Audio transcription, from the recording to the text
What audio transcription does
Audio transcription turns recorded speech into written text. You upload an audio file, Kotosay detects the language and writes down what was said, and the transcript opens in the editor with a timestamp on each line.
Speech to text and voice recording transcription are other names for the same job. Turning an audio file to text works the same way for a short voice memo as for a long interview.
Getting a recording off your phone
You can transcribe voice recordings right on the phone you made them with. Upload from its browser, for example from the Files app on an iPhone or the file manager on Android.
- iPhone Voice Memos. Open the recording, tap the More button, then choose Share or Save to Files. The file is an M4A.
- Android. Recorder apps differ from one phone maker to the next. Most have a Share option on each recording, and the files often sit in a Recordings folder.
- Call recordings. Where they end up depends on the phone and the app, so start with the app's own share menu.
- Voice messages from chat apps. Save or export the message first. These are often OGG or OPUS files, and they upload as they are.
Calls, interviews and podcasts
For an interview or a call, the timestamps lead you back to every answer, so you can check a quote against your own recording before you use it. Several people talking at once lower the accuracy, so ask people to take turns.
Laws on recording calls differ by country and state, so ask before you record. For a podcast, the transcript gives you text for show notes or a page per episode.
What makes a recording transcribe well
A clean recording gives a clean transcript. These habits help while you record.
- Keep the phone or microphone close to the person talking. A phone on the table in front of the speaker beats a better microphone across the room.
- Switch off music, the TV and any fan you can. Music and other voices under the speech cover words.
- Ask people to let each other finish. When two voices overlap, words from both tend to go missing.
- On a call, use the recording the app makes itself. A second phone held near a speakerphone picks up the room as well.
TipHave a noisy recording already? Upload it anyway and read the first 5 minutes, which are free. Fix what came out wrong in the editor, and check names and numbers against your own recording.
After the transcript
Every transcript opens in the editor, where you can fix words while the timestamps stay in place, then copy the text. With Unlimited you can also download TXT, DOCX, PDF, SRT or VTT and translate the text into another language.
If a recording causes trouble, write to support. We answer within 24 hours.