• #7764 pyVideoTrans 视频日语语音翻译成字幕 字幕能让字幕和语音同步，说话时有字幕，说话结束字幕也结束...

  💬 1
  [Last reply:21 days ago] 字幕与语音不同步的底层原因是：日语翻译成中文后，句子音节数量和朗读时长必然发生变化，配音/字幕时间轴与原字幕时间轴产生偏差。针对你「说话时字幕出现、说话结束字幕结束」的需求，解决方案如下：1. 调整 VAD 切分参数（让字幕边界贴合说话起止）进入 菜单 -> 工具/选项 -> 高级选项 -> 语音识别参数，调整以下三项：最长语音持续(秒)：设为 5 或更小最短语音持续(毫秒)：设...
  111.31**4 Posted at: 21 days ago

• #7763 怎么能让字幕说话时有 说话结束字幕就没了...

  💬 1
  [Last reply:21 days ago] 无法确定具体原因。请提供软件安装目录下的 logs/ 文件夹中最新的日志文件（复制底部约30行报错信息）以便分析。同时请补充说明：你使用的是哪个功能（视频翻译/语音转录/文字配音）、想调整的是「生成的字幕时间轴」还是「已有字幕文件」、以及当前软件版本号（主界面左上角）。...
  111.31**4 Posted at: 21 days ago

• #7762 Traceback (most recent call last): File...

  💬 1
  [Last reply:21 days ago] 这是代码包文件缺失/版本不一致导致的：报错 No module named 'videotrans.recognition._whisperxapi' 表示当前软件内的渠道实现文件 _whisperxapi.py 不存在，而 v4.14 的渠道注册表仍在调用它。解决方案你使用的是 WhisperX(本地API) 渠道，该渠道需要你自己另行部署，且当前版本安装包内该模块缺失或未覆...
  39.68**1 Posted at: 21 days ago

• #7761 pyvideotrans（v4.14）当前版本: v4.14 语音识别下来框里 怎么没有parakeet 日语 (内置)...

  💬 1
  [Last reply:21 days ago] parakeet日语(内置) 渠道已在 v4.12 版本中被移除，替换为 Nemotron-3.5-asr-0.6b(内置) 模型，因此 v4.14 语音识别下拉框中不会再有该渠道，属于正常变更。解决方法：若要识别日语，请改用 Nemotron-3.5-asr-0.6b(内置) 渠道（支持日语、韩语、越南语及欧洲系约 40 种语言）。或在 Huggingface_ASR(内置) 渠道中选择日语专用...
  111.31**4 Posted at: 21 days ago

• #7760 太慢了，翻译和配音...

  💬 1
  [Last reply:22 days ago] 翻译和配音阶段的耗时由所选的渠道类型和是否为本地模型决定，与语音识别阶段的 GPU 加速无关。核心原因：字幕翻译和配音主要消耗网络请求（在线 API）或 CPU/GPU 推理（本地模型）。若使用的是本地 TTS 模型（如 F5-TTS、Qwen3-TTS、ChatterBox 等），且未启用 CUDA 加速，速度会非常慢；若使用在线 API（Edge-TTS、DeepSeek 等），则取决于平台响...
  39.68**1 Posted at: 22 days ago

• #7759 [DeepSeek] API service returned error: Insufficient balance.Please check account balance on provider platform. API: https://api.deepseek.com/v1/...

  💬 1
  [Last reply:22 days ago] Your DeepSeek API account has no remaining balance. The error code 402 - Insufficient Balance returned by https://api.deepseek.com/v1/ means the API key is valid but the account has run out of credits...
  90.1**0 Posted at: 22 days ago

• #7758 1.mp4No subtitles recognized. Check: [1] Media contains clear human speech, [2] Selected spoken language matches audio, [3] Selected engine supports spoken language.Traceback (most...

