Observe your light

A color is always seen in a particular light, at a particular hour. How to observe daylight and bulbs, and what the number on the back of a chip does and does not mean.

1. Light is the medium Understand

Everything you see in a room is light that has bounced off a surface and reached your eye. Change the light and you change what reaches the eye, even though the paint, the sofa and the floor have not changed at all. Physical principle A surface does not have a color independent of the light falling on it; what it has is a way of reflecting whatever light arrives. This is why a chip that looked right in the store can look wrong at home, why a wall that was fine all winter can turn sallow in June, and why the course keeps saying "at the hours you use the room."

Daylight is not one thing. Through a single day it changes in amount (dim at dawn, strongest around midday, falling again toward dusk), in direction (low and from one side in the morning, high at noon, low and from the other side late in the afternoon), and in color (the common observation is that direct low sun leans warmer and that light from a clear blue sky, with no sun in it, leans cooler). It changes with weather: an overcast sky is a huge, soft, fairly even light source, and the same room under it has fewer shadows and a flatter, often cooler look than it has in sun. And it changes with season: the sun is lower in winter and reaches further into rooms, higher in summer and often blocked by an overhang or a tree that is bare in January and dense in July.

None of this is a reason to despair. It is a reason to look, more than once, at the times that matter to you, and to write down what you saw. The eye adapts quickly (Physical principle within a minute or two in a room, your visual system "corrects" for the color of the light so that a white page keeps looking white), which is precisely why memory is a poor instrument and a written log is a good one.

2. Compass direction is a hypothesis

You will meet the orientation rule everywhere: Convention in the northern hemisphere, north-facing rooms get steadier, cooler-leaning light with little direct sun; south-facing rooms get warmer light that changes through the day; east windows are bright in the morning; west windows are warm in the late afternoon. In the southern hemisphere, swap north and south. The convention exists because it is often true, and it is a reasonable place to start. It is not a rule, and this course will not let you use it as one.

What overrides it, routinely: a neighbor's house ten feet from the window; a deep porch roof or awning; a mature tree; a window that is small, or high, or one of several facing different ways; the color of whatever is outside the glass. That last one surprises people. Physical principle Light that bounces off a colored surface takes on some of its color: light off a lawn can arrive green-tinged; light off a red brick wall or a terracotta roof can arrive warmer; light off a white rendered wall opposite tends to be bright and fairly neutral; light off a dark fence is simply less. How much any of this shows in your room depends on the distances, the sun and the season, which is what the log is for. A "north room" with a sunlit brick wall opposite the window may be warmer at 3 p.m. than a "south room" under a porch. Direction tells you where the sun is; it does not tell you what is in the way or what the light bounced off on its way in.

So treat direction as the first guess the log will confirm or overturn. If you do not know your directions, it takes two minutes: open the compass app on a phone, stand at the window facing out, and read the bearing (within 45 degrees is plenty). Or, with no phone, notice where the sun rises and sets relative to the window; a window the morning sun comes through faces roughly east. Write the direction in the log as a hypothesis, then let the entries at your actual hours say whether the room behaves as the convention predicts. Pell Avenue's west windows, in the walkthrough below, do; Wren Lane's single window turns out not to matter much at all, because Joan is never in the room when the sun is.

3. Artificial light: reading a bulb

From dusk onward, and all day in some rooms, the light is whatever you screwed into the fixture. Three numbers on the packaging describe it, and once you can read them you can often fix an evening color problem without touching a paintbrush. They narrow the choice; your eyes, in the room, settle it.

