Color analysis quiz › Science
Color analysis science: what it can and cannot claim
Seasonal color analysis borrows the vocabulary of color science, but it is a rule-based classification, not a measurement of your skin. Two published studies test how well the labels can be reproduced: 55.6% accuracy on a face-image benchmark of seasonal types, and 65% on a four-season task using influencer-labeled cosmetics. This page sets out what the evidence supports, what it does not, and the exact inputs and arithmetic behind this quiz.
The claims, sorted by evidence
Six claims get bundled together when someone says "your season". They do not stand on the same footing.
| Claim | Status | What the evidence shows |
|---|
| Colors can be located on hue, value and chroma axes | Established color science | The Munsell system orders color by hue, value and chroma, and CIELAB places color in a perceptual space. Both describe color, not people. |
| Every person fits one of 12 seasonal types | Convention | Four-, 12- and 16-season systems coexist and use different type counts. A 2011 paper compares the seasonal skin-color classifications used in the UK, Japanese, Korean and German systems. The label comes from a rule set, not an instrument. |
| Your season predicts which colors flatter your face | Not established | We found no published experiment that tests this directly. The closest experiments point elsewhere: iris color, not skin color, drove clothing-color judgments in Perrett (2025), and clothing lightness changed perceived skin fairness in Zhang et al. (2025). |
| A model can reproduce the labels people assign | Testable, not settled | Ou and Chan (2026) classify face images from the Deep Armocromia benchmark and report 55.2% to 55.6% accuracy, with Spring recall below 0.42. Kawase et al. (2026) classify lipstick colors into influencers' seasonal labels: 65% across four seasons against a 25% chance level, with heavy overlap between all four in the a*–b* plane. |
| Seasonal categories are applied color science | Partly cultural | Kawase et al. describe seasonal labels as looser narrative constructs that combine lightness with culturally elaborated categories. Lee and Park (2025) note that seasonal and warm/cool classifications often produce inconsistent results in practice. |
| In-person draping is the reference standard | Not published | We found no head-to-head study that validates draping against an independent outcome. It is a professional convention and a useful physical test, not a calibrated benchmark. |
The useful claim is the modest one: a season is a compact way to describe three dimensions of your coloring and to bound a palette. The sensible way to hold it is as a heuristic: a short rule that produces a useful answer most of the time, without a validated measurement behind it. It does not settle an argument about a single garment.
What this quiz computes
The quiz asks 12 questions, four per dimension. Every input is your own judgment. Nothing is photographed, uploaded, sampled or measured, and no skin pixel is read. Each option carries a weight from -2 to +2.
| Dimension | The four questions | A positive score means |
|---|
| Undertone (u) | Wrist vein color; white versus cream fabric under the chin; gold versus silver next to the face; how your skin responds to sun. | Warmer (golden or peachy). Negative means cooler (pink or blue-red). |
| Value (v) | Natural or childhood hair depth; skin depth relative to people around you; eye depth; overall impression in a black-and-white daylight photo. | Lighter. Negative means deeper. |
| Chroma (c) | Eye brightness and iris-to-white contrast; hair-skin-eye contrast; saturated versus dusty colors; compliments on bold versus tonal outfits. | Clearer. Negative means softer. |
The 12 questions, verbatim
- In daylight, the veins on your inner wrist look mostly… (undertone)
- Holding pure white vs cream fabric under your chin, which looks better? (undertone)
- Which metal next to your face evens out your skin? (undertone)
- In the sun (without sunscreen) your skin tends to… (undertone)
- Your natural hair color (or your childhood color, if grey or dyed) is closest to… (value)
- Your skin, compared with most people around you, is… (value)
- Your eye color is… (value)
- In a black-and-white daylight photo, your overall impression is… (value)
- Your eyes look… (chroma)
- Contrast between your hair, skin and eyes is… (chroma)
- A saturated top vs a dusty, greyed version of the same color: which one looks like it belongs to you? (chroma)
- People more often compliment you in… (chroma)
Step 1. Sum the four weights in each dimension and divide by 8, the largest possible absolute total, so every score runs from -1 to +1.
Step 2. The dimension with the largest absolute score is the dominant trait.
