The Podcast Microphone Guide I Wish Every Host Read First

Buy a dynamic microphone on an XLR cable, and buy it before you buy anything else. A dynamic mic in a normal untreated room beats an expensive condenser in that same room, every single time. Dynamics have a tight pickup pattern and a low sensitivity, so they hear your mouth and largely ignore your air conditioner, your keyboard, and the road outside.

The popular broadcast dynamics all work. So do the cheaper ones. In the edit I honestly cannot pick a $99 dynamic from a $400 dynamic once both are placed correctly and gained correctly. What I can pick, instantly, is a mic sitting eighteen inches away on a desk stand next to a laptop fan.

If you are buying for two people in one room, buy two identical mics. Matching them saves me an hour of matching tone later, and it makes the episode sound like one show rather than two phone calls stitched together.

If you are still deciding on format and gear order, I laid out the full sequence in my walkthrough on how to start a video podcast, which covers cameras and lighting in the same breath.

Use a dynamic unless your room is treated, and use a condenser only when it is. That is the whole decision, and it has almost nothing to do with which mic sounds better on a spec sheet.

Condensers are more sensitive. That sensitivity is a feature in a padded studio and a curse in a spare bedroom. A condenser will faithfully record the hum of your fridge two rooms away, the reflection off your window, and the tiny slap off the desk in front of you. I can reduce that in post. I cannot remove it, because room reverb is baked into the same waveform as your voice.

Here is the practical test. Stand in your recording spot and clap once, hard. If you hear a short ring after the clap, you have reflections, and you want a dynamic mic plus something soft on the wall behind you. A bookshelf, a curtain, a duvet on a stand. It looks unglamorous on camera, but a video podcast can hide it just out of frame.

One more honest note. Condensers pick up mouth noise and breath more than dynamics do. If you are a close talker, that is another vote for dynamic.

One fist away, angled slightly off axis so you are not breathing straight into the capsule. That is roughly four inches. Most hosts sit at twelve to eighteen inches because that feels polite, and it costs them the entire low end of their voice.

Distance is the single biggest variable in podcast audio, bigger than the price of the mic. Move from four inches to twelve and you lose proximity bass, you double the amount of room you capture, and you force the gain up, which drags the noise floor up with it.

For video shows this creates a real tension. A boom arm bringing the mic in close will sit in frame. My advice is to stop fighting it. Let the mic be visible. Podcast audiences read a mic in shot as a signal that the show is serious. The alternative is a lav mic under a shirt collar, which is fine for audio but adds fabric rustle every time someone gestures.

If you record standing, or you move a lot, tape the cable to the stand so you do not tug the mic mid sentence. I have cut around that noise more times than I want to admit, and it always costs a good take.

Aim for peaks around minus twelve to minus six dBFS, and never let anything touch zero. Digital clipping is permanent. When a waveform goes flat at the top, the information is gone, and no editor on earth reconstructs it. I can hide a quiet recording. I cannot un-clip a loud one.

Set gain with the guest at their loudest, not their calmest. Everybody gets louder ten minutes in when they start enjoying themselves. Ask them to laugh during the level check, because that laugh is the true peak of your episode.

Watch the noise floor too. With the room quiet and nobody speaking, your track should sit down near minus sixty. If it is sitting at minus forty, something is generating hiss, usually a cheap interface being asked for too much gain on a low output dynamic. An inline gain booster solves it for about $80 and saves the whole show.

Record each person to a separate track. Always. Merged tracks mean I cannot duck one voice under the other, cannot fix one person coughing while another talks, and cannot cleanly cut a tangent. Separate tracks are free and they change what is possible in the edit.

Every person gets a microphone. Cameras do not need microphones, but they do need a reference. Those are two different jobs and people conflate them constantly.

One mic shared between two hosts sounds like one good voice and one distant voice, and there is no fix. Even with the mic centered, whoever leans away is now off axis and quieter. Two mics, two tracks, done.

Cameras matter for a different reason. If you are running two, three, or four angles, let every camera record its own scratch audio. Nobody uses camera audio in the final mix, but that scratch track is what lets us sync angles in seconds instead of eyeballing claps. This becomes very obvious on multi-cam editing, where an unsynced angle can quietly drift by a frame or two over an hour.

One practical trick: clap once at the top, on camera, in view of every lens. It takes two seconds and it is a free insurance policy against a dead timecode setting. Hosts tell us it feels silly. It is not silly when a battery dies mid record and the file restarts.

I can fix a lot. Consistent hiss, hum from a bad ground, plosives on hard P sounds, uneven levels between a loud host and a shy guest, mouth clicks, a little room reverb, and background noise that stays steady across the episode. Those are all routine, and they are included in the edit at $199 or $299 per episode rather than billed as some special restoration line item.

Here is what I cannot fix. Clipped audio. A voice recorded so far off mic that the room is louder than the person. Two people on one mic talking over each other. A dropped Bluetooth headset that swaps sample rate mid sentence. Music playing in the room behind you for forty minutes.

Because my editors watch the whole episode rather than running silence detection, we catch these problems in the first ten minutes and tell you, which at least means episode two does not repeat them. That human pass is also how we find the moments worth pulling for clips and where b-roll and cutaways will actually help rather than distract.

If your show interviews specialists, audio clarity matters even more, because names and terms have to survive. That is a big part of how I approach podcast editing for doctors and other expert led shows.

Record locally on both ends, then upload the files. Never rely on the Zoom or call recording as your master. Call audio is compressed, sample rate limited, and full of small drops you will not notice live but will hear on headphones later.

Most remote recording platforms already do this. They record a full quality file on each participant's machine and upload it in the background. Use that. Ask your guest to stay on the page until the upload bar finishes, because closing the tab early is the most common way a guest track disappears.

Also ask guests to use headphones. Not for their comfort, for your edit. Without headphones their speakers leak your voice back into their microphone, and now every word you say exists twice, slightly delayed. Echo cancellation makes that worse by pumping the guest track up and down.

Send guests a one paragraph brief before the call: headphones on, quiet room, mic a fist away, laptop plugged in, notifications off. It takes ninety seconds to write and it raises your audio quality more than any gear purchase. I usually fold that same brief into the show notes and prep workflow so it lives in one place per episode.