Short answer
The height of the curve sets lightness, where it is high sets hue, and how sharply it rises or falls sets how saturated the colour is. A flat curve is a neutral; a single steep edge is a clean yellow, orange or red; a hump with dark either side is a green or blue. Shape also reveals what a single L*a*b* value hides: two greens with different curves, such as a leaf and a pigmented plastic, will not keep the same relationship under different lights. Measured examples sit 22 to 29 ΔE00 apart depending on the illuminant.
Read a curve in three passes. First its overall height: fresh snow and ground gypsum stay above 80% everywhere and are white; weathered asphalt and galvanised steel stay near 10% and are dark greys. Second, where the high part is. The eye's lightness response peaks near 555 nm, so reflectance there counts most towards L*; reflectance at long wavelengths pushes the colour towards red, at short wavelengths towards blue. Third, how abruptly the curve moves between low and high. Sulphur climbs from 5% at 410 nm to 60% at 480 nm and then flattens, which gives a clean, light yellow. Pine plywood rises across the whole spectrum without an edge, which gives a duller, browner colour.
Three shapes recur in the measured set. An edge is a single step from low to high reflectance: cinnabar is dark and flat until about 570 nm and then rises by eight percentage points in a single 10 nm band; haematite does the same more gradually and from a darker floor, which is why it is a deeper, browner red. A hump is a region of reflection between two absorptions: malachite peaks near 540 nm with absorption in the violet and the red. A trough is the reverse: lazurite reflects in the blue, absorbs broadly around 600 nm and reflects again in the deep red, where the eye barely responds. The table gives each sample's mean reflectance in three regions and where its steepest step falls.
Lawn grass and a light green tarpaulin both peak in the green. The grass curve is low everywhere, with a modest peak near 550 nm, a dip around 670 nm where chlorophyll absorbs, and a steep climb beyond 700 nm. The tarpaulin has a tall peak near 520 nm and stays low in the far red. Under D65 they are simply different colours, one dark and yellowish, one bright and bluish. What the curves add is a prediction: because they differ in different places, each illuminant weights the disagreement differently, and the gap between them runs from about 22 to 29 ΔE00 across seven CIE illuminants for the 10° observer. A pair with the same curve shape at different heights would keep an almost constant gap.
A curve records one geometry. The USGS spectra used here were made for remote sensing, with a laboratory spectrometer on powders and leaves or a field spectrometer at 20° to 40°, and were corrected to absolute reflectance; a d/8° sphere or a 45°/0° instrument would give different heights for glossy or textured specimens. A curve says nothing about gloss, texture or how the colour changes with angle. It assumes the surface only reflects: a fluorescent material emits extra light that depends on the lamp, and its apparent reflectance can exceed 100%. And it describes the specimen measured. The galvanised sheet reads dark because a mirror-like metal sends little light towards a detector that is not at the mirror angle, not because zinc is dark.
| Material | Mean, 400–490 nm | Mean, 500–570 nm | Mean, 580–700 nm | Steepest 10 nm step | L*, a*, b* (D65/10°) |
|---|---|---|---|---|---|
| Green lawn grass | 3.3% | 7.5% | 5.6% | 720–730 nm, +11.5 points | 31.7, -6.9, 17.2 |
| Fresh oak leaf | 9.7% | 14.2% | 11.7% | 720–730 nm, +14.3 points | 43.6, -4.4, 10.8 |
| Golden dry grass | 8.3% | 15.2% | 21.9% | none above 1.5 points | 47.3, 3.9, 19.5 |
| Red-orange geranium flower | 2.8% | 1.5% | 48.9% | 590–600 nm, +11.6 points | 39.2, 58.9, 36.3 |
| Haematite (synthetic iron oxide red) | 2.0% | 2.8% | 18.3% | 580–590 nm, +3.4 points | 29.5, 30.4, 23.8 |
| Goethite (yellow-brown iron oxyhydroxide) | 2.8% | 9.5% | 20.5% | 550–560 nm, +2.4 points | 41.6, 15.5, 36.8 |
