Short answer
Young red wine gets its purple-red from grape-skin anthocyanins. Over time those react with tannins and with fermentation by-products to form new, more stable pigments that are redder-orange and less sensitive to pH, while free anthocyanins decline — so the colour moves from purple through ruby to brick and tawny. White wines deepen from green-gold to gold and amber mainly through slow oxidation of their phenolic compounds. Oxygen exposure speeds both. Laboratories follow the change with two published OIV methods, one built on three absorbance readings and one on CIELAB; neither comes with an official ladder of colour names.
Nearly all red grape varieties have colourless juice; the colour is in the skins. During fermentation and maceration, anthocyanins are extracted from the skins into the wine along with tannins. In a young wine many of these anthocyanins are free and behave like the pigments in red cabbage: wine's acidity keeps a share of them in the red flavylium form, and much of the rest is colourless at wine pH. Sulphur dioxide added as a preservative also bleaches free anthocyanins reversibly. The vivid purple of a new red is therefore a fragile colour.
As wine ages, anthocyanins react with yeast metabolites such as pyruvic acid and acetaldehyde, and with tannins, to form new pigments. Pyranoanthocyanins, one family of these, are coloured across the whole pH range, resist bleaching by sulphur dioxide and are more stable in storage than the anthocyanins they come from; their colours sit towards orange-red. Pigmented tannin polymers add brick and brown tones. Meanwhile free anthocyanins are lost. The net effect is a colour that becomes less purple and less intense but more stable — the ruby-to-garnet-to-tawny sequence a taster reads as age.
White wines have little anthocyanin, and their colour comes from much smaller amounts of phenolic compounds. Exposed to oxygen, these oxidise — enzymatically in the juice if it is handled carelessly, and chemically in the wine over years — and form yellow-brown products. A slow, controlled version gives the gold of an aged white; a fast, uncontrolled one is a fault called browning, often with stale, sherry-like aromas. Winemakers prevent it by limiting oxygen, using antioxidants such as sulphur dioxide, and can remove some brown compounds with fining agents.
Rosé and orange wines sit between the two: short skin contact extracts some pigment, and orange wines are white grapes fermented on their skins, extracting phenolics that give amber colour.
Tasters tilt the glass over a white surface and look at depth (how much light gets through the centre) and hue at the rim, where the wine is thinnest. A purple rim suggests youth; an orange or brick rim, age or oxidation. The tilt matters because a glass is a wedge: the same liquid is seen through a few millimetres at the edge and several centimetres at the centre, so one pour shows a whole series of depths at once. Lighting matters too: warm restaurant light makes every red look browner and older than it would in daylight.
The International Organisation of Vine and Wine publishes its analysis methods openly, and two of them deal with colour. The older convention, method OIV-MA-AS2-07B, applies to red and rosé wines. It reads absorbance at 420, 520, 620 nm — conventionally taken as the yellow-brown, red and blue parts of the colour — and refers each reading to a 1 cm path. Colour intensity is the three added together; shade is the first divided by the second. A dark red swallows too much light for a 1 cm cell, so the method lists cells down to 0.1 cm and asks for one that brings the reading between 0.3 and 0.7 before scaling. The second method, OIV-MA-AS2-11, records the whole transmission spectrum from 380 to 780 nm in 5 nm steps and converts it to CIELAB under illuminant D65 for the 10° observer, referred to 10 mm of wine. It reports lightness, the red–green and yellow–blue coordinates, and the chroma and hue angle derived from them.
A worked case: a red read in a 0.1 cm cell at 0.42, 0.55 and 0.13 has 1 cm absorbances of 4.2, 5.5 and 1.3, so its colour intensity is 11.0 and its shade 0.42 ÷ 0.55 = 0.764. The figures are invented to show the arithmetic, not taken from a real wine.
The red reading depends less on how much anthocyanin a wine holds than on what state it is in. At a typical wine pH of 3.4 to 3.6 only around a fifth to a quarter of the anthocyanin molecules are in the red flavylium form, and at pH 4 about a tenth are, so a small rise in pH costs colour without any pigment being lost. Sulphur dioxide does something similar by a different route: bisulfite adds to the flavylium ion and makes a colourless product, reversibly. Copigmentation works the other way. Anthocyanins stack with colourless phenolics such as flavanols and flavonols, which both intensifies the colour and shifts its absorption peak by some 5 to 20 nm towards longer wavelengths — the violet cast of a very young red. Reviews estimate that this stacking supplies 30 to 50% of a young red wine's colour. It is temporary, because the free anthocyanins it depends on are steadily being built into larger pigments.
By the end of alcoholic fermentation roughly a quarter of the anthocyanins are estimated to have already joined to tannin-type molecules, and after a year the share passes 40%. In these polymeric pigments the coloured part of the molecule is shielded from water and from sulphur dioxide, and the pyranoanthocyanins stay coloured right across the pH range, so the colour of a mature red no longer swings with the conditions that govern a new one. In the three-wavelength convention the same history shows as arithmetic: the red reading falls as free anthocyanins disappear, the yellow-brown reading holds or climbs as orange and brown pigments form, and shade, being one divided by the other, rises. The three-wavelength method is not scoped to whites, which have almost nothing absorbing in the red; their browning is followed with the CIELAB method instead, where a browner wine would be expected to read darker and yellower.
