White Balance & Colour Correction for Real Estate Interiors

Why colour accuracy matters

Buyers decide what a property is before they walk through the door, and they decide it from the photos. Wrong colour poisons all of that. An orange cast makes a perfectly good benchtop look dated. Fluorescent green turns a bathroom clinical. And a living room full of blue-grey shadows reads as cold and unlived-in, no matter how warm it feels standing in it.

It is not only an aesthetic problem. When the paint, the timber and the carpet look nothing like the photos, trust goes immediately, and the buyer starts wondering what else was dressed up. Agents report the same thing: accurate colour means fewer awkward moments at the inspection, and a shorter run from first enquiry to offer.

All of it starts with white balance, the setting that tells the sensor what "white" means under whatever light you happen to be standing in. Get that right in camera and roughly 80% of your colour problems never reach the edit.

Colour temperature and the Kelvin scale

Every light source has a colour temperature, measured in Kelvin (K). Low numbers are warm and orange. High numbers are cool and blue. White balance is your camera compensating for that, so a wall that looks white to your eye also looks white in the file.

The sources you will actually meet inside a house:

Interiors almost never have one colour temperature. A living room might be running 2700K lamps, 5500K daylight through the windows and a 4000K LED downlight in the ceiling, all landing on the same wall at the same time. That is where most colour problems in property photography begin.

The mixed lighting problem

When several light sources at different colour temperatures hit the same room, no single white balance setting is correct everywhere in the frame. Not "difficult". Not "a compromise". There is no correct answer, and it helps to accept that before you start fiddling.

Take a bedroom in the afternoon. Cool 5500K daylight comes through the window and puts a neutral-to-blue tone on the bed and the floor near the glass. A bedside lamp running a warm 2700K tungsten bulb throws orange across the headboard and pillows. A ceiling LED at 4000K covers the far wall in a slightly cool white. Set white balance to 5500K and the daylight is right while the lamp corner goes orange. Set it to 2700K and the lamp is right while the window turns blue. Both cannot be accurate at once.

Four ways photographers deal with it:

• Kill every artificial light and shoot daylight plus flash, with the flash matched to daylight at 5500K. That removes the mixed temperature problem outright. It is the logic behind flambient photography, where a flash frame and an ambient frame get blended in post
• Change the bulbs. Some photographers carry a bag of 5000K LEDs and swap out the mismatched lamps on arrival. Slow, but it works in a small property
• Split the difference in camera at roughly 4500-5000K, between daylight and tungsten, shoot RAW, and correct each zone separately with localised adjustments in Lightroom
• Hand it to AI photo editing tools, which analyse each light source zone independently and correct per zone. It is what a good editor does by hand with luminosity masks and local adjustments

Camera white balance settings

Which mode you choose decides how much post-processing you are signing up for later.

Auto white balance (AWB)

Convenient, and close to useless for this work. The camera judges every frame on its own, so two shots taken seconds apart in the same room can land on different casts, and that inconsistency turns into a real problem the moment you batch edit the listing. AWB also over-corrects warm interior light, stripping out the exact warmth that made the room feel inviting. Use it casually, or not at all.

Preset modes

The presets (Daylight, Cloudy, Tungsten, Fluorescent, Shade) each lock a fixed Kelvin value. Daylight, around 5200K, is a solid default for rooms where big windows do most of the work. Tungsten, around 3200K, is for rooms lit entirely by warm incandescents with no daylight at all. The Fluorescent preset, around 4000K, adds a slight magenta shift to cancel the green from fluorescent tubes. The real advantage over AWB is not accuracy, it is that a preset holds still across every frame in a room, which is what makes batch editing possible.

Custom Kelvin

Most mid-range and professional bodies let you dial an exact Kelvin value. For mixed-lighting interiors, 4500-5000K is a sensible starting point: it splits the difference between warm interior lights and cool daylight and gives you a slightly warm file that is easy to nudge in post. Once you find a number that works in a particular house, lock it and shoot every room on it. Consistency across the listing matters more than perfection in any one frame.

Grey card and ColorChecker

Nothing beats a reference frame with a grey card or an X-Rite ColorChecker Passport. At the start of each room, or whenever the light changes noticeably, hold the card in the scene and take one shot. In post, hit the card with the eyedropper to set an accurate white balance, then synchronise that across every frame from that room. The guesswork is gone.

A ColorChecker takes it further with standardised colour patches, so you can build a camera profile that corrects colour accuracy across the whole spectrum rather than just white balance. If you shoot several properties a week, the 30 seconds per room this costs you comes back many times over in editing time.

Correcting white balance in Lightroom

Careful in-camera settings still leave most property photos wanting a nudge in post, and Lightroom is where nearly everyone does it.

Temperature and tint

Temperature runs the blue-yellow axis. Right, to a higher Kelvin, warms the image. Left cools it. Shoot an interior at 4500K in camera and you will usually end up somewhere between 4200K and 5200K depending on which source dominates. Tint runs the green-magenta axis, and it is the one people forget. That is where fluorescent and LED casts get fixed: a green-tinted frame off fluorescent tubes normally wants +10 to +25 toward magenta.

Fastest route is the white balance eyedropper. Click something that should be neutral grey or white, a wall, a grey carpet, a ceiling, and Lightroom works out Temperature and Tint for you. Just do not click near anything strongly coloured. A white wall beside a red curtain is carrying red spill, and you will get a reading that sends the whole image the wrong way.