| Malachite (copper carbonate green) | 18.0% | 24.7% | 15.7% | 390–400 nm, +2.3 points | 54.6, -12.7, 6.7 |
| Azurite (copper carbonate blue) | 11.1% | 6.0% | 5.0% | 420–430 nm, +2.2 points | 30.0, 4.3, -20.9 |
| Lazurite (lapis lazuli) | 34.4% | 21.1% | 17.7% | 490–500 nm, −3.8 points | 52.3, -2.6, -24.7 |
| Cinnabar (mercury sulphide red) | 8.5% | 9.3% | 34.1% | 610–620 nm, +8.1 points | 43.0, 26.2, 13.8 |
| Sulphur (reagent grade) | 29.9% | 76.1% | 82.2% | 450–460 nm, +11.8 points | 89.6, -9.5, 42.2 |
| Gypsum (selenite), ground white | 89.9% | 91.5% | 92.5% | none above 1.5 points | 96.7, 0.1, 1.1 |
| Cement debris coated with gypsum | 44.6% | 51.0% | 55.2% | none above 1.5 points | 77.1, 0.9, 6.9 |
| Weathered road asphalt | 6.7% | 8.2% | 10.1% | none above 1.5 points | 35.2, 2.4, 6.4 |
| Fresh pine plywood | 18.3% | 29.9% | 52.9% | 410–420 nm, +2.7 points | 64.9, 11.9, 22.5 |
| Galvanised sheet metal | 9.3% | 8.9% | 8.0% | none above 1.5 points | 35.5, -1.4, -1.7 |
| Blue woven plastic tarpaulin | 28.0% | 14.6% | 6.9% | 390–400 nm, +4.4 points | 42.9, -9.9, -32.9 |
| Light green plastic tarpaulin | 11.2% | 29.6% | 10.8% | 500–510 nm, +8.5 points | 54.1, -40.8, 22.4 |
| White vinyl plastic sheet | 74.7% | 79.7% | 76.9% | 400–410 nm, +22.7 points | 91.3, -1.3, -0.5 |
| Fresh snow | 83.1% | 83.2% | 82.4% | none above 1.5 points | 93.1, -0.2, -0.1 |
Each statement is labelled by kind — established fact, a standard’s requirement, observed market data, a convention, or Colourwise’s own interpretation or analysis — with the strength of the evidence behind it.
FactStrong evidence
The USGS Digital Spectral Library splib05a contains measured reflectance spectra of minerals, vegetation and man-made materials, taken on a Beckman 5270 laboratory spectrometer or an ASD field spectrometer and corrected to absolute reflectance; the ASD measurements used incidence and emission angles typically between 20° and 40°.
Source: USGS Digital Spectral Library splib05a (Open-File Report 03-395)
Colourwise analysisModerate evidence
Measured lawn grass and a measured light green tarpaulin differ by between about 22 and 29 ΔE00 across seven CIE illuminants for the 10° observer, the size of the gap depending on the light.
Based on: Colourwise computed CIELAB for USGS splib05a records 11208 and 10440 under D65, D50, A, FL2, FL11, LED-B3 and LED-RGB1 with the 1964 observer, each relative to that illuminant's own white, and compared them with CIEDE2000.
Caveat: CIEDE2000 was fitted to small differences; values above about five are a rough ranking, not a measure of how different the pair looks. One specimen of each material.
Source: USGS Digital Spectral Library splib05a (Open-File Report 03-395); CIE datasets (colour-matching functions, illuminants); CIE 1964 colour-matching functions, 10 degree observer (data table)
Colourwise analysisModerate evidence
In the measured set, the samples that read as near-neutral (fresh snow, gypsum, weathered asphalt, galvanised steel) vary by less than seven percentage points of reflectance across 380–730 nm, while the most saturated samples (the geranium flower, sulphur, cinnabar, haematite and both tarpaulins) each have a band at least five times as reflective as another.
Based on: Read from the 10 nm band means of the twenty USGS splib05a spectra in the site's dataset, which Colourwise checks for the same relationships.
Caveat: Twenty specimens; a description of this set, not a rule about materials.
Source: USGS Digital Spectral Library splib05a (Open-File Report 03-395)
FactStrong evidence
Appearance is not completely specified by a colour calculated from a spectrum, because it is also influenced by properties such as gloss and texture.
Source: ASTM E308-22 Standard Practice for Computing the Colors of Objects by Using the CIE System
Reviewed 6 October 2026. Colourwise summarises its sources in its own words and does not reproduce standards text or proprietary colour data. Spotted an error? Tell us.