Watercolor Color Hues: An Objective Tint-Based Method
Tags:
watercolors

Why we did this

Artists need hue information they can rely on. Names like “Prussian Blue” or “Burnt Sienna” carry history, yet they do not always predict how a paint will behave in tints. Therefore, we built a method that starts with controlled swatches and objective measurements, then translates those numbers into artist‑friendly hue categories and warm/cool bias. The result is consistent across brands and batches, and it is anchored in established testing practice rather than marketing language.

From handmade paint swatches to instrument reading

Swatches you can reproduce

We prepare two swatches for every color on the same standard watercolor paper:

  • Masstone: a full‑strength application that shows undertone, granulation, and transparency.

  • Tint: a uniform wash adjusted to be close to 40% reflectance at the pigment’s maximum absorption (the darkest part of its spectrum). This target keeps the tint in the range that artists actually use while ensuring the instrument can measure it robustly.

Although the lightfastness standard specifies a drawdown on filter paper for formal tests, we use the same reflectance target when making our watercolor‑paper swatches. This keeps our tint data directly comparable to lightfastness panels and eliminates guesswork when you compare paints.

Diagram of CIELAB a-star b-star plane showing chroma C-ab-star as the radius from the neutral center; used to judge watercolor tint saturation and bias.
Chroma (C*ab) as distance from neutral on the CIELAB a*–b* plane. Higher values indicate greater saturation. We read warm/cool on this plane at tint lightness. Image: Nilsjohan, CC BY-SA 4.0, via Wikimedia Commons

Spectrophotometer setup

We read each swatch with a benchtop spectrophotometer using standard daylight illumination and a 10‑degree observer. The instrument records the spectral reflectance curve and then calculates CIE L*a*b* values:

  • L* (lightness) from 0 (black) to 100 (white)

  • a* (green to red axis)

  • b* (blue to yellow axis)

From a* and b*, we compute two helpful descriptors:

  • Hue angle (in degrees): Equation showing hue angle equals arctangent of b-star over a-star, multiplied by 180 over pi.   mapped to 0–360.

  • Chroma (saturation): Equation showing chroma equals the square root of a-star squared plus b-star squared.

We average multiple readings if a color has more than one acceptable tint swatch (for example, 36% and 44% reflectance). This smooths paper texture effects and granulation.

a-star b-star diagram with neutral circle and labeled axes
The a*–b* plane at tint lightness. The center circle shows the near-neutral zone; +a* is redward, −a* greenward; +b* is yellowward, −b* blueward.

How we improve objectivity in the tint range

To remove paper influence and reduce noise where artists actually work (in washes), we apply two steps before assigning warm or cool:

Vector on a-b plane from paper point to swatch point labeled delta a*, delta b*
Paper compensation: we subtract the paper’s a*, b* from the swatch to get Δa* and Δb*. This removes paper warmth and reveals the true undertone of the tint.

  1. Paper compensation. We measure the same sheet of watercolor paper (for this set: L* 95.13, a* −0.09, b* 3.64) and compute Δa* and Δb* for each tint row (swatch minus paper). This subtracts paper warmth and makes brands and batches comparable.

  2. Weighted use of many tints. Instead of relying on one tint, we use all rows between 20–70% reflectance and compute a Gaussian‑weighted average centered at 40%. This emphasizes the ASTM tint target while averaging out granulation and application variation. We refer to the resulting Δ value we use in a family as the Warmth_Score, and we report which axis was used (Score_Axis).

Bell curve with shaded 20–70 percent window and dashed line at 40 percent.
All tint rows between 20–70% reflectance contribute to the average, with those nearest 40% weighted most. This centers the analysis on the ASTM tint target.

Turning numbers into artist‑friendly hue categories

Two questions guide the assignment

  1. What is the core hue family? We determine this from masstone appearance and its measured hue angle. The family labels are the familiar artist primaries and secondaries: Red, Orange, Yellow, Green, Blue, Violet, plus Brown and Black/Gray for low‑chroma neutrals.

  2. Does the tint lean warm or cool? We answer this with the tint’s a*/b* bias, because artists experience warm/cool shifts most clearly in washes.

The decision rules we apply

We kept the rules simple enough to explain, yet precise enough to reproduce.

A. Hue family (first pass)

  • If chroma is very low in masstone, the color enters the neutral families (Black/Gray or Brown) rather than a chromatic family. Otherwise, the hue angle directs it to Red, Orange, Yellow, Green, Blue, or Violet.

