You wrap a great interview and open the files. One guest sounds like a radio host. The other sounds like a voicemail. That gap is almost never talent. It is the device each guest recorded on, and the automatic processing baked into it before the file ever reached you.
This guide covers how to record a podcast with guests on different devices without spending your night rescuing tracks. If you are still deciding whether to guest, sponsor, or launch your own podcast, settle that first. You will learn how to prep guests and what to control mid-session. You will also learn the best practices for editing podcasts recorded in different locations, so every voice sits in the same room. Clean inputs beat clever fixes every time.
To record a podcast with guests on different devices, standardize the recording app, disable automatic processing, and level-match every track afterward. Match every track to 16 LUFS integrated for stereo, or 19 LUFS for mono. Keep peaks under 1 dBTP so no voice jumps out.
1. Why guest devices make your audio clash
Every device decides for itself how to capture and shape sound. A laptop uses a small omnidirectional mic that grabs the voice plus keyboard clicks, fans, and room echo. An iPhone runs several mics through beamforming to push the voice forward. A USB headset often adds its own compression tuned for calls, not for editing.
Those choices do not match. Placed side by side on a timeline, the tonal and spatial gaps are obvious to any listener. EQ narrows the gap, yet it cannot turn a thin call-quality capture into a broadcast one. This is why controlling the capture matters more than owning great gear.
The processing that does the real damage
The worst offender is automatic gain control, or AGC. It watches the incoming level and constantly nudges it up or down to keep volume steady. On a phone call that helps. In a recording it creates a pumping effect that is hard to undo.
During a pause, AGC raises the gain and drags up the noise floor. When your guest speaks again, it drops the gain fast. The noise bed rises and falls under the voice, and that motion is welded into the file. You cannot lift it out cleanly later, so you have to stop it before it happens.
Before any session, ask the guest to disable noise suppression and echo cancellation. Have them switch off any "clear voice" or "enhanced" mode too. These features are built for live calls and fight you in the edit.
2. How to prep guests before you record
The best edit starts before anyone hits record. Two minutes of prep with a guest saves two hours at the desk. Give them a method, not a vague request to "sound good."
Standardize the recording app
Ask every guest to capture uncompressed local audio instead of leaning on call platforms or default voice apps. A dedicated recorder or a simple local editor like Audacity holds a steady format. For iPhone guests, skip Voice Memos, which records compressed AAC and throws away detail you later need for EQ and noise reduction.
Point iPhone guests to a recorder that captures WAV or AIFF instead. For Android guests, avoid the stock voice recorder. Set a dedicated app to record WAV at 48 kHz, with every enhancement switched off. A quick format check before you start is the cheapest insurance in podcasting.
Send this guest checklist every time
Guests forget instructions given by voice. Paste this into the calendar invite so the rules sit in front of them on the day. Copy it, adjust the app name to your pick, and reuse it for every booking.
Use headphones, not speakers, so my voice does not echo into your mic.
Record somewhere soft and quiet. A room with a rug, a sofa, or clothes beats a kitchen.
Sit 10 to 15 cm from the mic, slightly off to one side, and stay put.
Close spare apps and browser tabs, and silence your phone.
Use the app I sent, not your default voice recorder, and turn off noise reduction or clear voice.
Plug in your laptop so it does not throttle on battery.
Record a 30 second test and send it before we lock the full session.
Run a short sound test
A quick test recording surfaces problems while you can still fix them. Listen for background noise, room reflections, wrong gain, and processing artifacts. If the test sounds off, solve it now, because the same fault will run through the whole session.
3. Which recording platform records each guest locally
One feature matters more than price or video: whether the platform records each guest locally on their own device. When it does, a shaky connection only hurts the live call you hear, not the file you edit. That single fact separates clean multi-guest audio from stitched-together compromises.
Video call tools like Zoom send one compressed stream from the cloud, so a weak signal bakes dropouts into the recording. Local-recording platforms keep a full-quality file on each end and upload it after. Here is how the common options compare on the traits that decide guest audio quality.
| Method | Records each guest locally | Video | Best for |
|---|---|---|---|
| Riverside | Yes, per participant | Yes | Video podcasts and clips |
| Descript Rooms | Yes, per participant | Yes | Record and edit in one place |
| Cleanfeed | Yes, per participant | No | Audio-first, low-tech guests |
| Zoom | No, cloud stream | Yes | Guests who know nothing else |
| Double-ender | Manual, both sides record | No | Maximum control on a budget |
If a guest refuses new software, a double-ender still works. Both sides record their own audio during a normal call, then the guest sends you their file. You sync the two clean tracks and drop the call audio entirely.
Choosing the right platform is only part of a strong session. For a full walkthrough of remote workflows, guest prep, internet setup, and backups, read our remote podcast recording guide.
Recording video as well as audio? Prep guests for camera placement, lighting, and webcam quality on top of their mic setup. Our guide to video podcasts with webcam guests covers those visual best practices.
4. What to watch while the recording runs
Once the session is live, your job shifts to catching problems you cannot fix later. Watch levels, log trouble spots, and handle drops without derailing the talk.