  💬 1
  [Last reply:22 days ago] The error means the ASR engine returned no subtitle lines. In your cfg the mismatch is the main cause: you selected spoken language English, but the ASR channel Dolphin Asian (Built-in) is designed fo...
  2a01**1 Posted at: 22 days ago

• #7757 Subprocess Error: cannot import name 'TensorSpec' from 'safetensors._safetensors_rust' (D:\Videotrans\_internal\safetensors\_safetensors_rust.cp310-win_amd64.pyd)Traceback (most re...

  💬 1
  [Last reply:22 days ago] This is caused by an incomplete or mismatched upgrade — the safetensors library bundled in your _internal folder is outdated/partial, so the new Qwen3-TTS code cannot import TensorSpec (a symbol that ...
  37.52**0 Posted at: 22 days ago

• #7756 Authentication Fails, Your api key: ****4-UA is invalid (request_id: ac18c29e-ba09-4ef9-a6d0-079db71d9e0f)https://api.deepseek.com/v1/...

  💬 1
  [Last reply:22 days ago] This is a DeepSeek API authentication failure (HTTP 401) — the API key you entered is invalid or not recognized by DeepSeek's server. The error is returned directly by https://api.deepseek.com/v1/, no...
  90.1**0 Posted at: 22 days ago

• #7755 Authentication Fails, Your api key: ****4-UA is invalid (request_id: 7e669a45-a57f-428f-8fb5-500e9eae2b01)https://api.deepseek.com/v1/...

  💬 1
  [Last reply:22 days ago] The error is a 401 Authentication failure from DeepSeek's API — your DeepSeek API key is invalid, incorrectly entered, or was revoked. The traceback shows AuthenticationError: Error code: 401 - ... Yo...
  90.1**0 Posted at: 22 days ago

• #7754 Subprocess Error: cannot import name 'TensorSpec' from 'safetensors._safetensors_rust' (D:\Videotrans\_internal\safetensors\_safetensors_rust.cp310-win_amd64.pyd)Traceback (most re...

  💬 1
  [Last reply:22 days ago] The error cannot import name 'TensorSpec' from 'safetensors._safetensors_rust' indicates a version mismatch in the safetensors library — typically caused by an incomplete or partia...
  37.52**0 Posted at: 22 days ago

• #7753 429 Client Error: Too Many Requests for url: https://www.google.com/sorry/index?continue=https://translate.google.com/m%3Fsl%3Den%26tl%3Dfr%26hl%3Dfr%26q%3DHello%2520guys%250AI%252...

  💬 1
  [Last reply:22 days ago] Google翻译渠道返回429，是因为当前IP被Google判定为流量异常（通常是VPN节点不干净），或请求过于频繁。两种可行处置方式：更换翻译渠道（推荐，最稳定）菜单 → 翻译设置 → DeepSeek，填入API Key（可充值1元）主界面翻译渠道下拉改为 DeepSeek，重新翻译或改用本地 M2M100(内置)，首次使用会自动下载模型继续使用Google翻译的话更换更干净的VPN节点（IP...
  90.1**0 Posted at: 22 days ago

• #7752 429 Client Error: Too Many Requests for url: https://www.google.com/sorry/index?continue=https://translate.google.com/m%3Fsl%3Den%26tl%3Dfr%26hl%3Dfr%26q%3DHello%2520guys%250AI%252...

  💬 1
  [Last reply:22 days ago] The 429 error is Google Translate's rate-limiting / IP-blocking, not a bug in pyVideoTrans.Core causes:If you are using a VPN/proxy to reach Google, your node IP is likely flagged as "abnormal traffic...
  90.1**0 Posted at: 22 days ago

• #7751 stealth/space-bunny-alpha is not a valid model IDhttps://openrouter.ai/api/v1...