Color temperature, in kelvin

Correlated color temperature (CCT) describes how warm or cool a white light looks, in kelvin (K). Lower numbers are warmer (more yellow-orange); higher numbers are cooler (more blue-white). The U.S. Department of Energy's purchasing guidance groups bulbs into bands of about 2700 K (warm, the look of an old incandescent bulb), 3000 K (soft), 3500 K (neutral) and 4100 K (cool). Those four bands are reported here from the DOE's page on purchasing energy-efficient light bulbs, which also explains the other numbers below. Bulbs sold as "daylight" generally have a higher, bluer color temperature than "warm white" ones; the name does not mean they reproduce a day's changing light, so read the kelvin number on the box. Benjamin Moore's warm-and-cool page makes the same point from the paint side: bulbs above roughly 3500 K read cooler, and a wall color seen under them tends to as well. The kelvin number is a description of the white, not a full one: the DOE's fact sheet LED Color Characteristics (2012, a plain-language explainer of CCT and CRI) notes that two sources with the same CCT can look different, one slightly greenish and one slightly pinkish. Which band you prefer in a living room is a Preference; a common choice for sitting rooms is 2700 K or 3000 K, and sometimes higher where tasks need it, but that is a habit, not a rule.

Color rendering index

CRI is a 0 to 100 score for how faithfully a light source shows colors compared with a reference light of the same color temperature. Incandescent bulbs score 100. ENERGY STAR's former bulb specification required a CRI of at least 80 until that program ended at the end of 2024 (the EPA's lighting sunset memo announces the end); the same DOE page still cites the figure, and many bulbs still meet it. Two bulbs can share a kelvin number and have different CRIs. A higher CRI makes it more likely that reds and wood tones show faithfully, and some makers also print R9, the score for a strong red sample, which the main CRI number does not include. But the numbers do not reliably predict how your sofa or your floor will look. The same DOE fact sheet says that two light sources with the same CCT and CRI may not render colors the same way, that CRI scores should serve only as a rough guideline, and that a source with a lower CRI can sometimes make objects look more pleasing than one with a higher CRI. So use CRI to narrow the choice, then look: put the bulb in the lamp and see the sofa, the floor and the wall at the hour you sit there. Test it Where color matters to you, the course suggests 90 or above as a starting point for that look. Convention That 90 figure is a widely used rule of thumb among people who care about color, not a standard; 80 was the former ENERGY STAR floor, and many bulbs still meet it.

Lumens, not watts

Watts measure the energy a bulb uses; lumens measure the light it gives. An LED produces far more lumens per watt than the incandescent bulbs most of us grew up with, so "a 60-watt bulb" no longer means anything about brightness. Buy by lumens (the DOE page gives rough equivalents for older wattages). More lumens in the right lamp is often the cheapest way to make a dim room usable, and no paint color does the same job, which section 5 returns to.

Three more things on the box

• "Warm-dim" or "dim to warm." An ordinary dimmable LED gets dimmer but stays the same kelvin. A warm-dim LED drops in color temperature as it dims, the way an incandescent did, so a lamp turned low in the evening also gets warmer. If you liked the old bulbs for that, this is the feature to look for; the DOE's page on LED color-tunable products describes warm-dim and the related "tunable white" products whose color temperature can be adjusted. Product-specific Whether a given bulb does this is on its packaging, not something you can tell from the glass.
• Dimmer compatibility. Not every LED works on every dimmer; a mismatch flickers, buzzes or will not go fully low. The DOE page says to check the packaging for dimmer compatibility, and to match the bulb to the dimmer you have.
• Match what is there. When replacing one bulb in a room, match the existing color temperature (the DOE page says so plainly), or replace them all. Since bulbs with the same printed numbers can still look a little different, the same model throughout is the simplest way to keep one room's light alike. One 5000 K bulb in a room of 2700 K bulbs is section 4's problem.

Where to read it

The packaging carries a standard "lighting facts" style panel with the lumens, the color temperature (often as a warm-to-cool scale with the kelvin marked) and the wattage, plus the dimmable and warm-dim claims nearby. CRI is not always on the front; look on the side or back, or search the model number on the maker's page, and if you cannot find it treat it as unknown rather than guessing. For bulbs already in your fixtures, with the box long gone, let the bulb cool, take it out, and read the printing on the base or the neck: most carry the kelvin ("2700K"), the lumens and a model number, and the model number finds the rest online. Five minutes with a chair and a cloth, and the log's bulb sheet is filled.