Keep levels in a safe range
Aim for peaks around 12 dBFS to 6 dBFS and an average near 18 to 20 dBFS. Anything that touches 0 dBFS clips, and clipped audio cannot be repaired cleanly. Quiet audio is the safer miss, since you can raise it later, though lifting it also lifts noise.
Fix a bad level without making it weird
If a guest reads too low or too hot, do not frame it as a fault. Say you want to check their audio, ask them to say their name and count to five, then adjust and move on. The check takes ten seconds and spares you a whole session at the wrong level.
Log timestamps as you go
Keep a plain text file open beside your recorder. Every cough, siren, doorbell, or connection drop gets a timestamp and a one word note. When you edit, you jump straight to each spot instead of scrubbing a two hour file blind.
Handle a mid-session dropout
If the connection stutters, note the time and keep the conversation alive rather than stopping cold. When a whole answer gets buried by a sustained noise, ask the guest to repeat it on the spot. The room and their energy still match, which a later re-record cannot promise.
5. How to level-match and fix tracks in post
Post-production is for refinement, not rescue. Work in order: match loudness first, then shape tone, then clean noise only where it is needed. Rushing to noise reduction before leveling just hides the real gaps.
Level-match with LUFS, not peaks
Normalization sets peak levels, but peaks are not perceived loudness. Use a LUFS meter to match how loud each track actually feels. Apple's audio requirements set the podcast target at 16 LUFS integrated for stereo, with a true peak no higher than 1 dBTP. A mono spoken-word file lands around 19 LUFS to sound just as loud.
EQ to close the tonal gap
EQ exists here to make voices match, not to perfect each one alone. Trim low rumble below roughly 80 to 100 Hz. Cut boxiness around 200 to 400 Hz. Add gentle presence near 3 to 5 kHz where a voice sounds dull. Aim for consistency across guests, then stop.
Use repair tools with restraint
Tools like iZotope RX reduce noise and hum on damaged tracks, and AI cleaners can lift clarity. Push them too hard and you get a metallic, underwater sound that is worse than the noise. Check your work at a low playback volume, where tonal imbalances and artifacts are easier to catch than on loud monitors.
Match loudness before you touch EQ or noise reduction. Once every track sits at the same LUFS target, the real tonal gaps show up honestly. Then you fix the right thing instead of masking it.
Build your dream guest list
Great audio starts with booking the right shows and people. Search 3M+ podcasts in MillionPodcasts, filter to the ones that accept guests in your niche, unlock verified host and booker emails, and export the shortlist to CSV or Excel for your own outreach.
Search 3M+ podcasts →6. When to re-record instead of fixing it
Some tracks cannot be saved. Restoration costs more than a fresh take in three cases. When audio is clipped throughout, when noise dominates the voice, or when a hard, reflective room offers no control. Re-recording the broken parts is the faster path.
Find the unusable sections and re-record only those. These targeted takes, called pickups, work best when the new recording matches the original room as closely as possible. A pickup from a quieter or more echoey space creates a seam as jarring as the problem it replaced.
Record 30 seconds of room tone at the start of every session for exactly this moment. Laying that ambient bed under a pickup smooths the join. Takes captured days later, on a different device in a different room, are the hardest to blend. The wider the gap between the two, the more obvious the edit stays.
7. Podcast guest audio questions, answered
What LUFS should a podcast be for guests recorded in different locations?
Level every track to the same integrated loudness before you mix. The podcast standard is 16 LUFS integrated for stereo and 19 LUFS for mono. Keep the true peak no higher than 1 dBTP. Match all guest tracks to that target so no voice jumps out.
What is AGC and why does it ruin podcast recordings?
AGC, or automatic gain control, adjusts the recording level in real time. During pauses it raises the gain and pulls up background noise, then drops it fast when speech returns. The result is a pumping noise floor that is very hard to fix later. Disable it before recording.
How do I disable audio processing in Zoom for podcast recording?
In Zoom, open Settings, then Audio, then Audio Profile, and enable Original Sound for Musicians. This turns off echo cancellation, noise suppression, and automatic gain control, giving you a cleaner signal for editing.
What recording app should iPhone podcast guests use?
Avoid Voice Memos because it records compressed AAC. Ask iPhone guests to use Ferrite Recording Studio, set to WAV or AIFF. Those formats preserve more detail for EQ and noise reduction.
When should you re-record a podcast guest instead of fixing it in post?
Re-record when audio is clipped throughout. Re-record when background noise dominates the speech, or when a highly reflective room gives you no control. Fixing these in post costs more time than a targeted pickup and rarely sounds clean.
References
Apple Podcasts for Creators. (2026). Audio requirements. https://podcasters.apple.com/support/893-audio-requirements Podnews. (April 2025). How to do loudness: the LUFS and LKFS FAQ for podcasters. https://podnews.net/article/lufs-lkfs-for-podcasters International Telecommunication Union. (2015). ITU-R BS.1770, Algorithms to measure audio programme loudness and true-peak audio level. https://www.itu.int/rec/R-REC-BS.1770