LOOKSMITH

Notes

Am I cool or warm toned? How to tell, in five tests

Last updated:

Five tests in daylight, run in order, settle whether you are cool, warm or neutral, and three of the five cannot report neutral even when it is the right answer.

Run five tests in daylight, count the results and take the majority. Four or five in one direction means you are cool toned or warm toned. A three to two split, or two unreadable tests, means neutral. Neutral is the hardest answer to get out of a quiz, because three of the five standard tests ask a yes or no question and cannot return it.

What you are testing, and what you are not

Undertone is the colour underneath your skin. For practical purposes it holds still. Skin tone is the surface, and it moves all the time with sun, sleep, temperature and whatever you had to drink last night. A tan changes your tone and leaves your undertone where it was. If the two ideas are still tangled, separate them before you start.

The practical reason to care is narrow. Undertone predicts which foundation shade will disappear at your jaw instead of sitting on your face like a mask, and it cuts a wall of lipsticks down to a set you can work through. It predicts nothing about coverage, finish, wear time or whether a formula will irritate you.

The five tests, and what each one can return

TestWhat you look atReads coolReads warmCan it return neutral?
1. Wrist veinsThe veins inside your forearm, in daylightBlue or purpleGreen or oliveNo. Only "I cannot tell"
2. White against creamA pure white shirt or sheet of paper, then an ivory one, held under your jawPure white looks clean, cream looks dingyCream looks warm, pure white looks harshNo. Only "both look fine"
3. Silver against goldThe two metals flat against bare skin, same light, same minuteSilver looks brightGold looks brightNo. Only "both look fine"
4. Burn against tanHow your skin reacts to strong sun, over years rather than one afternoonBurns pink, tans slowlyTans with a golden cast, rarely burnsYes. Burns first, then tans
5. Blue red against orange redTwo red lipsticks, one berry-leaning, one brick-leaningThe blue red makes your teeth look whiterThe brick red makes your skin look clearerYes. Both look right

Test four is a borrowed instrument. Burn-and-tan response is the Fitzpatrick scale (Fitzpatrick, Archives of Dermatology, 1988, 124(6), 869 to 871), which classifies how skin reacts to ultraviolet light rather than what colour sits underneath it, and it was never validated on deeper skin tones. It tracks undertone loosely. That is why test four is a tiebreaker and not a verdict.

That last column is the whole problem with how the quizzes answer this. Three of the five tests are a coin with two faces. When your answer is in the middle they return a shrug, and a shrug is what a quiz throws away.

How to run them so the answer means something

Does the lighting really change the result?

More than anything else on this page. Daylight, and only daylight. Stand at a window between roughly ten in the morning and three in the afternoon, or step outside into open shade. A shop ceiling is lit to sell the product rather than to show you your own colouring, and a bathroom bulb has a colour temperature nobody chose on purpose.

Then remove the confounders. Bare face, no foundation and no bronzer. No fake tan and no fresh tan. Bare wrist for test one, with no watch strap pressing colour into the skin. Nail polish sitting next to a metal in test three will drag the result, so take it off or use your collarbone instead of your hand.

Run all five, in order, in one sitting. Write the five results down before you interpret any of them. Reading each one as you go is how you talk yourself into the answer you wanted.

Scoring, with no wriggle room

What each answer predicts

Your resultFoundationLipstickHair colour
CoolShades coded C, or described as pink or rosy. Between two candidates, take the pinker oneBlue reds, berry, plum, raspberry, cool pinks, true redAsh blonde, cool or beige brown, espresso, blue black, burgundy
WarmShades coded W, G or Y, or described as golden or yellow. Between two, take the more golden oneCoral, brick, terracotta, warm nude, orange red, warm brownHoney and golden blonde, caramel, copper, auburn, warm chocolate
NeutralShades coded N. Failing that, the two nearest shades will both look acceptable, so pick the one that vanishes at the jawline after ten minutesChoose by depth and how strong the colour is rather than by temperature. True red and a mid rose are the safest starting pairNearly the full range is open. Decide by how much contrast you want against your skin and eyes, which what hair colour would look good on me works through

Shade codes are not a standard. C, N and W mean cool, neutral and warm at most brands, and MAC runs the letters the other way: its own shade finder defines the C in NC as golden and the W in NW as rosy, so an NC range is sold to the skin most other brands would code W. Read the brand's own shade description rather than the letter.

