Studio Matrx Monthly · Volume 1 · Issue 4 · September 2026
Amogh N P
 In loving memory of Amogh N P — Architect · Designer · Visionary 
Artificial Light & Colour RenderingLesson 3.3
Colour Theory & Application/Module 3 · Colour & Light

Lesson 3.3 · Colour & Light

Artificial Light & Colour Rendering

After dark, you choose the sun - so warmth (CCT) and honesty (CRI and R9) become colour decisions, and the wrong lamp can quietly ruin a beautiful palette

13 min Interactive lessonFree · open lessonByAmogh N P· Architect & interior designer
The hook

You spent weeks on the palette, then handed the whole thing to a bulb someone bought for eighty rupees.

Once the sun goes down, you become the lighting designer of the sky - and most of the colour disasters in finished interiors happen right there, in the choice of lamp. A cheap, wrong bulb can take a carefully built palette and quietly strip the life out of it: the reds go brown, the skin goes grey, the timber goes flat, and everyone blames the paint. Artificial light is the one part of the light-and-colour equation you fully control, which means when it goes wrong, it is your decision that went wrong.

Two numbers do most of the work, and every designer should be fluent in both. The first is colour temperature - the warmth or coolness of the light, in kelvin. The second, and more important for colour, is the colour rendering index and its companion R9 - how honestly a lamp reveals colour, and especially red. Get these right and your daytime palette survives the evening; get them wrong and it does not. This lesson makes both numbers concrete, exposes the traps hiding in ordinary LEDs, and gives you a way to specify lamps so the colours you designed are the colours people actually see after dark.

Kelvin = warmth. CRI = honesty (90+). R9 = the red it hides (50+). Read the box before you buy the bulb.

Colour temperature: how warm or cool the light is

The first thing to decide about any lamp is its colour temperature, measured in kelvin (K) and printed on every decent bulb's box. The scale is a little counter-intuitive: low numbers are warm (yellow-orange), high numbers are cool (blue-white). A candle is around 1800K; a warm incandescent-style lamp about 2700K; a 'neutral' or 'natural' white around 4000K; and a cool 'daylight' lamp 5000-6500K. The name comes from heating a theoretical object until it glows - it runs red, then orange, then yellow-white, then blue-white as it gets hotter - which is why a hotter number looks cooler to the eye.

Colour temperature sets the mood of a room and the cast it throws over every surface. Warm light (around 2700K) is relaxing, flattering and domestic; it makes warm colours - terracotta, timber, warm whites, skin - glow, and is the natural choice for living rooms, bedrooms and restaurants. Cool light (4000K and up) is crisp, alert and clinical; it suits kitchens, bathrooms, workshops, offices and any task needing precision, and it keeps cool colours - blues, greys, greens - clean. Get it wrong and a room fights itself: a cosy bedroom under 6000K 'daylight' LEDs feels like an operating theatre, and a kitchen under dim 2700K can feel gloomy and make food and worktops look grubby.

> Colour temperature is the light's accent. Pick it for the room's job and for the colours you want to flatter - warm to make warm colours sing, cool to keep cool colours clean.

A practical warning: mixing colour temperatures carelessly across one space reads as a fault, not a feature. A room where a 2700K lamp sits beside a 4000K downlight and a 6500K bare fitting looks broken - one pool warm, the next cold - and it undermines any colour scheme. Choose a temperature per space (or per clearly separated zone) and hold to it.

COLOUR TEMPERATURE (KELVIN)WARMCOOL2700Kcosybedroom3000Kwarmliving4000Kneutralkitchen5000K+cooltask/clinicalLower kelvin looks warmer. Pick per room - and hold one temperature per space.
Zoom
Colour temperature runs from warm to cool across the kelvin scale - and, counter-intuitively, lower numbers look warmer. Around 2700K is cosy and flattering to warm colours and skin; 4000K is a neutral working white; 5000K and above is a cool, clinical 'daylight'. Choose the temperature for the room's job, and hold it consistent within a space.

Low K = warm (2700 cosy). High K = cool (5000+ clinical). Warm flatters warm colours + skin; cool keeps blues/greys clean. Don't mix carelessly.

