Podcast Color Grading: How I Match Cameras and Keep Every Episode Looking the Same

Podcast color grading is two jobs stacked on top of each other: correction, which makes every camera match and look neutral, then grading, which puts a deliberate look on top. Most shows only ever needed the first one done well.

Here is what goes wrong without it. Your A cam sits near a window and drifts warm as the sun moves. Your B cam is a different brand with a different sensor, so it reads greener in the shadows. Cut between them and your guest changes color every eight seconds. Viewers do not think "the white balance is off". They think the show feels amateur and they leave.

Correction is worth more than any look you buy. I would rather spend forty minutes matching three cameras than five minutes dropping a moody teal LUT on a mismatched mess. Once the cameras agree, a look takes minutes.

This matters most on multi-cam editing, where every cut is a chance for the color to jump. On a single camera show the risk is smaller but not zero, because light changes over a 70 minute record.

I match cameras by picking one hero camera and pulling everything else toward it, in a fixed order: exposure, then white balance, then contrast, then saturation, then the shadow tint.

The hero is usually the camera with the best light on the host, not the most expensive body. If the A cam is properly exposed and the B cam is a stop under, I lift B rather than crushing A, because dropping a good image to meet a bad one makes the whole show worse.

Steps I actually run:

1. Set every camera to the same color space and transform before touching a wheel. Mixed log profiles that are not normalised are the single most common reason podcast grades look muddy.
2. Balance exposure on a single reference frame per person, ideally a wide where you can see the forehead, cheek and shirt.
3. Neutralise with the eyes and teeth. Whites in the eyes should read close to neutral, teeth slightly warm. It is a faster reference than any grey card nobody remembered to shoot.
4. Check skin tone against a vectorscope skin line. If the guest has darker skin, I do not push saturation to compensate, I lift the luminance in the mids and let the hue sit where it belongs.
5. Cut the episode and watch it end to end. Matching on stills lies. Matching on cuts is the truth.

Studio at $299 per episode covers 2 to 4 cameras in 4K, and unlimited revision rounds, which matters because color is the one thing hosts change their mind about after seeing it on a phone.

Use a LUT as a starting point, never as the finish. A LUT is a fixed math transform. It does not know your guest is backlit, and it cannot see that your B cam is green.

Where LUTs earn their keep is consistency across a season. I build one show LUT after grading the first three episodes properly, then apply it under per-episode corrections. Episode 12 then starts 80% of the way there and the editor spends the time on the shots that need attention.

What I avoid: heavy film emulation LUTs on talking heads. Strong halation and lifted blacks look great on a landscape and terrible on a face in a room with two lamps. Podcast color should be invisible, not stylish. Your audience is watching a conversation, not a music video.

If the show already has a visual identity, the grade should follow it rather than fight it. That usually means matching the palette in your intros and branding so the cold open and the body of the episode do not feel like two different shows.

Clips get judged in under two seconds, on a phone, at half brightness, usually in daylight. That is a brutal test for a flat, underexposed frame.

A vertical clip that is a touch brighter and a touch more contrasty than the full episode will hold attention longer. I grade clips as their own pass, not as a crop of the master. Faces get a small lift, backgrounds get a small pull back, and captions stay readable over whatever is behind them.

That pass carries over into LinkedIn podcast clips, where the feed background is white and a dark clip disappears into the page, and into podcast highlight reels, where six moments from six different lighting setups have to sit next to each other without a jarring shift.

Vertical clips are $25 each as an add-on, and that includes the clip specific grade rather than a straight export of the master.

Most podcast color problems are captured, not created in post, and most of them are cheap to stop.

Auto white balance left on. The camera hunts every time someone leans forward and the whole frame breathes. Lock white balance before you record.

Mixed light sources. A warm practical lamp behind the host and a cold key on the face means one of them will be wrong. I can split the difference in post with a mask, but the honest fix is to gel or replace the lamp.

Recording in a log profile with no exposure discipline. Log needs more light, not less. Under exposed log turns into noisy shadows and noise reduction softens the face.

One guest lit, one guest not. Remote guests are the worst offender here. I can lift a dim guest by about a stop and a half before it falls apart. Past that, the fix is a $30 lamp on their desk.

Over sharpening in camera, then again in post. It makes skin look like sandpaper on a 4K TV.

A lot of this shows up in the packaging too, which is why the thumbnail and the grade should be decided together. My sister studio ClickTheory does the thumbnails at 8 for $250, and we pull the thumbnail frame from the graded master rather than a raw screenshot.

Color is included in the edit, not billed separately. 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 cover up to 60 minutes as the base, with shorter episodes up to 30 minutes cheaper and longer ones up to 90 minutes more.

That compares with a typical agency at $300 to $400 per episode, plus roughly $400 a month once you add clips. I will concede the obvious: an agency with a dedicated colorist and a calibrated suite will beat me on a high end brand film. For a weekly conversation show, that ceiling is not the thing that decides whether people watch.

Buy in blocks and it drops further. A 4-episode pack saves 15%, monthly saves 25%. Show notes and chapters are $35, rush 24-hour delivery is $75. Full numbers live on the pricing section.

You own everything. Footage, project files and exports stay yours, including the grade nodes and any show LUT I build, so you are never locked to me. That matters more for B2B podcast editing where a brand team may want the assets in house later.

Color does not rank a video. It does change the two metrics that do move: click through and retention.

A thumbnail pulled from a graded frame reads cleaner at 210 pixels wide than one pulled from a flat log frame. And a viewer who does not notice anything wrong in the first thirty seconds stays longer, which is the signal the platform actually reads.

The rest of the discoverability work is text and structure, which I cover in the podcast SEO guide: titles, chapters, descriptions and transcripts. Color is the part that keeps people once the text has done its job.

One practical habit. Pick your look in the first month and stop changing it. A show that looks the same every week builds recognition in the feed. A show that changes its grade every six episodes throws that away for no gain.