Pepys

Interview Recording Checklist

A pre-record run-through for gear, room, consent, and levels, so your audio starts clean.

Transcript accuracy is mostly decided before anyone says a word. Whatever your recorder captures is the best a person or an AI tool will ever have to work with, and no cleanup step fully recovers a muddy, echoey, or clipped file. This checklist is the short pass to run right before an interview: check the gear, quiet the room, get consent on tape, and record a test clip you actually listen back to.

Work top to bottom, tick each box, and only start the real conversation once the test clip sounds clean. When the interview is done, hand the file to how to transcribe an interview. If you are still deciding how to save the file, check the best audio format for transcription before you press record.

The template

Pre-Interview Recording Checklist

Fill in the details, then work down each section. Do not start the real interview until every box is ticked and the test clip sounds clean.

Interview / project:(name or reference)

Date:

Location:(room, video call, or phone)

Interviewer:

Participant(s):

1. Gear

Primary microphone connected and selected as the input.

Mic placed close to the speaker, roughly a fist's width from the mouth, angled slightly off to the side so hard 'p' and 'b' sounds do not pop.

One mic per speaker where possible, so voices stay separate and easier to tell apart.

Backup recorder running (a second recorder or a phone voice memo), so one failure does not lose the whole interview.

Headphones on to monitor the live sound, not just watching the meters.

Batteries fresh or fully charged, with a spare set within reach.

Storage checked: memory card or drive has free space, and the card is formatted for this recorder.

Recorder set to save a high-quality file (a lossless format such as WAV is a safe default for transcription).

Phones and laptops set to silent or do-not-disturb so alerts do not land on the recording.

2. Environment

Quietest available room chosen, with the door closed and a note outside asking others not to interrupt.

Mic kept off bare hard surfaces. Use a stand or a folded cloth so table thumps and vibration do not travel into it.

Soft items in the room (curtains, a rug, a sofa) to soften echo. Bare rooms with hard walls bounce sound and blur speech.

Away from steady hum: air conditioning, heating vents, fans, and the fridge switched off or avoided.

Windows closed to keep out traffic, wind, and outdoor voices.

Buzzing or noisy electronics moved away from the mic.

3. Consent

This section is a starter, not legal advice. Rules on recording consent vary by country and by state or region, and some places require every participant to agree. Check what applies where you and your participant are before you record, and when in doubt, ask.

Explained who you are, the purpose of the interview, and that you would like to record it.

Explained how the recording will be used, stored, and who may hear or read it.

Asked permission to record, and gave the person room to say no or ask questions.

After pressing record, captured a spoken 'yes' on the file: ask again and let them confirm on tape.

Noted anything they asked to keep off the record or any limits they set.

Gave them a way to reach you later if they want to add, change, or withdraw something.

4. Levels and test clip

Recorded a 20 to 30 second test clip with both people talking at normal interview volume.

Watched the level meter: the loudest moments sit comfortably below the top, with headroom to spare, so peaks never hit the ceiling and distort.

Played the test clip back on headphones and actually listened to it.

Both voices are clear, balanced, and roughly even in volume.

No clipping, hum, hiss, echo, or background noise loud enough to bury words.

Fixed anything you heard and re-tested. Do not settle for a test clip that already sounds rough.

5. Once you hit record

Confirmed the record light or timer is genuinely counting up before the first real question.

Said the date, names, and purpose out loud at the start as a spoken label for the file.

Backup recorder confirmed rolling too.

Glanced at levels now and then, and noted the time of any key moments to find later.

Interviewer:Date:

Participant (acknowledges recording):Date:

How to use it

  1. 1

    Open this the morning of the interview and gather the gear it lists: primary mic, a backup recorder, headphones, and spare batteries or a cleared memory card.

  2. 2

    In the room, run the gear and environment boxes: quiet the HVAC and fridge hum, move the mic off the bare tabletop, and close the door.

  3. 3

    Before the real questions, ask permission and record a spoken yes, then capture a 20 to 30 second test clip of both people talking.

  4. 4

    Listen back on headphones. If you hear clipping, hum, or echo, fix it and re-test before you commit to the session.

  5. 5

    After the interview, upload the file to a pay-as-you-go tool like Pepys to turn it into a searchable transcript.

Recorded it? Transcribe it here

Drop in the recording for a speaker-labeled, timestamped draft in minutes. Your first 60 minutes are free, no card.

or paste a link
InstagramTikTokYouTubeFacebookSpotifyApple Podcasts

More templates

Frequently asked questions

Does the recording really change transcript accuracy that much?

Yes. Background noise, room echo, overlapping voices, and levels that are too low or clipped all raise the error rate, whether a person or an AI tool does the transcribing. The cleaner the file going in, the less correcting you do afterward, which is why the recording sets the ceiling on how accurate the transcript can be.

What is the single most important item on the list?

Mic placement. Getting the microphone close to each speaker, roughly a fist's width away and angled slightly off-axis, does more for clarity than almost anything else, because it captures the voice louder than the room around it.

Do I need consent to record an interview?

It depends on where you and your participant are. Consent rules vary by country and by state or region, and some places require everyone in the conversation to agree. Treat the consent section here as a starting point, not legal advice: explain what you are doing, ask, and capture a spoken yes on the recording. When you are unsure, check the local rule first.

What recording level should I aim for?

Loud enough to be clear, but with headroom so the loudest moments never hit the top of the meter and distort. If the meter is pinning at the ceiling, the audio is clipping and words get mangled; if it barely moves, the voice is buried in noise. Aim for a strong, steady signal that leaves room at the top.

Is a backup recorder worth the hassle?

For an interview you cannot easily redo, yes. A second recorder or even a phone voice memo running alongside your main one means a dead battery, a full card, or a bumped cable costs you nothing. It is the cheapest insurance on this list.

Don't just take our word for it.

Ask ChatGPT, Claude, or Perplexity what Pepys is and who it's for. One click, and your favorite AI does the homework.