CRI and R9: how honestly a lamp renders colour

Colour temperature tells you the light's cast, but not whether colours look true under it - and those are different questions. Two lamps can both be a pleasant 3000K and yet render colour completely differently, because one has a full, continuous spectrum and the other has gaps. The number that captures this is the colour rendering index, or CRI: a score out of 100 for how faithfully a light shows a set of reference colours compared with a full-spectrum reference (daylight or incandescent). 100 is perfect; a good modern lamp for interiors should be 90 or above; below about 80, colours visibly suffer.

CRI is genuinely useful but it has a famous blind spot, and it is exactly the one that matters most in interiors. The general CRI score is an average over eight fairly muted test colours - and it does not include the strong, saturated red sample known as R9. Deep red is precisely the wavelength region where cheap LEDs are weakest (recall the spiky SPD from Lesson 3.1), so a lamp can post a respectable CRI of 80-plus while quietly rendering reds, and therefore skin tones, terracotta and warm woods, badly. This is why a lamp can 'test fine' on the box and still make a room look sickly. For colour work, always ask for R9 as well as CRI - a high CRI with a poor R9 is a lamp that lies about red.

The stakes are highest wherever red lives, which is almost everywhere that matters emotionally. Skin under a low-R9 lamp looks grey, tired and unwell - fatal in a bedroom, a restaurant, a salon or anywhere people gather. Food looks unappetising. Timber, brick, terracotta and warm textiles go flat and lifeless. Meanwhile the blues and greens might look fine, so the fault is easy to misdiagnose as a paint problem when it is really a lamp problem. For any space where colour and people matter, specify CRI 90+ and a solid R9 (say 50+), and treat a bargain bulb with no published R9 as a risk.

CRI 90+ / R9 GOOD vs R9 POORhigh CRI + high R9 (true)redwoodterracottagreenblueskingood CRI but LOW R9 (reds drained)dullflatgreyedokokgrey/tiredBlues and greens survive; reds, wood, terracotta and skin collapse - the R9 blind spot.
Zoom
CRI scores how honestly a lamp reveals colour, but its headline average leaves out R9 - the saturated red. So a lamp can post a decent CRI and still render red, skin, timber and terracotta poorly. The top row shows colours under a full-spectrum, high-CRI, high-R9 light; the bottom row shows the same colours under a low-R9 lamp, with the reds and warm tones drained.

CRI = how honest the light is (aim 90+). R9 = the RED it hides. High CRI + low R9 = a lamp that lies about red, skin, wood, terracotta.

The LED trap - and how to read a lamp

LEDs are the default now, and rightly so - efficient, long-lived and dimmable. But they created a new colour problem precisely because of how most of them make white light. A common LED starts with an intense blue emitter and coats it with a yellow phosphor; blue plus yellow reads as white to the eye and to a light meter, but the underlying spectrum is spiky, with a real dip in the cyan-greens and, crucially, weak deep red. That is the technical reason so many otherwise-modern rooms feel subtly wrong: the light is bright, efficient and 'white', and still starving the colours that carry warmth and life.

The fix is not to avoid LEDs but to read them properly and buy the good ones - the difference between a poor LED and an excellent one is now enormous, and often not much more expensive. On the box or datasheet, check four things in order: colour temperature in kelvin (does it suit the room?); CRI (aim 90+); R9 if published (aim 50+, and be suspicious if it is hidden); and consistency, expressed as a binning tolerance in 'MacAdam steps' - a low number means one batch of lamps will actually match each other, which matters when several light one wall.

text
Reading a lamp for colour - the quick checklist
  CCT (kelvin)   -> warmth: 2700 cosy / 3000 warm-neutral / 4000 neutral / 5000+ cool
  CRI            -> honesty: 90+ good, 95+ excellent, under 80 avoid for colour work
  R9             -> the red it hides: 50+ good; if not published, be wary
  Binning steps  -> consistency: fewer MacAdam steps = lamps that match each other
  Dimming        -> good LEDs 'warm-dim' toward 2200-2400K like incandescent; cheap ones just get dim and grey

One more subtlety worth knowing: many premium warm LEDs are designed to warm-dim, dropping toward 2200-2400K as you dim them, mimicking the cosy glow of a dimmed incandescent. Cheap LEDs just get dimmer and, if their R9 is poor, greyer and sadder. If evening ambience matters - and in living and dining spaces it usually does - warm-dim behaviour is worth specifying, because it keeps warm colours and skin flattering right down to candle-level brightness.

