Whether you are capturing a podcast conversation, a singer-songwriter and their guitarist, or two vocalists trading lines, the goal when you record two people simultaneously is the same: two clean, independent tracks with as little of each source bleeding into the other’s mic as possible. That separation is what lets you balance, edit, and process each person on their own later. This guide covers the gear you need, how to place two mics in one room, and the level discipline that keeps both channels usable.
The gear you need to record two people
The single most important requirement is an audio interface with at least two microphone preamps, each with its own gain control. A two-input interface such as a Focusrite Scarlett 2i2, Universal Audio Volt 2, or Audient iD14 lets you plug in two mics and set their levels independently, then record each to its own track in your DAW. A one-input interface simply cannot do this properly, so this is the piece to get right first.
For microphones, dynamic mics are usually the smarter choice for two people in one untreated room. A dynamic like the Shure SM58, SM7B, or a Rode PodMic rejects far more of the room and the other person’s voice than a sensitive condenser does. Condensers sound gorgeous in a treated space, but their wide, hot pickup means they hear everything, including the other performer, which makes clean separation harder. If you own two of the same model, use them both so the two channels share a tonal character.
Mic placement that minimises bleed
Bleed, sometimes called spill, is the sound of person A leaking into person B’s microphone. You cannot eliminate it entirely in a shared room, but smart placement dramatically reduces it. The two techniques that work are distance and rejection angle.
- Get close: Have each person work their mic at 10-15 cm (4-6 inches). The inverse square law means the closer the wanted source is, the louder it is relative to the bleed, so proximity is your best friend.
- Point mics away from each other: Every directional mic has a null, a point of least sensitivity, at the rear (cardioid) or the sides (figure-8). Angle each mic so its dead zone faces the other person. A common layout is the two performers facing each other with mics angled so each rejection null points across at the opposite talker.
- Use the null, not just the front: Aiming the back of a cardioid mic at the other source can reduce their bleed by 15-25 dB compared with the front. This is the single most effective free trick available.
- Add a barrier if you can: A gobo, a duvet on a stand, or even a bookshelf between the two people adds real isolation. Sitting them back to back with mics facing outward works well for spoken word.
If the two sources are wildly different in volume, for example a loud singer and a quiet acoustic guitar, physical separation into different rooms or heavy screening becomes more important, because no mic angle fully hides a loud source from a nearby mic.
Setting levels for two independent tracks
Route each mic to its own mono track in your DAW and arm both to record. Set each preamp’s gain individually while the person speaks or performs at their real, full performance level, not a polite soundcheck murmur. Aim for peaks around -12 to -6 dBFS on each channel. This leaves comfortable headroom so a sudden loud laugh or belted note does not clip, and clipping in a recording is permanent.
Two practical level tips make a big difference:
- Set gain for the loudest person first: If one performer is much louder, dial their channel so their peaks sit around -6 dBFS, then bring up the quieter person’s gain to a comfortable -12 dBFS. You are matching perceived loudness, not the knob positions.
- Record 24-bit: The wide dynamic range of 24-bit means you can track conservatively and still have a clean, low-noise recording, so there is no reason to push levels dangerously hot.
Enable input monitoring so both people can hear themselves, ideally on headphones to prevent that monitor sound from feeding back into the mics. If latency, the small delay between performing and hearing yourself, is distracting, lower your interface buffer size to 128 or 64 samples while recording. Raise it back to 512 or higher when you move on to editing so your computer can handle the plugin load without crackling.
Phase, editing and the final polish
When two mics are close together in one room, they can pick up the same source slightly out of time, causing phase issues that hollow out the sound. If a track sounds thin when you play both channels together, try nudging one track by a millisecond or two, or flip the polarity on one channel and keep whichever version sounds fuller. Because you recorded to separate tracks, you can also gate or manually silence the gaps where each person is not talking, which cleans up the bleed dramatically and tightens a conversation.
From there, treat each track on its own: a high-pass filter to remove rumble below 80-100 Hz, gentle compression to even out levels, and a touch of EQ so the two voices occupy their own space. Keeping the sources separate from the very start is what makes all of that possible, which is exactly why the two-preamp, well-angled setup is worth getting right before you hit record.
For deeper technique, our home recording guides cover cardioid pattern behaviour and gobos, the mixing cluster explains balancing dialogue and dual vocals, and the tools hub includes a level meter and a phase checker that help you confirm your two channels are clean before you commit.