All three rows carry the same caveat on top of that. Foundation oxidises, drifting darker or more orange as it meets air and skin oil, so the shade you approve under a counter light is not always the shade you are wearing at dinner. Undertone gets you to the right two shades. Only wearing one for a morning picks between them.

Why the quizzes almost never say neutral

Nobody has published a census of undertone across a general population, so anyone quoting you a percentage made it up. Undertone is a continuous quantity and not two bins. Cool and warm are the ends of a line, everything between them is neutral, and no test above tells you where on that line most people sit.

The reporting bias is simple to see once you look for it. A quiz that tells you your veins are blue and therefore you are cool has given you an identity and a shopping list. A quiz that tells you that you are neutral has given you a shrug. One of those gets shared. Three of the five tests above cannot report neutral even when neutral is correct, so the format pushes every borderline reader to an end of the line.

Neutral is the least restrictive answer available. The temperature rule stops being a constraint, and you choose on depth and finish instead.

The awkward cases

Why do the vein test and the jewellery test give me different answers?

Because they measure different things. Vein colour is not blood colour. Venous blood is dark red, and veins look blue or green because of the way skin scatters light back at different depths (Kienle and others, Applied Optics, 1996, 35(7), 1151 to 1160). What test one measures is how your skin scatters light, which tracks skin thickness more than it tracks undertone. Metals measure something else again, which is how your skin throws a warm or a cool light source back at itself. Someone with fair, thin skin and a golden cast underneath can get blue veins and read warm on the metals. When tests one and three split, stop trying to break the tie. The split is your answer, and the answer is neutral.

What does it mean if I look good in both silver and gold?

Neutral, in most cases. It can also mean the two pieces you compared were not comparable, since brushed gold and polished silver reflect different amounts of light regardless of your skin. Repeat it once with two similar finishes. If both still look fine, that is the answer. Every metal on the counter is open to you.

Can my undertone change?

Not on any timescale that affects your shopping. Your surface tone changes across a year, sometimes by enough that your foundation no longer matches between February and August, and that is why people think their undertone has flipped. What happened is that you got darker or lighter within the same temperature. Re-test your shade every season. Do not re-test your undertone.

Can an app tell me if I am cool or warm toned?

An app can guess. If one hands you a confident label with a percentage beside it, treat that number as marketing, because an accuracy claim needs a published method behind it. Any tool reading a phone photograph is also reading your camera's automatic white balance, the bulb over your head and the colour of the wall behind you, and without a known colour reference in the frame it cannot reliably separate those from your skin.

Looksmith takes the other route and does not print a verdict. Looksmith does not classify you and it does not give you a season. You save a Likeness once, which is a set of selfies that becomes your reusable face, then you pick a look and Looksmith renders that complete look onto your own photograph in about a minute, with the products it used listed beside it. The difference is what you are looking at: a coral lip and a berry lip on your own face rather than a word describing them. Looksmith's Style Me puts you in several looks at once, which is the case to use if your five tests came back neutral and the temperature rule gave you nothing to narrow with. Looksmith is coming soon to iOS, the first six looks are free, and there is no subscription.

Two honest limits. Looksmith renders a look, it does not measure a pigment, so it is a read rather than a rulebook. And Looksmith will only shift your skin tone by a single step, through product finish, because it will not reshape the face you brought.

What to do with the answer

If you came out cool or warm, use it as a filter and nothing more. It removes a large part of a shade wall. It tells you which of two near-identical foundations to take home. Worth the ten minutes.

If you came out neutral, stop looking for a tiebreaker. You skip the temperature question and spend the attention on depth and finish. Whether you like the look on your own face was always the part that decided it.

Looksmith renders a full makeup look onto your own face and lists the products it used. iOS, coming soon.

What Looksmith does