How to Record a Conversation in Person, Then Turn It Into Text

By whispertranscribe.ai Editorial Team · Last updated: November 2026 · 12 min read

How to record a conversation: tell everyone you're recording and get a clear yes, then start your phone's built-in recorder or a laptop app. Set the device on a soft surface, closest to the quietest speaker. Play back a short test first. Afterward, back up the original and upload it for a transcript with speaker labels.

This guide covers face-to-face talks: interviews, a café chat, a conference-room meeting. Phone calls have their own guide, how to record a phone call. Device steps follow Apple, Google, Samsung, and Microsoft help pages, checked November 2026.

Ask before you record. This is general information, not legal advice. US federal law sets a one-party consent minimum for in-person conversations, but some states require everyone's agreement. The Reporter's Recording Guide from the Reporters Committee for Freedom of the Press summarizes each state.

Quick answer: how to record a conversation in 4 steps

You need a device that records audio, a quiet spot, and permission. A phone is usually enough for two people at one table. For a bigger group or a noisy room, add an external microphone.

  1. Ask, then record the yes. Explain why, start the recorder, and ask again so the agreement is on the file.
  2. Place the device. Put it on a folded cloth or notebook, microphone uncovered, nearest the softest voice.
  3. Test, then record. Record a few seconds, play it back on headphones, then start the real recording.
  4. Save, back up, transcribe. Copy the original to a second location, then upload it and turn on speaker detection.

Which recording setup fits your conversation?

Match your situation to a starting setup. Every row assumes people already know they're being recorded.

Situation Recommended device Watch out for
Two people at a small table Phone on the table between you Calls and notifications interrupting
Journalist or research interview, one guest Phone in an interview mode, or a clip-on mic for the guest The guest's voice should be at least as loud as yours
Group of 3 to 6 around one table Phone or recorder in the middle, nearer the quieter end Voices at the far end drop off fast with distance
Conference room, 6 or more people External omnidirectional or boundary-style mic, or two recorders Air-conditioning hum and crosstalk; keep a second backup recording
Walking interview or noisy venue Clip-on (lavalier) mic on each speaker Clothing rustle and wind; keep the mic clipped firmly
Seated meeting with a laptop open Laptop with a USB or headset mic, or a browser recorder Typing and fan noise next to the mic
Long session, 2 hours or more Phone plugged into power, or a dedicated recorder Battery, storage, and per-file length limits

Choose your recording device

Start with the device you own. Distance between each mouth and the microphone changes the result more than the recorder's brand.

iPhone: Voice Memos

Voice Memos is built into iPhone. Apple's Voice Memos guide says it records with the built-in microphone, a supported headset, or an external microphone.

  1. Open Voice Memos (it lives in the Utilities folder unless you moved it).
  2. Tap the record button. To change the level, Apple's advice is to move the microphone closer to or farther from the source.
  3. Tap stop, then rename the recording; by default it's called New Recording or named after your location.

Settings that matter for conversations:

  • Voice Isolation or Wide Spectrum. In Control Center, tap Voice Memos at the top. Apple says Voice Isolation filters out background sounds and Wide Spectrum includes them; Wide Spectrum isn't on every model. Test both if people sit around the phone.
  • Recording Mode. Settings > Apps > Voice Memos > Recording Mode offers Mono, Stereo, or Spatial Audio (not on all models).
  • Other apps. You can switch apps while recording, but Apple says recording stops if another app starts playing audio.

Samsung Galaxy: Voice Recorder with Interview mode

Samsung's Voice Recorder has a recording mode built for one-on-one interviews. According to Samsung US Support, Interview mode captures sound aimed at the top and bottom of the phone and filters out other audio, while Standard mode records balanced sound for meetings and lectures.

  1. Open Voice Recorder, tap the menu, then Settings, then Recording mode.
  2. Choose Interview for two people facing each other, or Standard for a group.
  3. Lay the phone flat between you, top end pointing at one person and bottom end at the other.
  4. Tap record. Use the bookmark icon to flag key moments, then stop, name the file, and save.

