An interview transcript has one job most other transcripts do not: it has to survive being quoted. A meeting summary can be a little loose. An interview transcript that misquotes someone becomes your problem, not the AI's. This guide walks the actual steps for turning a recorded interview into a clean, speaker-labeled transcript you can trust for writing, research, or archiving.
If you only need the platform page, use interview recordings. If you want the diarization background first, see what is speaker diarization.
Before you start: what you need
- The recording file: usually WAV from a field recorder, M4A from a phone voice memo, or MP4 from a Zoom or video call
- Consent already handled. If you recorded a call, the other party should know
- A rough sense of speaker count (one source, two people, a panel) so you set up labels correctly
- Ten to twenty minutes for the cleanup pass, depending on length and audio quality
You do not need a separate app for each file type. File Transcribe accepts WAV, M4A, and MP4 the same way.
Step 1: Get the recording off the device
Field recorders and phones store audio locally first.
- Phone voice memo: Export or AirDrop/share the M4A file to your computer, or upload directly from the phone browser
- Field recorder (WAV): Copy the file via USB or the recorder's card reader
- Zoom or video call (MP4): Download from Zoom's cloud recordings, or use your local recording if you captured it yourself
Keep the original file name with the date and subject. You will thank yourself when you have six interviews from the same week.
Step 2: Upload the file
- Open File Transcribe or go straight to interview recordings.
- Drop the WAV, M4A, or MP4 file, or paste a supported URL if the recording lives online.
- Guest upload works for a first pass with no account (daily caps apply). Sign in free to save the transcript and unlock SRT/VTT export later.
Step 3: Turn on speaker labels
If more than one person is talking, enable speaker labels before you start transcription. This splits the draft into Speaker 1, Speaker 2, and so on, based on who is talking when.
- One-on-one interview: Two speakers. Labels are usually clean with decent audio.
- Panel or group interview: More speakers means more overlap and more cleanup later. Budget extra editing time.
- Solo voice memo (your own notes about the interview): Skip speaker labels. There is nothing to separate.
Background on how this works and where it breaks down: what is speaker diarization.
Step 4: Let the transcript process
Processing time scales with file length, not complexity. A 45-minute interview usually finishes well before you would finish a coffee. Longer field recordings and full panel discussions take proportionally longer.
Step 5: Fix names, speakers, and hard-to-hear words
This is the step people skip and then regret. AI drafts get proper nouns wrong more than anything else: names, company terms, acronyms, and industry jargon.
- Rename Speaker 1 / Speaker 2 to actual names once you confirm who is who.
- Click any segment to jump the audio to that exact moment, rather than scrubbing a timeline.
- Fix names and terms in one pass through the whole transcript before you chase small filler words.
- Re-listen to any quote you plan to publish directly. AI transcripts are drafts, not final copy, especially on a recording with room echo or a bad phone connection.
Step 6: Export in the format you need
| Format | Use it for |
|---|---|
| TXT | Quick copy-paste into notes or a draft article |
| DOCX | Formatted transcript for editors or research files |
| Archival copy or something you send to a source for review | |
| SRT / VTT | Captions if the interview is going into a video |
TXT and DOCX exports are available on guest and free accounts. SRT and VTT unlock after you sign in free. See pricing if you are transcribing interviews regularly and need higher daily caps.
Handling common interview audio problems
Phone recorded in a noisy room
Expect more cleanup on names and quiet asides. If you have a choice, record with the phone closer to the speaker than to yourself.
Overlapping talk (interruptions, cross-talk)
Diarization struggles here more than word accuracy does. Fix the speaker attribution manually on any segment you plan to quote.
Bad phone line or dropped audio
Some AI drafts will guess at unclear words rather than leave a gap. Cross-check anything that reads oddly against the audio before you trust it.
Long-format interviews (over an hour)
Break your editing pass into sections rather than trying to fix the whole transcript in one sitting. Fix names and key quotes first; polish filler words last, if at all.
Why upload instead of a meeting bot for interviews
A meeting bot works for recurring internal calls. It is a worse fit for interviews:
- Sources are often outside your org and will not accept a third-party bot joining the call
- Field interviews and phone memos never touch a meeting platform in the first place
- You often need to re-listen and re-edit days later, which a live bot transcript does not support well
Upload-first transcription sidesteps all of that: record however you already record, upload when you are ready, edit at your own pace. See File Transcribe vs Otter if you are weighing both approaches.
FAQ
How do I transcribe an interview for free?
Upload the WAV, M4A, or MP4 on File Transcribe without an account for a first pass (guest daily caps apply). A free account adds saved transcripts and SRT/VTT export.
Does File Transcribe label speakers automatically?
Yes, when you enable speaker labels before transcription. You rename the generic labels (Speaker 1, Speaker 2) to real names during the edit pass.
What file format is best for interview recordings?
WAV gives the cleanest audio if your recorder supports it. M4A from a phone is fine for most interviews. MP4 works when the interview happened on video. File Transcribe accepts all three.
How accurate is AI transcription for interviews?
It depends heavily on the recording: mic distance, room noise, accents, and overlapping speech all move accuracy more than the tool you pick. Always re-listen to any quote before publishing it. More: what impacts AI transcription accuracy.
Can I export interview transcripts as captions?
Yes. Export SRT or VTT after signing in free if the interview is headed into a video, documentary, or YouTube upload.
Next step
Upload your interview file on the homepage and run through the steps above. For the freelancer-specific version of this workflow across interviews, podcasts, and captions, see best AI transcription apps for freelancers.
Related: Interview recordings · What is speaker diarization · Transcript format guide · Pricing
More guides
- Transcription guides
- Try File Transcribe free
- Transcript format guide
- Best transcription software
- Transcribe Zoom meetings