Troubleshooting
Troubleshooting
Section titled “Troubleshooting”Common Issues
Section titled “Common Issues”Transcription is taking too long
- Large files (>1 hour) may take 15-30 minutes
- Check your internet connection
- Very poor audio quality requires more processing
Poor accuracy
- Ensure audio quality is sufficient (clear speech, minimal noise)
- Check that the audio contains clear speech in a supported language
- Background music or noise significantly impacts accuracy
Speakers not identified correctly
- Very similar voices may be confused
- Overlapping speech can cause misattribution
- Short utterances may not be correctly attributed
File won’t upload
- Check file format is supported
- Ensure file is under 500 MB
- Try converting to MP3 format
Error Messages
Section titled “Error Messages”| Error | Cause | Solution |
|---|---|---|
| ”File too large” | Exceeds 500 MB limit | Compress or split the file |
| ”Unsupported format” | File type not recognized | Convert to MP3 or WAV |
| ”Processing failed” | Audio couldn’t be processed | Check audio quality, re-upload |
| ”Insufficient hours” | Usage limit reached | Upgrade subscription or wait for reset |
| ”Audio too short” | The recording is under 10 seconds, usually an empty or ring-only export from a phone system | Check the exported recording; re-export the call if a real recording exists. Retrying will not help |
| ”No speech detected” | The audio contained no usable speech, so the transcript was discarded rather than invented | Confirm the recording is audible and contains speech. Re-uploading the same audio gives the same result |