B. Warm versus cool (tint‑based)
We use the tint’s sign and magnitude along the a*/b* axes to decide the lean within the chosen family:

Family Warm means… Cool means…
Blue a* ≥ 0 → violet‑blue a* < 0 → green‑blue
Green b* ≥ 0 → yellow‑green b* < 0 → blue‑green
Yellow a* ≥ 0 → orange‑yellow a* < 0 → green‑yellow
Orange a* ≥ 0 → red‑orange a* < 0 → yellow‑orange
Red b* ≥ 0 → red‑orange b* < 0 → red‑violet
Violet a* ≥ 0 → red‑violet a* < 0 → blue‑violet
Brown hue angle < 55° → red‑brown hue angle ≥ 55° → olive/yellow‑brown
Black/Gray b* ≥ +1.5 (or a* ≥ +1.5) → warm (brownish) b* ≤ −1.5 (or a* ≤ −1.5) → cool (bluish/greenish)


These thresholds reflect how painters describe bias in mixtures and washes. They also keep the system legible on category pages and filter swatches.

Why tints, not just masstones?

Painters mix in the tint range. Therefore, a paint’s warm/cool behavior must be interpreted in terms of how water dilutes the film and how paper contributes light. Tints reveal the undertone that masstone can hide. For example, a blue that looks neutral in a dense patch can swing greenish in a wash. Our rules capture that behavior.

Special handling for neutrals and earths

Low‑chroma colors deserve their own logic because small numeric shifts can flip a hue label without reflecting what artists see.

  • Black / Gray (near‑neutrals). In tints, the chroma is tiny, so we classify the undertone using paper‑compensated deltas. If Δb* ≥ +1.5 or Δa* ≥ +1.5, we label Black — Warm (a brownish cast). If Δb* ≤ −1.5 or Δa* ≤ −1.5, we label Black — Cool (a bluish or greenish cast). When both deltas are within ±1.5, we present Black — Neutral. These thresholds are high enough to ignore instrument noise yet low enough to catch meaningful undertones in mixes.

Horizontal ruler labeled delta b-star with ticks at −2.0, −1.5, 0, +1.5, +2.0 and zones for cool, neutral, warm.
Black undertone in tints: Δb* ≤ −1.5 → Black — Cool; |Δb*| < 1.5 → Black — Neutral; Δb* ≥ +1.5 → Black — Warm. If Δa* is larger in magnitude than Δb*, we use Δa* instead.

  • Brown / Earth. Browns live at low‑to‑moderate chroma and sit between orange and yellow at tint. Paper light tends to push b* positive, which can make many browns look cool if you rely only on hue angle. Therefore, we anchor the decision on the red–green cue (Δa*) that artists read in mixtures, and we allow a neutral zone when that cue is essentially zero.

Brown watercolors: an objective way to decide warm vs cool

Browns in watercolor often appear “warm” to the eye because paper light adds yellow, which raises b* and pulls the hue angle toward about 90°. If we categorize warm/cool colors using only the hue angle, many browns appear cool on paper, even when their mixtures feel reddish. To match practical mixing, we center the decision on the red–green axis (a*) and use hue angle only when a* is essentially neutral.

The rule we apply to browns at tint

  1. Detect brown at tint when chroma is modest and the hue angle sits in the orange–yellow sector:

    8 ≤ C*ab ≤ 30 and 25° ≤ h° ≤ 110°.

    (We extend the upper bound from 95° to 110° so burnt umbers near yellow still classify as brown.)

  2. Warm / cool / neutral split using paper‑compensated Δa*, averaged across many tints with Gaussian weight centered at 40%:

    • Brown — Warm if Δa* ≥ +0.5.

    • Brown — Cool if Δa* ≤ −0.5.

    • Brown — Neutral if −0.5 < Δa* < +0.5.

    We publish Warmth_Score = weighted Δa* so you can gauge how strong the bias is.

This approach privileges the red/green cue artists use when neutralizing flesh, greens, and skies, while dampening paper and application effects through averaging.

a-b plane with a brown wedge and vertical lines at minus 0.5 and plus 0.5 a-star marking cool, neutral, warm zones
Browns in the orange–yellow sector: Δa* ≤ −0.5 → Brown — Cool; −0.5 < Δa* < +0.5 → Brown — Neutral; Δa* ≥ +0.5 → Brown — Warm. Using Δa* matches how artists judge red versus olive bias in washes.

Two measured examples

SKU Color name Tint a* Tint b* Tint hue (°) Result
850‑621 Italian Burnt Umber Warm −1.19 +10.96 96.20 Brown — Cool (a* is slightly negative; yellow‑leaning)
850‑618 Italian Burnt Umber −0.16 +10.68 90.86 Brown — Cool (a* is near zero but negative; tiebreaker keeps it cool)


Reconciling names with measurement. Product names follow historical or supplier conventions. Our site labels reflect measured tint behavior because that predicts mixtures. When a legacy name says “Warm” but the tint a* is negative, we list it as Brown — Cool and show the swatch and data so you can judge by eye. When helpful, product pages include a note such as “Measured hue bias: cool.”

