Podcast Lighting Guide: Setups That Look Good and Edit Fast

The best podcast lighting setup is one key light per person at 45 degrees off camera and slightly above eye level, a soft fill on the opposite side at about half the key's brightness, and one background light so the room behind you is not a black hole. That is it. Three lights per person maximum, and two is usually enough.

The key does the work. Put it high enough that the shadow under the nose points down, not sideways, and close enough that the light source looks big relative to the face. A 60cm softbox at one metre is soft. The same softbox at three metres is a hard, unflattering point. Distance matters more than wattage.

The fill is not there to remove shadow. It is there to keep shadow inside the range the camera can record. If I can see detail on the shadow side of your face, I can grade your episode in one pass instead of three. When the fill is missing, the dark side crushes to pure black, and no amount of lifting in post brings it back without noise.

Background light is the one people skip and the one that makes footage look professional. A single lamp behind you hitting a wall, a shelf, or a plant separates you from the set. It costs about $40 and it is the cheapest upgrade to perceived production value I know of. If you want the full order of operations from record to publish, I lay it out in my podcast workflow guide.

Spend $150 to $400 total for a two person setup and stop there. I have graded footage from $3,000 cinema fixtures and from $60 panels off a marketplace, and past a certain point the difference in the final file is smaller than the difference a repositioned stand makes.

Here is where the money actually helps. Higher CRI means skin tones land where they should instead of going green or magenta, and that saves me a correction step per camera. Flicker free operation matters if you shoot at 50fps or higher, because cheap LEDs pulse and the pulse shows up as banding that cannot be removed. Bi-colour control lets you match your lights to whatever daylight is leaking through the window rather than fighting it.

What does not help: more lights, bigger lights, and RGB colour effects. I get episodes lit with six fixtures and four colours, and they take longer to cut because the colour changes as people lean forward and back. Two clean lights beat six clever ones every single time.

Budget the money you saved into the parts of the show people notice. A proper intro and branding package does more for retention than a fourth light. So does a thumbnail people actually click, which is why our sister studio ClickTheory does 8 thumbnails for $250 rather than us pretending a nice rim light will do that job for you.

Lighting matters most on multi camera podcasts because every angle has to match, and if it does not, every cut looks like a mistake. On a single camera show a slightly warm image is just a look. On a three camera show, one warm angle and two neutral ones means the viewer sees a colour jump every time I cut, roughly every 8 to 15 seconds.

The fix is boring and it works. Set your white balance manually on every body to the same Kelvin value, usually 5600K if you have daylight in the room and 3200K if you do not. Do not use auto white balance. Auto shifts mid sentence when someone in a white shirt leans into frame, and that shift is baked into the file forever.

The second fix is to check your wide angle before you roll. Wide cameras see the light stands, the window, and the doorway your key light was hiding. I have received plenty of episodes where the close ups are beautiful and the wide shows a bare bulb in the corner. On a Studio plan episode with 2 to 4 cameras I am grading every one of those angles, so consistency across them is the single biggest thing you control.

If your show runs more than one camera, the practical camera and lighting decisions are worth reading together in my notes on multi-cam editing.

Overhead room lights left on. Ceiling lights put shadow in the eye sockets and a hot spot on the forehead. Turn them off. Your key light replaces them.

Glasses glare. This is the one that costs the most in post, because a bright rectangle on a lens cannot be removed, only cropped around. Raise your key higher and push it wider until the reflection slides off the lens. Ask the guest to tilt their glasses down a few degrees at the temples. Check it on camera, not by eye.

Mixed colour temperature. A warm lamp on one side and a cold window on the other means one half of the face is orange and one half is blue. Close the blind or match the lamp. Skin tone is the thing viewers judge, even when they cannot name what is wrong.

Light too far back. If the key sits behind the camera pointing straight at the face, the image goes flat and the nose disappears. Move it 45 degrees to the side and you get shape back for free.

Exposing for the room instead of the face. Cameras average the frame. If the wall behind you is bright, the camera darkens you to compensate. Expose for the skin and let the background fall where it falls. Underexposed faces are the number one reason footage looks noisy, and noise is what makes an episode look cheap on a phone screen.

You light a remote guest by sending them three instructions before the call, not by fixing it afterwards. Post cannot invent light that was never recorded. This is the single biggest quality gap on interview shows, and it is entirely solvable in a two minute pre call.

Instruction one: face a window, do not sit with your back to it. Backlit guests turn into silhouettes, and the platform's auto exposure then blows out the window trying to compensate. Instruction two: put the laptop on a stack of books so the camera is at eye level. A camera pointing up at a chin is a lighting problem as much as a framing one, because the light in the room is above and you are shooting into shadow. Instruction three: turn off the ceiling light and put any lamp you own beside the screen, not behind you.

If the guest is a repeat contributor or the show is built on interviews, a $40 clip on panel posted to them pays for itself in the first episode. I have never regretted telling a host to do that.

When guest footage is still uneven, the honest move is to lean on the good angle and cut around the bad one, which is a judgement call a human makes and an automatic tool does not. I get into the reasoning in human editing versus AI tools.

Good lighting does not make your show interesting, but bad lighting gives people permission to leave. Viewers decide in the first few seconds whether a video looks worth their time, and brightness, contrast and colour are what they read first, before they process a word you said.

The compounding effect is in clips. A 30 second vertical clip is watched on a phone at arm's length, often outdoors, often at half brightness. Footage that looked acceptable on your studio monitor turns into mud on that screen. Contrast survives compression. Detail in the shadows does not. When I cut vertical clips at $25 each, the clips that perform are almost always the ones where the face is clearly separated from the background.

This is also why background light earns its place. On a cropped vertical frame the background is most of the image. A lit wall with some depth reads as a set. An unlit wall reads as a spare room.

If you want to know whether any of this is moving the needle on your own show rather than taking my word for it, the numbers to watch and the ones to ignore are in my podcast analytics guide. And if you are cutting a launch piece, lighting consistency matters double, because a trailer intercuts clips shot weeks apart, which I cover in trailer editing.

Once you send the footage, a human editor watches the whole episode. Nothing is auto cut by silence detection, which matters for lighting because a silence detector does not know that the take where the lamp fell over is the take with your best answer in it.

On colour, I match every camera to a single reference frame, correct exposure and white balance per angle, then apply one grade across the episode so it reads as one room. If your lighting was consistent, that is a fast pass. If it was not, it is a per clip job, and per clip work is where turnaround pressure comes from.

Starter is $199 per episode for a single camera at 1080p with one revision round. Studio is $299 per episode for 2 to 4 cameras at 4K with unlimited revisions. Both are priced per episode up to 60 minutes as the base, with shorter episodes up to 30 minutes costing less and longer ones up to 90 minutes costing more. A 4-episode pack saves 15% and monthly saves 25%. Add ons are vertical clips at $25 each, show notes and chapters at $35, and rush 24 hour delivery at $75.

First cut comes back in 48 hours and revisions land the next working day. You own everything: footage, project files and exports are yours, with us keeping a copy alongside you. The full step by step is on how it works and the numbers live on pricing.