Podcast Recording: What You Actually Need

Acoustics versus soundproofing, mic technique that fixes most problems free, multi-cam setups for video podcasts, and the audio mistakes post cannot fix.

Listeners forgive bad video. They will watch a slightly soft, slightly badly framed shot for an hour without complaint. They will not forgive bad audio - echo, room noise, or a guest who sounds like they are phoning in from a stairwell gets the episode closed within a minute.

So the order of priority is: room, then microphone technique, then microphone, then everything else. Most people buy that list backwards.

Acoustics and soundproofing are not the same thing

These get used interchangeably and they solve completely different problems.

Soundproofing is about isolation - stopping outside noise from getting in. Traffic, a lift shaft, a neighbouring unit, aircraft. It is a construction problem, solved with mass and separation, and it is expensive. A room-within-a-room build - a structurally separate inner room sitting inside the outer shell - is the proper solution, which is why studios that need real isolation build one.

Acoustic treatment is about what the room does to sound made inside it. Hard parallel surfaces bounce your voice back into the microphone a few milliseconds late, and that is what makes a recording sound hollow, boxy or distant. It is solved with absorption - panels, soft furnishings, broken-up surfaces.

You can have one without the other, and both failure modes are audible. A perfectly isolated bare concrete box still sounds terrible. A well-treated room next to a busy road still picks up buses. Scope's Studio B is a sound-treated room-within-a-room, and the podcast room is set up specifically for recorded speech.

Testing a room in ten seconds

Stand in the middle and clap once, hard. If you hear a ring, a flutter or a tail after the clap, the room will do that to every word you say. Then stand still and listen for thirty seconds with your eyes shut. Air conditioning, a fridge, a computer fan, traffic - all of it is going under your recording and none of it comes out cleanly afterwards.

You cannot fix a room in post. You can only fix it in the room.

Microphone technique is free and fixes most of it

The single biggest improvement available to most podcasts costs nothing: get closer to the microphone and stay there.

Distance is the whole game. The closer the mic is to the mouth, the more of what it captures is voice and the less is room. Double the distance and you get dramatically more room in the recording. A hand's width or so, consistently, is the usual target for a cardioid dynamic - but the exact distance matters less than not moving.

  • Speak across the mic rather than directly into it. This is what stops hard P and B sounds hitting the capsule as a thump of air.
  • Use a pop filter or a foam windshield, and use it as well as good technique, not instead of it.
  • Keep your distance constant. Someone who leans in when animated and back when listening produces a track that changes tone every twenty seconds and is a nuisance to level.
  • Do not touch the table. Every knock, tap and laptop keystroke travels straight up a boom arm into the recording. Shock mounts help; not tapping helps more.
  • Headphones on, always. If you are not monitoring, you do not know something has gone wrong until the session has ended.

The signal chain

Deliberately generic - specific models change constantly, and any studio will tell you what they hold if you ask.

One microphone per person, always. Two people sharing a mic guarantees an unbalanced recording and removes any chance of fixing one voice without affecting the other. Dynamic microphones are the standard for podcasting because they reject room sound and handle loud voices well; condensers are more detailed and considerably less forgiving of an untreated room.

Every mic goes to its own channel on the interface or recorder, and you record every channel as a separate file. This is the most important technical decision in the whole setup. A single mixed stereo file means one person's cough is permanently over the other person's best answer. Separate tracks mean you can fix each voice independently, and it costs nothing but disk space.

Set levels before you start, with everyone talking at the volume they will actually use - people get louder once the conversation is going, so leave headroom. And record a backup. A second recorder, a phone on the table, anything. Sessions that only exist in one place are sessions you will eventually lose.

Remote guests

Never rely on the call audio. Have the guest record their own local track and send it over afterwards, and use the call recording only as a sync reference and a safety net. Ask them to use wired headphones, sit somewhere soft rather than in a kitchen, and turn off notifications. Five minutes of instruction before the call is worth more than any amount of processing afterwards.

Video podcasts: cameras and framing

Most podcasts are now filmed, largely because vertical clips are how episodes get discovered. That changes the setup requirements.

How many cameras

Three is the working standard for a two-person show: one on each person, and a wide covering both. That gives an editor everything needed to cut a conversation - reaction shots, a clean single for every line, and a wide to reset on. Two cameras works. One camera locked wide is watchable but leaves you almost no editing options and no way to cut around a stumble.

If vertical clips are a deliverable, plan for them rather than cropping later. Either frame the singles loose enough to survive a 9:16 crop, or add a dedicated vertical camera. Cropping a tight 16:9 two-shot into vertical rarely produces anything usable.

Sync and cards

Clap at the top of every session, on camera and on mic. Timecode is better if you have it, but a clap costs nothing and has saved more edits than any piece of equipment. Check card space and battery before each recording block, not after.

Lighting for a video podcast

The aim is flattering, consistent and unnoticeable. It should look like a well-lit room, not a photo shoot.

  • A soft key on each person, slightly off to one side and above eye level. Large and soft beats small and bright every time.
  • Gentle fill on the shadow side so faces do not go half-dark - a bounce board is often enough.
  • A little separation from behind or above so heads do not merge into the background.
  • Something in the background with depth. A practical lamp, a shelf, a colour wash. A flat wall directly behind someone reads as a hostage video.
  • Match your colour temperature across all sources. Mixed daylight and tungsten gives you two differently coloured faces and a grading job.

One quiet warning: LED units with fans are common, and a fan two metres from a microphone is an audible noise floor for the entire episode. Test with the lights on before you commit.

The mistakes that cannot be fixed later

  • Recording in an untreated, echoing room. Reverb is baked in permanently.
  • Clipping. Audio recorded too hot is distorted for good - record conservatively and raise it afterwards.
  • One mixed track instead of separate ones. Removes every option you would want in the edit.
  • Air conditioning or a fan running under the whole session. Noise reduction will get some of it out and make voices sound underwater doing it.
  • No backup recording. The failure is rare and total.
  • Nobody wearing headphones, so a fault that started in minute three is discovered in minute ninety.

How much time to book

A one-hour episode is not a one-hour booking. Allow set-up, level checks, a guest who arrives late, a conversation that runs long, and a clean pack-down. If you are recording several episodes, batch them - the setup is done once and each additional episode costs only its own runtime plus a short reset.

That is why the podcast room is priced by the hour rather than in day blocks. Booking three hours and recording three episodes is a far better use of money than three separate sessions. Rates are on the rates page.