Transcribe oral histories and archival recordings
Transcribe older, noisy or archival audio into text with timestamps, editable corrections and a preserved original for editorial notes.
Availability depends on the configured speech provider.
Guest preview: up to 50 MB and 10m. Results are kept for a limited time unless you create an account.
Archival recordings bring old microphones, tape hiss and vocabulary that modern models rarely hear. VoqText gives you a starting transcript, keeps the untouched provider output, and lets you annotate corrections paragraph by paragraph.
Dedicated uncertain-word markers and editorial note fields are planned; today, edited paragraphs are flagged and the original is one hover away.
What you get
- Timestamped transcript with the provider original preserved
- Edit flags showing where a human corrected the text
- JSON export carrying timestamps for archival systems
- Checkpoints to preserve reviewed states
Who it is for
- Oral history projects
- Museum and library audio collections
- Family and community archives
Frequently asked questions
How does it cope with poor audio?
Expect more errors than with modern recordings. Cleaning the audio beforehand helps; the editor is built for fast correction.
Can I mark words as uncertain?
Marker syntax and editorial notes are planned. Until then, use a convention such as [unclear] in the text; it is preserved in exports.
Which languages are supported for older recordings?
The same provider language list applies; dialects and historical speech may be recognized less accurately.