  💬 1
  [Last reply:22 days ago] The core cause: OpenRouter rejects the request because the model ID stealth/space-bunny-alpha is invalid. Note that in the error payload the model value is ' stealth/space-bunny-alpha' — the...
  37.52**0 Posted at: 22 days ago

• #7750 语音识别阶段出错[faster-whisper(本地内置)] No transcription results returned. Please check the original audio/video or model and try again.info=TranscriptionInfo(language='en', language_probab...

  💬 1
  [Last reply:22 days ago] 核心原因是 VAD 语音活动检测后有效语音时长为 0（duration_after_vad=0.0），即整段音频未检测到符合阈值的人类语音，因此返回空转录结果。常见原因及解决办法：视频本身无有效人声：该文件为音乐教程/Audiomodern 演示视频，可能全程只有背景音乐、鼓点或纯音乐，无清晰讲解人声。请先本地播放确认是否有人声说话。背景音乐过大掩盖人声：主界面点击设置更多参数，勾选分离人声背景声...
  125.67**6 Posted at: 22 days ago

• #7749 Secure connection failed, check system time or network settingsTraceback (most recent call last):...

  💬 1
  [Last reply:22 days ago] 这是 SSL 握手被中断导致的连接失败，底层原因是你的网络无法与 open.bigmodel.cn（智谱AI）建立安全的 HTTPS 连接，并非软件本身的 BUG。常见原因及按优先级排查：代理/VPN 干扰（最常见）智谱AI 是国内服务，不需要走代理，但如果你主界面填写了代理地址，且代理出口在境外或不稳定，会导致 SSL 握手被中断。操作：清空主界面 对齐控制 行右侧的网络代理文本框，保存后重试。...
  2602**84e Posted at: 22 days ago

• #7748 配音阶段出错 [302.AI] {...

  💬 5
  [Last reply:22 days ago] 不是。付费的必要性取决于你选择的配音渠道，绝大多数情况下可以完全免费。免费方案（推荐）：Edge-TTS(免费)：微软免费接口，中文配音效果自然，零成本，适合绝大多数场景。内置本地模型：Qwen3-TTS(内置)、OmniVoice-TTS(内置)、F5-TTS(内置)、ZipVoice(内置)、Confucius-TTS(内置) 等，全部免费，仅需首次在线下载模型（几百MB~几GB），本地运行不...
  2a03**513 Posted at: 22 days ago

• #7747 配音阶段出错 [302.AI] {...

  💬 1
  [Last reply:22 days ago] 错误为 302.AI 返回 HTTP 401，表示身份认证失败，通常是 API Key 未填写、填写错误、已失效或账户无权限/余额不足。排查步骤：打开 菜单(Menu) -> 翻译设置/配音设置 -> 302.AI，确认 API Key（appkey）已正确填写，无多余空格。登录 302.ai 官网检查该 Key 是否有效、账户是否有余额、是否具备语音合成（TTS）权限。若刚修改过 K...
  2a03**513 Posted at: 22 days ago

• #7746 Invalid API Key, for details please view 302.AIhttps://api.302.ai/v1/...

  💬 1
  [Last reply:22 days ago] 这是 302.AI 平台返回的 401 认证失败，即你填写的 302.AI API Key 无效（错误码 -10002），与软件本身无关。解决方案：登录 302.AI 官网，进入控制台确认 API Key 是否正确、是否已过期或被重置；确认该 Key 对应的账户有余额（余额为 0 时也常返回类似无效报错）；打开软件 菜单 → 翻译设置 → 302.AI，重新粘贴完整的 API Key（注意不要有多...
  2a03**513 Posted at: 22 days ago

• #7745 ASR Error [FunASR(Local)] Traceback (most recent call last): File...

  💬 1
  [Last reply:22 days ago] This error means the FunASR (paraformer-zh) engine returned a result dict that does not contain the key sentence_info, so the recognition step failed inside stt_fun.py.Possible causes:The paraformer-z...
  202.150**5 Posted at: 22 days ago

Related resource

• Document
• Download
• GitHub repo
• My Post