4. Mixed color temperatures: "something is off"

If a room looks wrong in the evening and you cannot say why, count the bulbs before you blame the paint. A ceiling fixture at 5000 K and a table lamp at 2700 K throw two different whites onto the same wall; the wall near the lamp can go yellow, the wall under the ceiling light blue-gray, a cream rug can read as two colors, and the white paper you hold up may look different depending on which way you turn. Physical principle Your eye adapts to one color of light well; it cannot fully adapt to two at once, so the difference tends to stay visible. This is a common cause of "something is off," and it is one of the cheapest to fix: bulbs cost roughly a few dollars each, and even a renter who cannot touch a wired fixture can usually change what is in it.

The order of operations matters. If you are going to change the bulbs, do it before you sample paint, so that every observation is made under the light the room will actually have. Sampling under the old mixed bulbs and then replacing them means sampling twice. Wren Lane, below, is the worked example: Joan standardizes the bulbs first, looks again at her evening hours, and only then brings chips into the room.

5. Light reflectance value, correctly

On the back of many paint chips, and on most manufacturers' color pages, you will find a number called LRV, the light reflectance value. It is widely quoted and widely misunderstood, so here is what it is. Product-specific LRV is a percentage, from 0 to 100, of the visible light a paint color reflects, as measured by the manufacturer for that product; it is a measured quantity (a British standard, BS 8493, exists purely to specify the test method), and the manufacturer's figure applies to their paint, not to a color name in general. Where it is printed varies by brand: Sherwin-Williams' white-paint article uses LRV as a comparison aid between its whites; the consultant guide reports that Benjamin Moore lists it on each color's page and Sherwin-Williams prints it on the back of chips; check the chip or the maker's color page.

LRV is closely related to value from Module 2: a high LRV is a light color, a low LRV a dark one, and the number lets you compare the depth of two chips from different brands without squinting. That is its real use. Now the list of things it is not, because every one of these mistakes is common:

• Not warmth. A warm off-white and a cool off-white can share an LRV of 80. The number says nothing about which way a color leans.
• Not undertone. Two grays at LRV 55 can lean green and violet respectively. Only a side-by-side comparison (Module 3) shows that.
• Not bulb output. LRV describes what a surface reflects; lumens describe what a lamp sends out. They are measured by different people of different things.
• Not a guarantee of brightness. A paint reflects a percentage of the light it receives. If little light arrives, a high-LRV wall reflects a high percentage of little, and the room is still dim. As the consultant blog Kylie M Interiors: Paint colors and LRV (a practitioner's guide to reading the number) puts it, "No paint color will save you if you don't have enough natural or artificial light." Adding lumens, or a lamp where there was none, does what no chip can.

Practitioners do group LRVs roughly by depth, and it helps when reading a chip rack to have a sense of the scale. Convention One practitioner's rough grouping (the Kylie M Interiors guide is a typical example) runs: very roughly, the 70s and above read as light to off-white; the 50s and 60s as light-to-medium; the 30s and 40s as medium; below that, dark. The boundaries are a habit of speech, not a standard, and the same color in a dim room and in a sunlit one can feel like two different depths. No LRV cutoff decides a room's color for you; the course's position is that what you see in the room at your hours decides, and that nothing in the number says a small or dim room must be light.

6. The hours that count, and why photos are not the room

A color has to work when you are there. If you leave at 8 a.m. and come home at 7 p.m., the living room's noon light is irrelevant to you, and a color chosen at noon has been chosen for someone else. So the log is built around use hours: the morning hour if you have breakfast there, the evening hours if that is when you sit, the whole day for a room where someone works. Setup asked when the rooms are used; the "Your home" box below turns that into the hours to log. If the answer is "mostly evenings," then the bulbs, not the window, are most of your light, and the five-minute bulb check may be the most useful step in this module.

