Skip to content

Old recordings

Transcribe old recordings: cassettes and oral histories

A shoebox of cassettes, a grandparent's interview, an oral-history project: the sound is worth keeping and the text is what makes it searchable. The first step is not here: the tape has to become a digital file. Once it is one, upload it and read it.

Free preview: one file up to 15 minutes and 25 MB, 3 tries a day, no account needed.

Want to keep your transcripts? Create a free account

Step one: make the tape a file

This site transcribes files, not tapes. A cassette needs to be played into a computer and recorded as an audio file. The usual ways are a cassette player with a USB connection, or an ordinary player with a line output connected to the computer's line input or a USB audio adapter. Free audio editors such as Audacity can record that input and save it as WAV or MP3. Record each side as its own file, and keep the WAV as your archive copy: a transcript can always be made again, a tape cannot be re-recorded.

Step two: upload and read

  1. Choose the digitised file in the box above. WAV, MP3 and M4A all work, and the formats page lists the rest.
  2. Set the spoken language if the speakers are not using English. Dialects and old-fashioned vocabulary are where a transcript struggles most, so leave extra time to check.
  3. Listen while you read. Correct names, places and dates first: they are what a family or an archive will search for.
  4. Export text for reading, or timestamped text to keep each passage tied to its place on the tape.

What old recordings do to a transcript

  • Tape hiss and a quiet voice: raising the volume in an editor before you upload can help a little.
  • Wow and flutter, where the tape speed wobbles, makes voices sound unsteady; it limits accuracy and is not something a transcript can fix.
  • A microphone far from the speaker, a room with echo, a television or a kettle in the background.
  • Two or more people at once, which is normal in a family interview and hard for any transcript.
  • Dialect, an accent the language model rarely sees, and words that have fallen out of use.

Oral-history practice: permission and records

An interview is the speaker's story, and projects that collect them usually get written permission for how the recording and the text may be used. Keep the signed form with the file. Uploading a recording sends it to our server, so use it only for material you may share that way; the privacy page says how long files are kept, and the self-hosted option is for recordings that must stay on your own machine. Treat the transcript as a working draft: if the project publishes it, the speaker or a person who knew them should review it.

Long tapes

Without an account one file can be up to 15 minutes and 25 MB. A free account lifts those per-file limits and keeps your transcripts in a library, so for a longer recording create one instead of cutting the file up. A C-90 cassette is 45 minutes a side, so a free account is the practical route, or you can record each side in two or three files.

How it works

  1. Add a recording

    Drop in a video or audio file, choose one from your device, or record from the microphone. Nothing to install.

  2. It transcribes

    Speech becomes timestamped text, with speaker labels when more than one person talks.

  3. Read, fix, export

    Click any line to correct it, then download TXT, subtitles, Markdown, JSON or CSV.

Questions

Can you digitise my tapes?
No. Scriptorium transcribes audio files. You need to record the tape into a computer first.
Will it understand a strong dialect or an old language form?
It will try and will get some of it wrong. Read the transcript against the audio, and expect to correct more than for modern, clear speech.
Can I name the people speaking?
Yes, speaker labels cover up to eight people with a free account. Who is who is decided by you, not guessed.
Is the transcript good enough to archive?
Keep the audio as the record and the transcript as a finding aid. Check it before you rely on it.
Is it free?
Yes — one recording up to 15 minutes and 25 MB with no account, up to 3 tries a day. A free account keeps your transcripts in a library and lifts the guest limits; the current limits and paid plans are on the pricing page.
What can I export?
Plain text, timestamped text, SRT, VTT, Markdown, JSON with word timings and CSV, all from the same edited transcript.

Want your transcripts saved and longer files? Create a free account · See pricing · All use cases

Try it with your file

No account needed for a first transcript.

More pages like this