Fix Shot Matching Across Cameras for Real Estate | Leumos AI

Shot matching across cameras in real estate videography breaks down in three places: exposure drift between Sony A7C interiors and Mavic 3 aerials, mixed tungsten and daylight rooms that defeat single-LUT pipelines, and the 200+ clips per property that compound every micro-mismatch into visibly uneven listings. The fix is a strict order of operations — equalize color space first, match tonality second, grade creatively last.

I've been a DaVinci Resolve Certified colourist for four years, and the real estate videography reels I've graded for friends running their own shoots have a recurring failure mode I see in maybe seventy percent of self-graded work: the drone shot is teal, the interior is muddy yellow, and the exterior establishing shot looks like a different listing entirely. Same property. Same day. Three cameras, no consistency.

This page is the diagnostic I'd run if you sent me a folder of 200 unmatched clips at 11pm with a noon delivery. It's not magic — there's a specific sequence that fixes about eighty percent of the problem before you touch a single creative slider. The other twenty percent is taste, and we'll cover that too.

Why Your A7C, Pocket, and Mavic Don't Match in the First Place

Three cameras, three sensors, three color sciences — even when you shoot the same scene at the same time. The Sony A7C and FX30 share Sony's color science but the FX30's APS-C sensor renders skin and warm tones slightly differently than the A7C's full-frame Exmor R. The DJI Pocket 3 uses a 1-inch CMOS with DJI's own color profile baked in. The Mavic 3's Hasselblad-tuned Four Thirds sensor produces yet another rendering of the same light. You're stitching three color philosophies into one continuous tour.

Now add the real killer: mixed lighting. A typical reception room has tungsten downlights at 3200K bleeding into south-facing windows at 5600K, bouncing off a beige carpet, with maybe a 4000K LED in the kitchen. Your A7C's auto white-balance picked something. Your Pocket picked something different. Your drone, shot ninety seconds later in pure daylight, picked 6500K. Three different white points, baked into three different files, all supposed to cut together as one continuous home tour.

If you're shooting 4K H.265 with no log profile — which most run-and-gun real estate shoots are — those white points are baked. There's no untangling them without doing real color work. That's the diagnosis. Here's the treatment.

The Three-Step Fix That Actually Works

The mistake almost everyone makes is grabbing a real estate LUT pack off Gumroad and slapping the same .cube onto every clip. That only works if your footage already matches — which is the entire problem you're trying to solve. The order matters:

Step 1: Normalize color space. If you shot S-Log3 on the A7C or FX30 and D-Log on the Mavic 3, transform each to Rec.709 first using the correct input transform per camera. Same destination color space, different input paths. In Resolve that's CST nodes. In Premiere it's the Lumetri input LUT. Either way, do this before you touch anything else.

Step 2: Match tonality. Pick your hero shot — usually a clean exterior or your most-watchable interior wide. Then equalize every other clip's exposure, contrast, and white balance to that hero. This is where most colorists lose three hours per listing: 200 clips, manually adjusting each one to match. It's the bottleneck I built Match All to kill.

Step 3: Apply creative look last. Only now drop your chosen LUT or reference image. Because your clips are already equalized, the same look lands the same way on the drone shot, the kitchen interior, and the backyard pull-back. One look, every clip, consistent listing.

If you're a real estate videographer, we're building this for you. Leumos AI launches in ~30 days — join the early-access list and you'll be in the first 500 (50% off the first year).

Reference Looks: How Boutique Real Estate Group and BoxBrownie Reels Stay Consistent

If you've studied The Boutique Real Estate Group's video brand, you'll notice something specific: every property tour they put out sits in the same warm-clean grade. Slightly lifted shadows, neutral skin where present, a touch of warmth in the highlights, and crucially — the drone establishing shot has the exact same color signature as the master bedroom interior thirty seconds later. That continuity is what makes the brand feel premium. It's not the camera. It's the grade discipline.

BoxBrownie-style luxury property reels operate on the same principle: bright, clean, slightly cool in the highlights, warm in the midtones, no green spill in the windows. Eli Jones's real estate video pipeline is a masterclass in repeatable looks at volume — he's delivering at four to ten properties per week and every reel sits in the same tonal pocket. Iman Gadzhi's properties content leans warmer and more cinematic, but again: the consistency between exterior, interior, and aerial is non-negotiable.

The technique behind all of these isn't a secret LUT. It's a reference frame. Pick one still — a photograph from the listing, a frame from a film you love, a still from a competitor's reel — and grade every clip in your timeline toward it. That's the workflow Reference Image Grading is designed around.

Where AI Color Grading Fits (and Where It Still Breaks)

I've tested every AI color tool on the market over the last twelve months — Colourlab AI, fylm.ai, color.io, Dehancer. They're genuinely impressive at specific tasks. Colourlab AI's shot-matching is the closest thing to a magic button I've used, and fylm.ai's UI for film emulation is excellent. None of them, however, are built for the specific problem of four-to-ten property turnarounds per week from a browser.

What AI does well right now: tonal equalization across cuts of the same camera, reference-image matching when the source and target are both in clean light, and removing the most obvious exposure drift. What AI still struggles with: skin tones under mixed tungsten-and-daylight (the classic real-estate-with-an-agent-on-camera problem), bridging a 1-inch drone sensor to a full-frame interior without color bias creeping in, and any creative day-for-dusk push.