Step 3. The dominant trait and its sign choose the season.
| Dominant dimension | Direction | Season |
|---|
| Value | Light and warm | Light Spring |
| Value | Light and cool | Light Summer |
| Value | Deep and warm | Deep Autumn |
| Value | Deep and cool | Deep Winter |
| Chroma | Clear and warm | Bright Spring |
| Chroma | Clear and cool | Bright Winter |
| Chroma | Soft and warm | Soft Autumn |
| Chroma | Soft and cool | Soft Summer |
| Undertone | Warm, with value + chroma above 0 | True Spring |
| Undertone | Warm, with value + chroma at or below 0 | True Autumn |
| Undertone | Cool, with chroma − value above 0 | True Winter |
| Undertone | Cool, with chroma − value at or below 0 | True Summer |
Edge rules. An undertone score of exactly 0 counts as warm. When two dimensions tie for the largest absolute score, the order is undertone, then value, then chroma. A person who picks the middle option in all 12 questions gets three scores of 0 and the result True Autumn. That is the tie-break rule, not a finding about anyone's coloring.
Output. The quiz returns three scores, one dominant trait, one of 12 labels, and the two neighboring seasons. It does not return a probability, a match percentage or a skin measurement. The result page prints 8 colors and 4 neutrals as sRGB hex codes, the colors the season advises against near the face, and makeup, hair and clothing notes.
Where the label is most likely to be wrong
- The light you answer in. Indoor bulbs and screens shift warm/cool judgments, and the quiz cannot detect or correct that because it never sees an image.
- Dyed hair, self-tanner, makeup and tinted lenses. The value questions ask for natural or childhood hair for this reason; recent color changes can move the answer.
- Neutral and olive undertones. The warm/cool axis is binary, and a neutral score is rounded to warm by rule. Olive skin often reads as neither.
- A close dominant score. One answer can change the dominant dimension. When two of the three absolute scores are within a question's weight of each other, treat the result as a boundary pair and compare both palettes.
- Color vision and fatigue. The questions ask you to judge hue and contrast; color vision deficiency, tiredness or a poor screen changes the input, not the season.
- Culture, language and the system. Warm and cool boundaries, season names and palettes differ between the 12-season systems used in different countries.
How to get a usable answer anyway
- Answer in daylight. Bare-faced, hair back, no filter. The quiz cannot detect a warm bulb; your answers have to.
- Take it twice on different days. If the label changes, you are on a boundary. Keep both seasons rather than picking the one you like.
- Compare the two neighbor palettes against your face in daylight. The one that leaves your skin even and awake wins.
- Read the scores as directions, not distances. The quiz uses the sign and the dominant dimension; +0.9 cool and +0.1 cool both count as cool.
- For a high-stakes choice (wedding, uniform, brand palette), use physical fabric or book an in-person draping. A screen hex is not a textile.
What would change this page
Three results would move the evidence from thin to useful: a preregistered study showing analysts agree on season labels well above chance, a trial showing season-matched colors are rated more flattering than mismatched colors by blinded observers, or test-retest reliability for quiz-style self-assessment. A search on 2026-10-30 did not find any of the three.
Sources
Searched on 2026-10-30 in OpenAlex and general web search for "seasonal color analysis", "personal color diagnosis", "personal colour analysis" and "clothing color skin tone perception". The list includes every study cited above, plus the standards and trade books the vocabulary comes from.
- Ou, N. Y. P. and Chan, J. H. (2026). Lumina: Seasonal Color Analysis System with Explainable Machine Learning. Proceedings of the 14th International Conference on Advances in Information Technology (IAIT 2026), 1–6. peer-reviewed conference paper. doi.org/10.1145/3816713.3818796. Supports: 55.2% (XGBoost) and 55.6% (ensemble) accuracy on the Deep Armocromia benchmark, Spring recall below 0.42, and the authors' warning that uncontrolled illumination can distort skin LAB values independently of true undertone. Checked 2026-10-30.
- Kawase, A., Adachi, J. and Nagashima, K. (2026). Personal Colour between Perceptual Space and Social Practice. ColourTurn 2(1). journal article. doi.org/10.25538/tct.v2i1.8733. Supports: 1,600 product entries from five Japanese Instagram influencers; 65% accuracy on a four-season task against a 25% chance level; extensive overlap between all four seasons in the a*–b* plane; seasonal categories described as looser narrative constructs. Checked 2026-10-30.
