Check Your Order Status
Enter the Job ID from your confirmation to check the status of your transcription.
Contact Us
Common Questions
How long does transcription take?
Most files are processed within minutes. Longer files (over 1 hour) may take up to 15 minutes. You'll be redirected to a status page where you can monitor progress.
What file formats are supported?
MP3, MP4, WAV, MOV, AVI, MKV, FLAC, OGG, AAC, M4A, WMA, AMR, MPEG, WMV, FLV, WEBM, 3GP, M4V, and more — up to 500 MB per file.
What is the SHA-256 hash?
A cryptographic fingerprint computed from your original file in your browser before upload. It appears on every page of your deliverables to prove the file was not modified during transmission or processing. You can re-compute it at any time to verify your file matches.
What does "not legible" mean?
When our transcription engine is less than 50% confident in a segment, it's marked as "not legible" rather than guessing. This means the audio in that section may contain background noise, overlapping speech, or distortion that prevented reliable transcription. We'd rather flag uncertainty than present guesses as fact.
What are the confidence percentages?
Every sentence in your transcript carries a confidence score from our processing engine. 85% and above is high confidence, 50–84% is low confidence (flagged for review), and below 50% is marked "not legible." These are estimates — even high-confidence segments should be verified for critical use.
What is the retrieval code?
When your deliverables are ready, you receive a unique retrieval code. This code is the only way to access your encrypted archive, which is stored for a minimum of one year. Store it securely — lost codes cannot be recovered or regenerated.
How long are my files stored?
Your source audio or video is permanently deleted the moment transcription completes — we never retain your original media. Your output deliverables (transcript, segments, PDF, hash file) are hashed, encrypted, and archived for a minimum of one year. After the retention period, archived deliverables are permanently deleted.
How does speaker diarization work?
Our engine automatically detects and separates different voices in your recording. Each speaker is labeled (Speaker 1, Speaker 2, etc.) with color-coded output. Speaker labels are consistent throughout the transcript — the same voice gets the same label every time it appears.
Can I get a refund?
If processing fails or produces no output, you will not be charged. For quality concerns, contact us within 7 days with your Job ID and a description of the issue. Refunds are evaluated on a case-by-case basis. See our full terms for details.
Is my transcription admissible in court?
Perceptive Vision provides automated transcription with cryptographic integrity verification — not a certified legal record. Admissibility depends on your jurisdiction and the specific legal context. The SHA-256 hash and chain-of-custody metadata can support authenticity arguments, but all output should be reviewed by a qualified professional before use in any legal proceeding. See our legal page for full disclaimers.