How to Transcribe Recordings to Text: Meetings, Interviews and Lectures

Save hours of typing

Typing out a one-hour meeting by hand usually takes three to four hours. With an AI draft, you only fix the mistakes. For recordings that need to become text — transcripts, minutes, interview articles, lecture notes — this is by far the most efficient route.

Transcribing with AI

1. In Speech to Text, choose the Spoken language.
2. Add a phone voice memo (M4A), an MP3, or a Zoom/Teams recording as is.
3. The transcript comes out one sentence per line. Edit, then Copy it into your document or Save it as TXT.

Need timestamps? Choose SRT as the format and each line gets a time code, making it easy to check against the recording.

Record for better accuracy

• Keep the mic close: put the phone in the middle of the table. Even a metre away hurts accuracy.
• Choose a quiet room: air conditioning, café music and typing add errors.
• One speaker at a time: overlapping voices are the hardest part for AI.

If the recording is already noisy, run it through Noise Reduction first; it can improve the result.

Turning the transcript into a document

• Skim through once and fix just names, numbers and jargon.
• For meeting minutes, keep the raw transcript separately and summarise decisions and action items.
• Unclear passages are easier to check slowed down to about 0.75× with the Audio Speed Changer.

Consent and privacy

Tell people you're recording and get their agreement first. Recording laws differ by country and state, and secretly recording a conversation you're not part of can be illegal. Recordings and transcripts contain personal information, so share them carefully.

FAQ