Docs
🎞️ Audio & Video File Transcription
Offline mode allows you to transcribe audio or video files from your local disk. All processing is done locally, ensuring privacy and security.
1. Supported Media Formats
Owl Meeting features strong file compatibility, capable of handling various recording and video formats:
- Media Support: MP3, WAV, M4A, MP4, MKV, MOV, and almost all other mainstream formats.
- Built-in Converter: You can use the Format Conversion tool within the interface to prepare files.
2. Segmentation Methods
Proper audio segmentation is the foundation of efficient transcript organization. Choose the split strategy that best fits your content:
- Time Interval: Automatic segmentation based on Voice Activity Detection (VAD). Best for speeches, podcasts, or solo narrations.
- Speaker Diarization: Split based on different speaker voice characteristics. Ideal for
meetings or interviews.
- Speaker Labeling: Automatically assign identity tags to different segments, supporting quick subsequent edits.
3. Smart Mode
In this mode, you can assign specific recognition models to different speakers for targeted identification, which greatly improves speed and accuracy.
By using the most matching model for different languages or accents, you can effectively tackle complex multi-speaker dialog scenarios.
4. Test Mode
Randomly extract 3-minute samples from long audio files for recognition. Quickly preview effects and dynamically adjust parameters or models based on test results.
5. Performance
Blazing Fast CPU Inference: Thanks to a deeply optimized engine, you can achieve high-speed transcription even on regular PC CPUs:
- i5-11400H (5 years old CPU): 30 minutes of audio processed in about 1 minute.
- i5-4210m (10 years old CPU): 30 minutes of audio processed in about 3 minutes.