ClarumAUDIO Free trial →

Mic placement for a two-person podcast

Clarum Audio · September 2026

Two people. One room. Two mics. Most of the bleed you will fight later is decided here, before anyone opens a DAW.

A plugin can turn down the inactive mic. It cannot move the chairs.

Sit farther apart than feels polite

Face each other, or sit on a corner, not shoulder to shoulder. The other voice is quieter on your mic when there is distance and the mouth is not aimed at their capsule.

A small table is worse than a large one. If the table is small, move the mics, not the conversation.

Table dynamics vs lavs

Podcast interviews usually use table dynamics (SM7B / RE20 / PodMic class) or lavs on video. People do not shop for “a cardioid mic.”

Start with distance. Then match the mic type.

Table dynamic. Many have rear rejection. Point the dead side at the other speaker. Point the front at the person who owns that mic. A dynamic forgives a noisy room more than a condenser. It does not remove the other voice. It only makes it quieter before it hits the track.

Lav. Most lavs are omni. There is no rear to aim. Clip each lav on the outer side of the chest, away from the other person. Do not put both capsules on the near edges so they face each other. More space between seats still matters.

Head-worn or cheek mics are closer still. Same rule for omni: the other mouth should not sit right next to the capsule.

What placement cannot fix

Talk-over. Both mouths are open. Both mics should stay open. Distance only lowers the bleed under the real voice. It does not delete the other person.

A quiet talker next to a loud one. The quiet mic opens on the loud voice. Placement helps. It does not balance two very different levels by itself.

A live room. Hard walls, a glass table, a bare floor. Move blankets and absorption if you can. Do not expect a gate to clean what the room already printed onto both tracks.

After the mics are placed

Leave the placement alone for the whole session. Do not slide chairs in mid-interview and expect the edit to match the start.

Then record a minute of each person alone, and a minute of both. Listen to the other track. If the inactive mic is almost as loud as the active one, move the mics before you record the hour.

If placement is as good as the room allows and the inactive mic is still messy, that leftover is what a gain-sharing automixer is for. It compares the tracks. It does not replace the distance you skipped.

Longer note on gates vs automix: Mic bleed vs noise gate vs automixer.