The same settings let you set Recording quality (High, Medium, or Low) and Block calls while recording. If recordings stop when the screen locks, Samsung suggests setting the app's battery to Unrestricted under Settings > Apps > Voice Recorder > Battery.

Google Pixel: Recorder

The Recorder app works on Pixel 3 and later phones and Pixel Tablet, per Pixel Phone Help. It keeps recording when the screen sleeps, and a single recording can run up to 18 hours.

On Pixel 6 and later, Recorder can add speaker labels (Speaker 1, Speaker 2) to its own transcript, but Google's speaker labels page says the feature works only in US English.

A laptop or desktop computer

A computer suits a seated meeting: steady power, plenty of storage.

  • Windows: Microsoft's Sound Recorder FAQ says each file can last up to three hours. Ctrl+R starts, Space pauses, and Esc stops and saves.
  • Mac: Voice Memos is built in too. Under Voice Memos > Settings, Apple's Mac settings page offers Audio Quality: Compressed (smaller files, lower quality) or Lossless (larger, higher quality).
  • Any computer with a web browser: whispertranscribe.ai's online voice recorder lets you pick the language and microphone, switch on speaker detection, and record straight into a time-stamped transcript.

External microphones

Add a microphone when someone still sounds distant after you've moved the device. Pickup pattern matters more than price. The first three rows follow Shure's guide to polar patterns; the last is a mic style rather than a pattern.

Pickup pattern Hears sound from Good for a conversation when Placement tip
Omnidirectional All directions equally Several people sit around one table in a quiet room Center it among the speakers
Cardioid Mostly the front One speaker per mic, or a noisy room Aim the front at the mouth; Shure says cardioids reduce ambient noise by roughly two-thirds compared with omnis
Bidirectional (figure-8) Front and back, not the sides Two people facing each other across a table Put it between them, each person facing one side
Clip-on lavalier Mainly the person wearing it, because it sits close Walking interviews, noisy venues, panels Clip it on the chest, clear of scarves and zippers

Check compatibility first; Apple points iPhone users to accessories labeled "Made for iPhone" or "Works with iPhone."

Where to place the recorder: a room checklist

Distance is the biggest factor you control. Shure's Microphone Techniques for Recording explains that direct sound drops about 6 dB each time the distance doubles, while the room's ambient sound stays at nearly the same level everywhere. So the farthest voice sinks toward the room noise.

  • Closest to the quietest voice. Shift the device toward the softest speaker, not the loudest.
  • On something soft. A notebook or folded cloth keeps table knocks out of the mic.
  • Microphone uncovered. Don't leave the phone in a pocket, a bag, or under papers.
  • Away from steady noise. Avoid air vents, projectors, fridges, and coffee machines.
  • Fewer hard surfaces. Glass and bare walls bounce sound; carpet, curtains, and bookshelves help.
  • Interruptions off. Turn on Do Not Disturb and silence other devices on the table.
  • Power and storage. Charge, check free space, or plug in.
  • A real test. Have the farthest person speak normally, and listen on headphones.
  • Two mics? Space them out. Shure's 3-to-1 rule says the distance between microphones should be at least three times the distance from each mic to its own speaker.

We haven't measured these setups, so we don't give exact distances. Your test recording in the real room is what counts.

For groups: help the transcript later

Speaker detection separates voices but can't name them. Have each person say their name at the start, avoid talking over each other on decisions, and repeat key numbers clearly.

Get agreement first, and capture it on the recording. In-person rules depend on the state or country and on whether people expect privacy. This section is general information, not legal advice.