CRI 90+ / R9 GOOD vs R9 POORhigh CRI + high R9 (true)redwoodterracottagreenblueskingood CRI but LOW R9 (reds drained)dullflatgreyedokokgrey/tiredBlues and greens survive; reds, wood, terracotta and skin collapse - the R9 blind spot.
Zoom
CRI scores how honestly a lamp reveals colour, but its headline average leaves out R9 - the saturated red. So a lamp can post a decent CRI and still render red, skin, timber and terracotta poorly. The top row shows colours under a full-spectrum, high-CRI, high-R9 light; the bottom row shows the same colours under a low-R9 lamp, with the reds and warm tones drained.

Specifying light so colours read true

Pulling it together, choosing artificial light for colour is a short, disciplined sequence, and it belongs in the colour decision, not in a separate electrical package handed off and forgotten. First, decide colour temperature by room and mood: warm (2700-3000K) for living, dining, bedrooms and hospitality where you want warm colours and skin to glow; neutral to cool (3500-4000K, occasionally higher) for kitchens, bathrooms, studies and workshops where precision and clean cool colours matter. Then hold that temperature consistent within each space or clearly zoned area, so the room does not read as patched together.

Second, demand colour honesty everywhere people and finishes matter: CRI 90+ as a floor, a published R9 (50+) wherever red, skin, wood or terracotta appear, and matched binning where several lamps wash one surface. Treat an unbranded bulb with no colour data as unknown until proven otherwise - it is the single most common way a good scheme dies after dark. Third, think about layers, not one ceiling light: a mix of general, task and accent light lets you use warmth and brightness deliberately, and a good dimmer (ideally warm-dimming) turns one fitting into several moods.

Finally - and this connects straight to the next lesson - approve colour under the actual lamps that will be installed, not the site's temporary work lights. A colour signed off under a bare, cool, low-CRI construction bulb is a colour you have not really seen. The safest practice is to view final samples under the specified lamps and in daylight, because a colour has to survive both. Two colours that match beautifully under one light can diverge under another - the phenomenon of metamerism - which is exactly why on-site approval under real, final light is non-negotiable, and exactly where Lesson 3.4 begins.

Warm for living/hospitality, cool for task rooms. CRI 90+, R9 50+, matched binning. Approve under the real installed lamps - not site work-lights.

Terms & tools you'll meet in this lesson

CCT (correlated colour temperature)

The warmth or coolness of a white light, in kelvin (K)

Low = warm (2700K cosy), high = cool (5000K+ clinical). Sets mood and cast; choose per room and hold it consistent.

CRI (colour rendering index)

A 0-100 score for how faithfully a lamp reveals colour versus a full-spectrum reference

Aim 90+ for interiors. Useful but averaged over muted samples - it can hide a real weakness in red.

R9

The saturated-red test colour left out of the headline CRI average

The number that reveals skin, wood and terracotta honestly. Aim 50+; be wary of any lamp that does not publish it.

Binning / MacAdam steps

A tolerance for how closely a batch of lamps match each other in colour

Fewer steps = lamps that actually match. Matters when several fittings wash one wall or ceiling.

Hands-on workshop

Workshop - put your lamps on trial

This is a quick, revealing test you can run at home or in a shop with a lamp aisle. It makes the abstract numbers - CCT, CRI, R9 - visible on the things you care about most: red, wood and your own skin. Once you see the difference, lamp choice stops being an afterthought.

Two or three different bulbs or lit spaces, a strong red object, a piece of wood or terracotta, your own hand. No software.

Given & goal
Goal: see how different lamps render red, wood and skin
Inputs: 2-3 different bulbs or lit rooms + a red object + a piece of wood + your own hand
Time: ~15 minutes
  1. 1Gather a warm cheap LED, a cool cheap LED and, if you can, a good CRI-90+ warm lamp (or use daylight as the honest reference). Note each one's box: kelvin, CRI, and whether R9 is even mentioned.
  2. 2Hold a strong red object (an apple, a tomato, a red book) under each light in turn. Rank them: which red looks juicy and alive, which goes dull, brownish or muddy?
  3. 3Now hold your own hand under each. Watch for the tell-tale grey, tired look under a low-R9 lamp versus a healthy warmth under a good one - skin is the most honest test there is.
  4. 4Add a piece of timber or a terracotta pot and repeat. Note which lamp makes the grain and warmth sing and which flattens it.
  5. 5Write down which lamp you would specify for a bedroom and which for a bathroom, and why - in terms of CCT and R9, not brand.