What this gives you as an artist

  • Transparent filtering: Category pages show hue families you already speak. The warm/cool split helps you pick complements and split primaries quickly.

  • Predictable mixing: Because assignments come from the tint range, the buckets predict how a color leans in actual mixes. You can plan secondaries and neutrals with fewer test strips.

  • Cross‑brand consistency: The same rules apply to every paint. Therefore, a “cool blue” in our system behaves like a cool blue, regardless of name.

Why perception sometimes disagrees with the label

Language is slippery. “Greenish blue” is Blue—Cool in our rules because a* is negative at tint. However, granulating pigments, optical brighteners in paper, and differences in binders can nudge readings. We average multiple measurements and inspect edge cases, but we always show the measured outcome instead of the marketing name. When your eye says one thing and the label says another, check the warm/cool tag and the sample image; they directly describe tint behavior.

Limitations and transparency

  • Paper matters. Watercolor paper tone and sizing influence the tint’s lightness and chroma. We control this by using the same stock for all swatches and by targeting the same reflectance range.

  • Granulation and texture can introduce local variation. We average multiple readings and standardize the swatch method to reduce noise.

  • Multi‑pigment mixes are classified by their measured tint behavior. If a mixture straddles boundaries, we flag it for manual review and keep the choice most useful to painters.

A quick example: from swatch to hue bucket

  1. Prepare a uniform tint wash close to 40% reflectance at the pigment’s absorption minimum.

  2. Measure L*a*b*; compute hue angle and chroma.

  3. Select a family based on masstone hue (or low‑chroma rules for neutrals).

  4. Apply the tint‑bias rule for that family to assign Warm or Cool.

  5. For blacks and browns, use the undertone and the specialized thresholds above.

Flowchart from handmade swatch to spectrophotometer to paper compensation to weighted tints to hue family to warm/cool/neutral label.
From swatch to label: masstone sets the hue family; paper-compensated, weighted tints determine warm/cool (or neutral) on the appropriate axis.

Pocket formulas

  • Hue angle: Hue angle(wrap to 0–360)

  • Chroma: Chroma

  • Warm/Cool rules: see the table under “The decision rules we apply.”

What matters most for contemporary practice

This approach respects tradition while serving modern workflows. It is reproducible, brand‑agnostic, and sensitive to the part of a paint’s behavior you actually use—the tint. Therefore, you can build split‑primary palettes with confidence, compare earths by undertone, and choose blacks that land the neutral you want without guesswork. Because the system is based on measured tints, the labels on our site reflect what will appear on your paper.

Bibliography

ASTM International. Standard Specification for Artists’ Watercolor Paints (ASTM D5067‑16, reapproved 2021). West Conshohocken, PA: ASTM International, 2016 (reapproved 2021). https://doi.org/10.1520/D5067-16.
Annotation: Defines performance and labeling requirements for artists’ watercolors and the specimen preparation used in lightfastness testing. Its tint drawdown target (approximately the mid‑reflectance range) informed our choice to center measurements around a 40% reflectance tint when assigning hue.

ASTM International. Standard Test Methods for Lightfastness of Colorants Used in Artists’ Materials (ASTM D4303‑10(2022)). West Conshohocken, PA: ASTM International, 2022. https://doi.org/10.1520/D4303-10R22.
Annotation: Specifies exposure regimes and evaluation procedures for rating lightfastness of artists’ media. While our focus here is hue assignment, we mirror its specimen logic so tint data relate meaningfully to permanence evaluations.

International Organization for Standardization; Commission Internationale de l’Éclairage. ISO/CIE 11664‑4:2019 Colorimetry — Part 4: CIE 1976 Lab colour space.* Geneva/Vienna: ISO and CIE, 2019. https://www.iso.org/standard/74166.html.
Annotation: Establishes the definition of the CIE 1976 Lab* colour space that we use to compute a*, b*, chroma, and hue angle from spectrophotometer readings of our tints.

Commission Internationale de l’Éclairage (CIE). CIE 015:2018 Colorimetry, 4th ed. Vienna: CIE, 2018. https://cie.co.at/publications/colorimetry-4th-edition.
Annotation: Authoritative reference for observers, illuminants, and recommended practice in color measurement. We follow these recommendations when selecting instrument geometry and the standard observer/illuminant for our measurements.