About photographs. A phone camera does not record the light; it guesses it. Its automatic white balance looks at the scene, decides what it thinks should be neutral, and shifts every color so that it is. Point it at a warm-lit room and it usually cools the picture toward what it assumes you wanted; point it at the same wall by a blue-sky window and it may warm it. Two photographs of the same wall an hour apart can disagree with each other and with your eyes. The course therefore uses photographs for exactly two things: value checks (switch a photo to black and white to see light-and-dark relationships, as in Module 2) and records (a picture of where a sample board stood, with the date). Never for a color decision, and never as a substitute for standing in the room.

7. See it: the same wall at four hours See it

The first demonstration is deliberately unlike most "see your color in different light" tools. The wall is one screen color, #cfc6b6, in every panel, and the course has not altered it. Only the caption and a small sky or lamp indicator change. That is the point: the course will not simulate light on a screen, because a simulation would be a fake with the authority of a picture, and you would learn to trust it instead of your eyes. What the panels give you is a checklist: at each hour, what to look for in the actual room.

The second demonstration shows what LRV does describe. Four swatches of one gray at four depths, labelled with the kind of LRV each might carry if it were a product (it is not; these are screen grays). Read left to right and notice that nothing about warmth or undertone changes; only depth does.

8. Practice: the light log Practice

Time: about 30 minutes in all on one day (5 minutes for the bulb check, then about 10 minutes per look), plus optional extra days. Materials: a sheet of plain white paper and a mid-gray card (the same two sheets every time), a phone for notes, and the light observation log. Do it in the first-project room. Nothing is bought.

The minimum is the bulb check, one look in daylight and one look at each hour the room is mostly used, with a line or two noted each time. For most decisions that is enough to go on. A longer log (several days, an overcast day and a sunny one, more than one wall) is worth it when a decision is hard, expensive or contested, or when your first looks disagree with each other; step 6 describes it, and it is optional.

1. Do the bulb check first. For every fixture and lamp in the room, read the kelvin, CRI and lumens from the base of the bulb or its packaging, and note whether it is dimmable. Enter each one on the log's bulb sheet, with whether you can change it. If two kelvin values appear in one room, decide now whether you will standardize before the log, and if so, do it before step 3.
2. Find the window directions with the phone compass or by sunrise and sunset. Write them in the log as a hypothesis, and note anything outside each window that could color or block the light: a lawn, a brick wall, a tree, a neighbor's wall, an awning.
3. Daylight entry. Stand in the room with the lamps off. Hold the white paper against the main wall, then the gray card. Write what each looked like compared with how it looked in the store or in another room ("paper slightly blue," "gray card looks greenish"). Note the sky (sunny, partly cloudy, overcast), where the light is strongest, whether any patch of wall is in direct sun, and whether the room is in use at this hour. If the room is used in daylight, make this look at that hour.
4. Entry at your actual use hour, with the lamps as you normally have them. Same paper, same card, same wall. If you use the room in the evening, this is the entry that matters most. Note which lamp was nearest, and whether the paper changed color as you moved around the room (that is the mixed-bulb sign). If the room is used at more than one time of day, make one look at each.
5. Read it back. Which entries disagree with each other? Which hours are the ones you are actually in the room? Those hours are where every sample in Module 9 will be judged, and the log entries are the baseline you will compare the samples against.
6. Optional: the longer log. If the choice is hard or expensive, the household disagrees, or your looks contradicted each other, repeat the entries over a few days, try to catch one overcast day and one sunny one, and add a morning and a midday look. If the weather will not cooperate, log what you get and mark the gap; a log with honest gaps beats one with guesses.

If you cannot easily stand for ten minutes, or kneel, do the entries from a chair with the paper and card on a clipboard against the wall beside you; the log needs your eyes at your usual height, which for most people in an evening room is seated anyway.