  • Federal minimum. The RCFP's introduction to its recording guide says federal law requires the consent of at least one party before recording in-person, telephone, or electronic conversations, and that states may set stricter rules.
  • Some states want everyone's consent. The same page says about 11 states mainly require all parties to agree. Our guide to call recording laws by state lists them with statute links.
  • Expectation of privacy. Federal law's definition of "oral communication" covers speech made by someone expecting it isn't being intercepted, under circumstances that justify that expectation (18 U.S.C. § 2510(2)). The RCFP says courts look at where the talk happened, what it was about, and how loudly people spoke, and that conversations in a home carry the greatest expectation of privacy.
  • Visible recorder. The RCFP says consent is generally presumed when a face-to-face interview is recorded with the device in plain view, but calls recording verbal consent a best practice.
  • Outside the US. Rules differ by country; also check your employer's or school's policy.

Say something like this after you press record, then wait for each person to answer. (This script is an example written for this article.)

"It's [day], we're at [place]. I'm recording this conversation so I can get an accurate transcript. It'll be stored [where] and shared with [who]. Is everyone OK with that? Could each of you say your name and 'yes'?"

If anyone hesitates, take written notes instead. If someone changes their mind midway, stop and agree on what happens to the part already recorded.

After you record: name, back up, and pick a format

Deal with the file the same day. Microsoft notes that Sound Recorder stores recordings inside the app, so uninstalling it deletes them.

Name and back up the original

  1. Rename it so a search finds it. Include the purpose and the people, for example interview_lee-patel_original.m4a (made-up example).
  2. Keep the original untouched. Save edits under a new name; Pixel Recorder saves crops as a copy.
  3. Make a second copy on another drive or in cloud storage you control.
  4. Write a short note beside it: who's speaking, where, and the time code where consent was given.
  5. Plan a delete date for sensitive recordings.

M4A, MP3, or WAV: which format to record in?

Keep your device's default unless you have a reason to change. Converting later can't add back detail that wasn't captured.

Format What it is Approximate size per hour Pick it when
M4A (AAC) Compressed; the usual default on iPhone and many recorders About 58 MB at 128 kbps Everyday interviews and meetings
MP3 Compressed; plays almost anywhere About 86 MB at 192 kbps You need to share or play the file on older devices
WAV, 16-bit, 44.1 kHz, mono Uncompressed About 318 MB You plan to edit the audio, and the room was quiet
WAV, 24-bit, 48 kHz, stereo Uncompressed, two channels About 1.04 GB Podcast or broadcast production

Sizes are our arithmetic (bit rate × 3,600 seconds ÷ 8), not measurements. They matter because whispertranscribe.ai accepts files up to 5GB and 10 hours; by that math, a 24-bit, 48 kHz stereo WAV passes 5GB at roughly 4.8 hours. If a long WAV is too big, convert it to MP3 first, or split it.

How to transcribe a recorded conversation

A transcript makes an hour of audio searchable and quotable. Here's the workflow with whispertranscribe.ai, a browser-based transcription tool powered by OpenAI's Whisper model. Everything happens in the browser, so there's nothing to install.

  1. Open an upload page. For an interview, use interview transcription. To try it without signing up, the free transcription page with no account required works the same way. Free limits are shown on the page and can change. Recorded in Voice Memos? Use the voice memo transcription page; our voice memo transcription guide covers exporting the file.
  2. Choose the spoken language. Pick from 130+ languages. Many have a Noise / Low-Quality Audio/Video option for café or street recordings, and there's a Multilingual Mix option for conversations that switch languages.
  3. Upload the file. M4A, MP3, WAV, AAC, FLAC, OPUS, and video such as MP4 or MOV are accepted.
  4. Tick Enable Speaker Detection. The text is split by speaker; match labels to the names said at the start.
  5. Review, then export. Accuracy reaches up to 99% on clear audio; noise, crosstalk, distance, and strong accents lower it, so check names, numbers, and quotes. Export DOCX, PDF, TXT, CSV, SRT, or VTT.

At a laptop and not recorded yet? The browser voice recorder records and transcribes in one step. For calls, see phone call transcription.

Paid plans start from $9.99/month billed annually; see pricing and the full list of features.