- Stacchio, L., Paolanti, M., Spigarelli, F. and Frontoni, E. (2025). Deep Armocromia: A Novel Dataset for Face Seasonal Color Analysis and Classification. Computer Vision – ECCV 2024 Workshops, Lecture Notes in Computer Science, 352–367. conference paper and dataset. doi.org/10.1007/978-3-031-91569-7_22. Supports: the labeled benchmark used by Ou and Chan; its labels are the target the model is scored against, not an independent measurement. Checked 2026-10-30.
- Lee, S. and Park, Y. (2025). The Study on Image-Based Perception of Color Harmony in Personal Color Diagnosis. Journal of Korea Society of Color Studies 39(4), 55–63. journal article. doi.org/10.17289/jkscs.39.4.2025.11.55. Supports: the authors' statement that seasonal and warm/cool classifications often produce inconsistent results in practice, and their finding that high-contrast images were judged more harmonious with vivid and dark tones while low-contrast images matched pale and dull tones. Checked 2026-10-30.
- Perrett, D. I. (2025). Eye color is more important than skin color for clothing color aesthetics. Psychology of Aesthetics, Creativity, and the Arts 19(5), 1099–1110. peer-reviewed journal article. doi.org/10.1037/aca0000626. Supports: iris color dominated clothing-color choices across two experiments, skin color had little influence, and participants told to match clothing to skin color still used eye color. Checked 2026-10-30.
- Zhang, Q., Chen, Z., Tian, B., Guan, Y. and Liu, Q. (2025). The impact of clothing colour on skin tone perception and consumer preference. Coloration Technology 142(2), 259–271. peer-reviewed journal article. doi.org/10.1111/cote.12828. Supports: 22 plain-colored sweatshirts in a real observation setting; lower clothing lightness made skin appear fairer; young Chinese participants preferred low colorfulness and a blue tint. Checked 2026-10-30.
- 차호연 and Kim, J. (2011). A Comparative Analysis of the Skin Color According to Seasonal Types in Personal Color Systems: Focused on UK, Japanese, Korean, German Personal Color Systems. Journal of Investigative Cosmetology 7(4), 339–348. journal article. doi.org/10.15810/jic.2011.7.4.002. Supports: that several national personal-color systems define seasonal skin color separately; the label depends on which system a list uses. Checked 2026-10-30.
- Foster, D. H. (2011). Color constancy. Vision Research 51(7), 674–700. review. doi.org/10.1016/j.visres.2010.09.006. Supports: the perceived color of a surface depends on the illumination and its surroundings, which is why a photo or a screen can shift a warm/cool judgment. Checked 2026-10-30.
- Munsell Color Science Laboratory, Rochester Institute of Technology. Munsell Color System. reference. www.rit.edu/science/munsell-color-science-laboratory. Supports: the hue, value and chroma ordering that the 12-season vocabulary borrows. Checked 2026-10-30.
- International Commission on Illumination (CIE). Colorimetry, 4th edition. standard. cie.co.at/publications/colorimetry-4th-edition. Supports: the CIELAB perceptual color space used by both classification studies cited above. Checked 2026-10-30.
- Itten, J. (1961). The Art of Color. trade book, not a study. openlibrary.org/search?q=the+art+of+color+itten. Supports: the seasonal color language that image consultants later adapted; it is a design theory, not an empirical test. Checked 2026-10-30.
- Jackson, C. (1980). Color Me Beautiful. trade book, not a study. openlibrary.org/search?q=color+me+beautiful+jackson. Supports: the four-season system that made seasonal typing mainstream. Checked 2026-10-30.
- Spillane, M. (1985). Color Me Beautiful's Looking Your Best. trade book, not a study. openlibrary.org/search?q=looking+your+best+spillane. Supports: the trade tradition that expanded the four seasons into a finer set of types. Checked 2026-10-30.
Each link was checked on 2026-10-30. Citation details are given so the claims can be checked against the abstract, not taken from this page.
Next: how the 12-season grid works and what the free and paid tiers include.