9. Your home Your home

10. Check your work Check your work

Criteria

• The bulb sheet lists every fixture and lamp in the room with a kelvin value and, where printed, a CRI; "unknown" is written where it is unknown, not a guess.
• The log has at least one daylight entry and one entry at each hour the room is actually used, each with sky, lamps, and what the white paper and gray card looked like. Extra days are optional, for hard or expensive decisions or looks that disagreed.
• Window directions are written as hypotheses, and at least one entry says whether the room behaved as the convention predicted.
• Reflected-light sources outside each window are noted, or "none noticed" is written.
• If the room has mixed color temperatures, the log says what you will do about it and whether it was done before or after the entries.
• No entry relies on a photograph for a color judgment.

Worked example

Theo at Pell Avenue (fictional) chose the longer log, because the west sun made the room change so much, and logged six entries: Saturday 9 a.m. overcast, lamps off, paper "slightly gray, flat," carpet "cool brown"; Saturday 1 p.m. partly cloudy, "paper white, room even"; Saturday 5 p.m. sunny, "paper warm, almost cream, carpet orange-brown where the sun hits, cooler in the corner"; Saturday 8 p.m., lamps on, "paper warmer by the 2700 K lamp, cooler by the unmarked one; the two sides of the room disagree"; Monday 7 a.m. sunny, "cool, dim, sky light only, paper faintly blue"; Monday 8 p.m., after the unmarked bulb was replaced with a matching 2700 K, "paper the same across the room." Room: the bedroom, logged first. Direction: one west window (confirmed by the 5 p.m. sun). Reflected: pale rendered wall opposite, "brightens afternoons." Use hours: 4 to 6 p.m. and evenings. That is more than the minimum (the Saturday 5 p.m. entry, which is both a daylight look and one of his use hours, the 8 p.m. entry, and one evening look after the bulb change would have been enough to go on), and it is a good log: honest about the weather it got, specific about what disagreed, and clear about which hours count.

Common mistakes

• Logging the hours you are not there. Three noon entries for an evening room describe someone else's room.
• Trusting the direction. "It faces north so the light is cool" written before looking. The log exists to check that sentence, not to repeat it.
• Reading the wall instead of the paper. The existing wall color is itself a colored surface; the white paper and gray card are the controls that let you see the light rather than the paint.
• Different paper each day. Printer paper, a napkin and the back of a chip are three different whites. Keep the same two sheets.
• Treating LRV as brightness. Choosing a high-LRV color "because the room is dark" without first looking at the lamps and the lumens. The log should say how much light there is before any chip is chosen.
• Treating CRI as a verdict. "It is CRI 95, so the sofa will look right." CRI is a rough guide; two bulbs with the same numbers can still look different. Look at the sofa under the bulb.
• Guessing CRI. If it is not printed and cannot be found by model number, write "unknown." A bulb with no CRI claim could be 80 or lower, and that is worth knowing.

What is still unresolved

• What to do with fixed materials that look different at the hours you logged: Module 5.
• How much of a candidate's behavior at your use hours is the paint and how much is the light: Module 9's sample protocol, which places boards in the room and logs them against these entries.
• Whether to change the lighting itself (a lamp where there is none, a warmer or cooler bulb, a dimmer): Module 10 treats lighting as part of the scheme and ranks it against other changes by effect per dollar.

11. Knowledge check and scenario

Before you move on

• I read the kelvin, CRI (or "unknown") and lumens of every bulb in the first-project room and put them on the bulb sheet.
• I know the window directions as a hypothesis, and what is outside each window that could color or block the light.
• My log has at least one daylight look and one look at each hour the room is actually used, with the same white paper and gray card (and more days, if the decision is hard or expensive).
• If the room had mixed color temperatures, I decided what to do about it, and if I changed bulbs, I did it before sampling anything.
• I can say in one sentence what LRV is and name three things it is not.
• I have not used a photograph to judge a color, only to check value or to record where something stood.