What happens to the file you upload

Under whispertranscribe.ai's privacy policy, the file you upload is removed automatically once processing finishes, MP3 copies made during processing are kept for 30 days by default, and transcripts remain in your account until you remove them. The same policy states that your content isn't used for AI training or reviewed by staff, and that it's protected by TLS 1.3 during transfer and AES-256 when stored. Tell participants where the file will go, and read the full policy before uploading anything sensitive.

Conversation recording troubleshooting

Most problems come from placement, power, or the file itself.

Symptom Likely cause Fix
One person is faint or sounds far away Recorder was nearer the louder speaker Move it toward the quiet voice, or give that person a clip-on mic
Voices sound hollow or boomy Hard, echoey room, or two mics picking up the same voice Choose a softer room; space mics using the 3-to-1 rule
Constant hum or hiss Vent, projector, or laptop fan near the mic Move the device and test again
iPhone recording stopped early Another app started playing audio Keep media apps silent during recording
Galaxy recording stopped when the screen locked Battery optimization paused the app Set Voice Recorder battery use to Unrestricted
Windows recording cut off Sound Recorder files max out at three hours Stop and start a new file at a natural break
Upload refused File is over 5GB or longer than 10 hours Convert to MP3 or split into parts
Speakers merged in the transcript People talked over each other, or Speaker Detection was off Re-run with Speaker Detection on; restate key points one at a time next session
Wrong words for names and jargon Unfamiliar terms, noise, or accent Proofread against the recording, using the names spoken at the start

FAQs

Often, but it depends on where you are. US federal law sets a one-party consent minimum for in-person conversations, and about 11 states mainly require everyone's agreement. Many laws apply only where people reasonably expect privacy, which courts judge by place, topic, and volume. A recorder in plain view usually helps. Ask on the recording, and check state or local law. This is general information, not legal advice.

Where should I put my phone to record two or more people?

Put it on a soft surface, microphone uncovered, closest to whoever speaks most quietly. For two people, lay it flat between you; on a Samsung Galaxy, Interview mode records from the top and bottom ends. For a group, start in the middle and shift it toward the far end. Direct sound drops about 6 dB each time distance doubles, so test from the farthest seat.

Do I need an external microphone to record a conversation?

Not usually for two people in a quiet room. A phone on the table often works if a test recording sounds clear. Add a mic when someone still sounds distant after you move the device, when the room is noisy, or when more than a handful of people speak. A clip-on mic per speaker suits walking or noisy interviews. Check compatibility with your phone or computer first.

What audio format is best for recording a conversation?

For most conversations, the default M4A (AAC) on your phone is fine and keeps files small, around 58 MB an hour at 128 kbps. Use WAV only if you plan to edit the audio. A 24-bit, 48 kHz stereo WAV runs about 1 GB an hour, so it can pass a 5GB upload limit in under five hours. Converting later can't restore detail, so choose before recording.

How long can I record a conversation?

It depends on the app. Google's Recorder on Pixel handles up to 18 hours in one recording. Microsoft's Sound Recorder FAQ says each file can last up to three hours. Voice Memos on iPhone stops if another app plays audio. For long sessions, plug in power and check storage. To transcribe, whispertranscribe.ai accepts files up to 10 hours and 5GB, so split longer recordings.

How do I get a transcript that shows who said what?

Upload the recording to a transcription tool with speaker detection, then rename the labels. On whispertranscribe.ai, choose the language, upload the file, tick Enable Speaker Detection, and export DOCX, PDF, TXT, CSV, SRT, or VTT. Ask everyone to say their name at the start so you can match voices to labels. Accuracy reaches up to 99% on clear audio; noise, overlap, and accents lower it.


Recorded the conversation? Get the text. Open whispertranscribe.ai interview transcription, pick from 130+ languages, switch on speaker detection, and download the result as DOCX, PDF, TXT, CSV, SRT, or VTT.

References

Ready to turn your recordings into text?

Transcribe audio and video with whispertranscribe.ai.

Start transcribing ↗