#2078 studyfix_merged.mp4No subs recognized. Check Audio/Lang.

2001:171b* Posted at: 9 hours ago 👁8

studyfix_merged.mp4No subs recognized. Check Audio/Lang.

=
system:Windows-10-10.0.26200-SP0
version:v3.90
frozen:False
language:en
root_dir:C:/tools/pyvideotrans
Python: 3.11.9 (tags/v3.11.9:de54cf5, Apr 2 2024, 10:12:12) [MSC v.1938 64 bit (AMD64)]

1 Replies

Floor #1 DeepSeek Posted at 9 hours ago

The error "No subs recognized. Check Audio/Lang" indicates that the speech recognition (ASR) process did not detect any speech in your video file. This is a common issue with several potential causes.

Here are the most likely solutions, in order of probability:

  1. Check the Audio Track: The video file studyfix_merged.mp4 might not contain an audio track with human speech, or the audio is corrupted. Use the Toolbox -> Audio/Video Merge/Separate feature to extract the audio and verify it plays correctly in another media player.
  2. Verify Language Settings: In the main interface, ensure you have selected the correct "Source Language" for the speech in your video. If you are unsure, set it to "auto".

  3. Adjust VAD Sensitivity (Most Common Fix): The Voice Activity Detection (VAD) might be too strict. Go to Menu -> Tools -> Advanced Options.

    • In the "Faster/OpenAI/Whisper" section, lower the "Speech Threshold" value (e.g., from 0.45 to 0.3 or 0.2). This makes the detection more sensitive to quieter speech.
    • You can also try increasing the "Max Speech Duration (seconds)" if the speech segments are very long.
  4. Try a Different Recognition Mode: In the main interface's ASR row, switch from "Overall Recognition" to "Batch Inference". This uses a different segmentation method that might work better for your audio.
  5. Enable Noise Reduction: Before starting the process, click "Set More Parameters" on the main interface and check the "Noise Reduction" option.
  6. Use a Different ASR Model: If using faster-whisper, try a larger model like large-v2 or large-v3 for better accuracy, especially if the audio quality is poor. Alternatively, for Chinese content, try the "Ali FunASR (Local)" model.

If none of these steps work, please provide the complete log file from the C:/tools/pyvideotrans/logs/ directory for further diagnosis.

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