The HSL panel

HSL (Hue, Saturation, Luminance) gives you per-channel control, which is how you fix things Temperature and Tint simply cannot reach. In interiors it comes down to a handful of moves. Pull Orange saturation back by -10 to -20 when a timber floor is shouting. Shift the Green hue toward yellow to kill the sickly cast off cheap LEDs. Drop Blue luminance by -5 to -15 to settle the blue shadows that show up in daylight-heavy rooms.

Localised colour adjustments

In a badly mixed room, global adjustments will not save you. Lightroom's masking tools (radial gradient, linear gradient, brush) let you correct different zones differently. The usual approach is a gradient mask over the window cooled by -500K to -800K, and a warming brush mask at +300K to +500K over the artificially lit part of the room. It works. It also adds 3-5 minutes an image, which is why this is the part of the job where AI-powered tools save the most. They run the same zone-by-zone analysis and apply the corrections without you drawing a single mask.

Room by room

Every room fails in its own way. Knowing which failure is coming lets you set the camera before you unpack the tripod.

Kitchens

Modern kitchens usually run cool-white LED downlights in the 4000-5000K range, and those sit close enough to window daylight that the balance is easier than elsewhere in the house. The trouble is the surfaces. Glossy benchtops, stainless appliances and tiled splashbacks bounce colour off every source in the room and leave you with local casts that no global slider will touch. Set 4800-5000K, then go looking for orange reflections in the stainless from any warm accent lighting.

Bathrooms

Bathrooms are the worst rooms in the house for this. Vanity lights run warm at 2700-3000K to flatter whoever is at the mirror, while the ceiling might be cool fluorescent or LED. The space is small, so both sources matter, and the tiles throw every cast straight back at you. White tile is a colour amplifier: a tint you would never notice on a painted wall becomes obvious. Start at 4000K and take the tint slider toward magenta, +5 to +15, to counter the green that bathroom-grade fluorescent tubes almost always bring.

Living rooms

This is the mixed lighting problem at full strength. Big windows push 5500K daylight into one side while table and floor lamps push 2700K into the other, and the middle of the room sits in the gradient between them. For HDR bracket shoots, fix white balance at 4500K. On a single exposure, either switch the lamps off and work with daylight plus flash, or keep the lamps deliberately and balance for the daylight. That second option is more defensible than it sounds. Warm lamp pools against cool daylight can look genuinely natural, as long as neither end is running away with the frame.

Bedrooms

Bedrooms lean warm and rarely have much cool light in them. Bedside lamps at 2700K, warm-white pendants at 3000K, not much window light and often what there is comes through sheers. The lighting is fairly uniform, which makes this the easiest room in the house to balance. Set 3500-4000K and you will usually get something pleasing with very little work after. Do not chase the warmth all the way out, though: a slightly warm bedroom reads as cosy, and a perfectly neutral one reads as a display suite.

Common mistakes, and batch editing

Experience does not make you immune to any of these. It just makes you faster at spotting them.

• Over-warming. Pushing warmth to make a room feel "cosy" drags the whites into yellow and the whole image starts to look like it was shot through an amber filter. White walls should read white, not cream. If they are going yellow, your Temperature slider has gone too far
• Ignoring the tint axis. Fluorescent and cheap LED lighting lays down a green cast that Temperature alone will never fix, no matter how far you drag it. You have to move Tint toward magenta as well. Photographers who only touch temperature end up with a whole listing that looks faintly unwell
• Blue shadows. In strong daylight, shadows pick up blue from the sky. It is physically correct, since open shade sits at 7000-8000K, and it still looks bad in a property photo. Brush a mask over the shadow areas and warm them +300K to +500K, or pull Blue luminance down in the HSL panel
• Drift across the listing. Each frame can be acceptable on its own and the set can still look amateur, because a buyer scrolling the gallery absolutely notices when the kitchen is cool and the bedroom is orange

Batch editing for consistency

Consistency across a listing counts for as much as accuracy in any one frame, and Lightroom's presets and synchronisation exist for exactly this. Edit one image per room type until the colour is right, then sync those settings across every image shot in the same room under the same light.

Build yourself a small set of presets for the situations that keep recurring: daylight-dominant rooms at 5200K with Tint +5, warm-lit rooms at 3800K with Tint +8, fluorescent-lit spaces at 4200K with Tint +20, mixed lighting at 4600K with Tint +10. Apply the closest one, then fine-tune with 2-3 sliders. That takes per-image colour correction from 3-5 minutes down to 30-60 seconds, and it holds the 25-40 images in a listing together visually, which is the part that actually shows.

How AI handles white balance

The expensive part of colour correction was never one image. It is doing the same thing consistently across a whole shoot, which is the bit AI editing has actually changed.

Traditional correction treats the frame as one zone. One Temperature value, one Tint value, applied evenly. That is precisely why mixed lighting defeats it: one setting cannot be right for areas lit by different sources. Zonal correction with Lightroom masks or Photoshop layers does solve it, at the cost of minutes per image and an eye trained to spot where one light stops and the next begins.

AI correction goes at it differently. Tools like Pixestate read each image for the light sources present and their rough colour temperatures, then segment the frame into zones by whichever source dominates each area and correct each zone on its own terms. The daylight near the window gets one treatment, the lamp-lit corner gets another, and the transition between them is blended.

It bites hardest on exactly the situations that cost you the most time by hand: living rooms with warm lamps against cool window light, bathrooms where the vanity and the ceiling do not match, and open-plan spaces where the kitchen, dining and living zones are each lit by something different but all land in one wide-angle frame.

The real gain is consistency at volume. Run 30 images from a shoot through it and the last one gets the same analysis and the same standard of correction as the first, whatever the lighting was doing in that room. Your judgment drifts over a long editing session. Software's does not. If you are shooting several properties a week, that removes the most tedious stretch of the whole post-production run.