-03 12:27:36,563 - DEBUG - Processing segment at 00:00.000
2026-01-03 12:27:37,149 - DEBUG - Log probability threshold is not met with temperature 0.0 (-1.358847 < -1.000000)
2026-01-03 12:27:37,245 - DEBUG - Log probability threshold is not met with temperature 0.2 (-1.358847 < -1.000000)
2026-01-03 12:27:37,323 - DEBUG - Log probability threshold is not met with temperature 0.4 (-2.065878 < -1.000000)
2026-01-03 12:27:37,395 - DEBUG - Log probability threshold is not met with temperature 0.6 (-1.506457 < -1.000000)
2026-01-03 12:27:37,477 - DEBUG - Log probability threshold is not met with temperature 0.8 (-3.251741 < -1.000000)
2026-01-03 12:27:37,574 - DEBUG - Log probability threshold is not met with temperature 1.0 (-1.851574 < -1.000000)
2026-01-03 12:27:37,616 - DEBUG - Processing segment at 00:30.000
2026-01-03 12:27:38,070 - DEBUG - Processing segment at 01:00.000
2026-01-03 12:27:38,523 - DEBUG - Processing segment at 01:30.000
2026-01-03 12:27:38,974 - DEBUG - Processing segment at 02:00.000
2026-01-03 12:27:39,426 - DEBUG - Processing segment at 02:30.000
2026-01-03 12:27:39,881 - DEBUG - Processing segment at 03:00.000
2026-01-03 12:27:40,332 - DEBUG - Processing segment at 03:30.000
2026-01-03 12:27:40,788 - DEBUG - Processing segment at 04:00.000
2026-01-03 12:27:41,241 - DEBUG - Processing segment at 04:30.000
2026-01-03 12:27:41,694 - DEBUG - Processing segment at 05:00.000
2026-01-03 12:27:42,146 - DEBUG - Processing segment at 05:30.000
2026-01-03 12:27:42,598 - DEBUG - Processing segment at 06:00.000
2026-01-03 12:27:43,048 - DEBUG - Processing segment at 06:30.000
2026-01-03 12:27:43,503 - DEBUG - Processing segment at 07:00.000
2026-01-03 12:27:43,953 - DEBUG - Processing segment at 07:30.000
2026-01-03 12:27:44,405 - DEBUG - Processing segment at 08:00.000
2026-01-03 12:27:44,857 - DEBUG - Processing segment at 08:30.000
Traceback (most recent call last):
File "videotrans\task\_only_one.py", line 43, in run
File "videotrans\task\trans_create.py", line 379, in recogn
RuntimeError: #2.mp4没有识别出字幕，请检查是否包含人类说话声音，以及说话语言是否和你选择的原始语言匹配

1 Replies

Floor #1 DeepSeek Posted at 2 hours ago

日志显示视频 #2.mp4 在语音识别阶段失败，提示“没有识别出字幕”。这通常由以下原因导致：

  1. 视频本身不含有效人声：视频可能只有背景音乐、音效或环境噪音，没有清晰的说话声。
  2. 语音识别模型选择不当：您选择了 small.en 模型，这是一个仅针对英语优化的较小模型。如果视频中说话者的口音较重、语速过快或背景噪音较大，该模型可能无法有效识别。
  3. VAD（语音活动检测）过滤过度：日志显示 VAD filter removed 04:34.059 of audio，这意味着有超过4分钟的音频被判定为静音或噪音而过滤掉了。如果视频中人声片段本身较短或间隔较长，可能被误过滤。

请按以下步骤排查和解决：

第一步：检查视频内容

  • 使用播放器打开原视频，确认其中包含清晰可辨的人类说话声。

第二步：调整语音识别设置

  1. 在主界面 第2行“语音识别” 区域：

    • 更换模型：将模型从 small.en 更换为更强大的 large-v2 或 large-v3。如果显存不足，可先尝试 medium。
    • 调整识别模式：如果当前是“整体识别”，可尝试切换到“批量推理”模式，有时对复杂音频效果更好。

  2. 在 菜单 -> 工具 -> 高级选项 中：

    • 调整VAD参数：在“Faster/OpenAI/Whisper 语音识别微调”区域，尝试将 语音阈值 从默认的 0.45 调低（例如 0.35），使检测更灵敏。同时可将 静音分割持续毫秒 适当调大（例如 200），减少过度切割。
    • 降低识别要求：将 识别准确度 (beam_size / best_of) 暂时都设为 1，以降低计算复杂度和显存占用。

第三步：确保语言设置正确

  • 在主界面 第3行“翻译设置”，确认 发音语言 已正确设置为视频的实际语言（例如“英语”）。如果不确定，可先设为 auto 让软件自动检测。

第四步：尝试降噪（如果背景噪音大）

  • 在主界面 第7行“快速参数”，勾选 降噪 选项，这可能会提升语音识别的清晰度。

如果以上步骤尝试后问题依旧，请提供更完整的日志文件（位于软件目录下的 logs 文件夹中），以便进一步分析。

请查看更多参考文档：

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.

Donate: https://ko-fi.com/jianchang512

Trending Questions