You’ll walk away with
A short ranked note comparing 2-3 lamps on how they render red, wood and skin, with a specification decision (which lamp for which room) justified by CCT, CRI and R9.

The worked example

Three altitudes on the same idea

Read the band that fits you — or all three.

For the architectColour on facades, in space and in context

Write light quality into the specification, not just the layout and the lux levels. A facade uplight, a lobby, a corridor - each needs a stated CCT and a minimum CRI, and R9 wherever brick, timber, stone or people are on show, or the contractor will value-engineer in the cheapest bulb that hits the lux target and quietly wreck the material palette. Coordinate colour temperature across a scheme so spaces relate, and insist on matched binning where lamps wash a single surface.

For the interior designerColour room by room, in materials & light

Choose the lamp with the same care as the paint, because after dark the lamp is half the colour. Set colour temperature by room and mood - warm for living, dining and bedrooms so skin and warm colours glow; cooler for task spaces. Insist on CRI 90+ and a real R9 wherever people, food, wood or terracotta appear. Design in layers and a warm-dimming dimmer, hold one temperature per space, and always approve your final colours under the actual specified lamps.

For the studentA trained colour eye and vocabulary

Learn to read a lamp box like a label: kelvin for warmth, CRI for honesty, R9 for the red it hides. Do the experiment - hold a red apple and your own hand under a cheap cool LED and under a warm CRI-90 lamp, and watch the difference. Once you have seen skin go grey under a low-R9 bulb you will never un-see it, and you will start noticing bad lighting in cafes, shops and homes everywhere - the fastest way to build a professional eye for light.

Misconception check

A high CRI number means colours will look great - it is the one figure that matters for good lighting.

CRI is important but incomplete, and trusting it alone is a classic trap. The headline CRI is an average over eight fairly muted test colours and deliberately excludes the strong saturated red, R9 - the very wavelength region where cheap LEDs are weakest. So a lamp can advertise a respectable CRI of 82 while rendering reds, skin, timber and terracotta poorly, making a room look subtly sickly. For colour work you need CRI 90+ and a published R9 (aim 50+). A high CRI with a hidden or low R9 is a lamp that tells the truth about most colours and lies about red - the colour that matters most for people and warm materials.
Try it

Do it yourself

No tools needed - reason it through.

  1. 1On the kelvin scale, does a lower number mean warmer or cooler light - and roughly what CCT would you pick for a cosy bedroom?
  2. 2In one line, what does CRI measure, and what is a good target for an interior?
  3. 3Why can a lamp with a decent CRI still make skin and reds look bad? Name the number that exposes this.
  4. 4Explain why a cheap 'white' LED can be bright and efficient yet still starve certain colours.
  5. 5Why should final colour samples be approved under the actual installed lamps rather than the site's work lights?
Take this with you

The one line to carry out

After dark you choose the sun, so make it a colour decision: set colour temperature (CCT) for the room's mood, and demand colour honesty (CRI 90+, R9 50+) so reds, skin and warm materials read true.
Take it further
References & further reading

Peer-reviewed journals & authoritative standards

  1. 01Color rendering indexWikipedia, 2026.
  2. 02Color temperatureWikipedia, 2026.
  3. 03LightingWikipedia, 2026.
Related lessons
Recap
Artificial light is the half of colour you fully control, and two numbers govern it. Colour temperature (CCT, in kelvin) sets warmth - low is warm and cosy, high is cool and clinical - and should be chosen per room and held consistent. The colour rendering index (CRI) scores how honestly a lamp reveals colour; aim for 90+, but know that its headline average hides R9, the saturated red where cheap LEDs are weakest, so always check R9 wherever skin, wood or terracotta matter. Read every lamp for CCT, CRI, R9 and binning, prefer warm-dimming LEDs for ambience, and approve final colours under the actual specified lamps.
Carry forward →

You now control both the warmth and the honesty of your light - but a new trap appears when colours are chosen under one light and viewed under another. Lesson 3.4 closes the module with metamerism and colour constancy: why two samples that match perfectly in the shop can diverge on site, and why colour must always be approved under real, final light.

A

The author

Amogh N P

Architect, interior designer, and creative polymath. Studio Matrx began in his notebooks — his vision of design made honest, useful, and open to everyone. Its Academy is written and taught in his memory, and free, forever.

More about Amogh →