The honest take is that AI gets you about eighty percent of the way to a matched timeline in five minutes, and the last twenty percent is still you with a primaries panel. That's the reality I'm designing Leumos around — automate bulk equalization with Match All and reference matching, but leave Manual Primaries right there so you can fix the one shot where the AI guessed wrong.

A Realistic 24-72 Hour Real Estate Turnaround Workflow

Here's the workflow I'm building Leumos to handle, end to end, in roughly twenty minutes of active work per property:

1. Ingest and organize. Drop your A7C, Pocket 3, and Mavic 3 clips into the upload queue. AI Scene Cut Detection chops continuous walk-throughs into individual shots on a timeline — no manual node-per-clip setup like Resolve. Manual Cut Tool catches anything the AI misses.
2. Normalize input. Apply the correct Input Color Space LUT per camera — S-Log3 for the A7C or FX30, D-Log for the Mavic 3, standard Rec.709 if you shot the Pocket 3 in normal profile. One click each.
3. Equalize. Run Match All once across the timeline. It handles exposure drift, white balance variance, and saturation mismatch between cameras automatically.
4. Apply your brand look. Drop a reference frame from your last delivered listing using Reference Image Grading, or pick from the Preset LUT Library with the intensity slider for fine control. Upload your own .cube if you've already built a house style.
5. Spot-fix. The two or three shots the AI missed — usually a window-blown kitchen or a mixed-light hallway — get cleaned up with Manual Primaries in under a minute each.
6. Export and deliver. That's the loop. Same look on every listing, every week.

The reason I'm building this is that I was tired of node soup. Working colorists and videographers shouldn't be spending Sunday nights at three a.m. matching drone clips to interiors at the cost of their actual life. If you're shooting four to ten properties a week at $250 to $1,500 each, you need a tool that respects your turnaround instead of fighting it.

Frequently asked questions

How long should it realistically take to shot-match a real estate reel across three cameras?

For a 200-clip listing shot on an A7C interior, Pocket 3 walkthrough, and Mavic 3 aerial, a clean Resolve workflow with CST nodes, group grades, and manual matching takes me about 90 minutes. With AI equalization handling the bulk equalization pass, that drops to roughly 20 minutes of active work — the AI does the first eighty percent in under a minute, and the remaining time is you spot-fixing the four or five shots the AI guessed wrong on. That's the math behind a 24-hour turnaround at scale.

Can I get the A7C and Mavic 3 to match in-camera so I don't have to fix it in post?

Partially. Set a fixed white balance on both — 5600K outdoors, 3200K interiors — instead of auto. Match picture profiles where possible: S-Log3 on the A7C and D-Log on the Mavic both transform cleanly to Rec.709 with the right input LUT. But the sensors themselves render color differently; you cannot eliminate the gap entirely in-camera. Aim to minimize it on the shoot so post is shot-matching rather than rescue work. The first ten minutes of every job should be a white-balance card on every camera.

What's the best LUT for real estate videography?

There isn't one. The Boutique Real Estate Group, BoxBrownie reels, Eli Jones, and Iman Gadzhi all use different looks — warm-clean, bright-cool, cinematic-warm respectively — and they all work because each operator picked one and applied it consistently. The better question is which look fits your listings (luxury, family, urban) and your agent's brand. Build or buy one .cube file, apply it as the final step after your footage is already equalized, and use the intensity slider to dial it back on shots where it's too aggressive.

Should I shoot S-Log3 or just standard profile for real estate?

For four-to-ten-property weekly turnarounds, standard profile (PP6 or Rec.709) is usually the right call. S-Log3 gives you more grading latitude but requires a proper exposure pass and a color space transform on ingest, which adds minutes per clip. If you're hand-grading every listing and want the flexibility, S-Log3 is worth it. If you're trying to keep edit time under an hour per property, shoot Rec.709 with a baked LUT preview on the camera monitor and skip the log workflow entirely.

Why do my interiors always look yellow even after I correct white balance?

Almost always because the tungsten downlights are warmer than your white balance setting accounts for, and the camera's auto-WB or your manual 3200K setting is averaging it against window daylight. The fix is to correct in two passes: pull a custom white balance off a neutral card under the dominant interior light source, then use a secondary correction (or AI tonal match) to reduce the cyan-magenta drift the daylight introduces. If you're shooting 4K H.265 with no log, this correction has to happen in software — the white point is baked into the file.

How does AI shot-matching compare to manually grading in Resolve?

Resolve is more powerful and more precise — you have full nodes, qualifiers, tracking, and HDR pipelines. AI shot-matching is faster and good enough for about eighty percent of property listings, especially when consistency across cuts matters more than artisanal taste on every frame. I still use Resolve for festival features and high-end commercial work. For high-volume real estate where you're delivering a recognizable brand look across 4-10 listings a week, AI equalization plus a reference frame is the right tool. They're different jobs.

What does Leumos AI cost, and when can I actually use it?

Leumos launches in roughly 30 days from today. Pricing on launch is Free ($0, 2 uploads/day, 400MB max), Creator ($15/month, 8 uploads/day, 1GB max), and Pro ($39/month, 20 uploads/day, 2GB max) — the Pro tier is sized around exactly the four-to-ten properties per week real estate workflow. The first 500 early-access signups get 50% off the first year, which is the cheapest the product will ever be. Sign up on the coming-soon page and you'll get launch-day access and the discount locked in.

---

Leumos AI launches mid-2026. The first 500 early-access signups get 50% off the first year. Join the early-access list →