Neither method yields a colour you could print. Both look through a thin, fixed layer of clarified wine, while a glass presents every depth from a film to several centimetres, lit from whatever the room provides. That is why this site gives no colour chip for wine, and why no official table turns a shade value into ruby or garnet: the tasting vocabularies used in wine education are teaching conventions, not measurement standards. The numbers are still useful for what they were built for, which is comparing wines measured the same way and tracking one wine over time. A comparison of the two approaches on 112 wines found that they agreed only for dark reds and that CIELAB followed what tasters saw more closely, with neither capturing everything. Agreement between laboratories is also uneven: in the OIV's own validation, the limit within which two laboratories should agree on lightness ranged from under half a unit for the palest wine to more than eleven for a mid-depth one.
| Quantity | How it is formed | Reported to | Method |
|---|---|---|---|
| Colour intensity (I) | A420 + A520 + A620, each referred to a 1 cm path | 3 decimal places | OIV-MA-AS2-07B |
| Shade (N) | A420 ÷ A520 | 3 decimal places | OIV-MA-AS2-07B |
| Lightness L* | From the transmission spectrum; 0 is black, 100 colourless | 1 decimal place | OIV-MA-AS2-11 |
| a* and b* | Red–green and yellow–blue coordinates | 2 decimal places | OIV-MA-AS2-11 |
| Chroma C* and hue angle H* | Length and direction of the (a*, b*) point | 2 decimal places | OIV-MA-AS2-11 |
| Wine | L* (mean) | a* (mean) | b* (mean) | Chroma C* | Hue angle | Between-laboratory limit on L* |
|---|---|---|---|---|---|---|
| B | 98.0 | -0.86 | 9.04 | 9.1 | 95.4° | ±0.4 |
| A | 96.8 | -0.26 | 10.95 | 11.0 | 91.4° | ±1.7 |
| C | 91.6 | 2.99 | 17.75 | 18.0 | 80.4° | ±3.3 |
| D | 86.0 | 11.11 | 17.10 | 20.4 | 57.0° | ±5.5 |
| E | 77.4 | 20.51 | 19.68 | 28.4 | 43.8° | ±2.2 |
| F | 67.0 | 29.29 | 26.51 | 39.5 | 42.1° | ±11.5 |
| G | 34.6 | 52.13 | 45.82 | 69.4 | 41.3° | ±2.8 |
| H | 17.6 | 47.55 | 30.07 | 56.3 | 32.3° | ±2.8 |
| Driver | What it does | Reported size |
|---|---|---|
| pH | Sets the share of anthocyanin in the red flavylium form | 20–25% at pH 3.4–3.6; about 10% at pH 4.0 |
| Sulphur dioxide | Bisulfite adds to free anthocyanins, giving a colourless product; reversible | Not quantified here |
| Copigmentation | Stacking with colourless phenolics deepens colour and shifts it towards violet | 30–50% of young red colour; peak moves 5–20 nm |
| Polymeric pigments | Anthocyanins bound to tannin-type molecules; protected from bleaching | About 25% of anthocyanins after fermentation, over 40% after a year |
| Pyranoanthocyanins | Orange-toned, coloured at any pH, resist sulphur dioxide | Not quantified here |
Why: Oxygen exposure during making or storage has oxidised its phenolics.
Fix: Check the closure and storage temperature; smell for stale, bruised-apple notes that often accompany browning.
Why: Warm, dim lighting shifts reds towards brown.
Fix: Judge colour under daylight or a neutral lamp against white.
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.
FactModerate evidence
Pyranoanthocyanins form during winemaking from anthocyanins and yeast metabolites such as pyruvic acid and acetaldehyde; they express colour across all pH values and are more resistant to bleaching and more stable in storage than their parent anthocyanins.
FactStrong evidence
Wine browning results from enzymatic and non-enzymatic oxidation of polyphenols and is prevented mainly by limiting oxygen and using antioxidants; fining agents can remove some brown compounds.
FactModerate evidence
Anthocyanin colour depends on pH, and sulphur dioxide bleaches free anthocyanins, which makes the colour of young red wine less stable than that of aged wine.
Source: Anthocyanins: Factors Affecting Their Stability and Degradation; Influence of the Anthocyanin and Cofactor Structure on the Formation Efficiency of Naturally Derived Pyranoanthocyanins
Colourwise interpretationModerate evidence
The shift of red wine from purple towards brick with age is best read as a change in which pigments dominate rather than simple fading.
Based on: Colourwise's synthesis of the pyranoanthocyanin and oxidation literature cited on this page; the relative contribution of each pigment family varies by wine.
StandardStrong evidence
OIV method OIV-MA-AS2-07B, for red and rosé wines, defines colour intensity as the sum of the absorbances at 420, 520 and 620 nm for a 1 cm optical path and shade as the absorbance at 420 nm divided by that at 520 nm, each expressed to three decimal places.
Caveat: A Type IV (auxiliary) method in the OIV's classification; the indices are conventions and the method attaches no colour names to them.
StandardStrong evidence
OIV method OIV-MA-AS2-11 (Resolution Oeno 1/2006) derives L*, a*, b*, chroma and hue angle from transmittance measured between 380 and 780 nm at 5 nm intervals, using illuminant D65 and the 10° observer, with readings from thinner cells converted to a 10 mm path.
FactModerate evidence
At a wine pH of 3.4 to 3.6 about 20 to 25% of anthocyanins are in the coloured flavylium form, falling to about 10% at pH 4.0; bisulfite from sulphur dioxide adds reversibly to the flavylium ion to give a colourless product.
Caveat: Percentages reported by a review from earlier model-solution work; real wines vary with composition and temperature.
Source: Anthocyanins and Their Variation in Red Wines I. Monomeric Anthocyanins and Their Color Expression
FactModerate evidence
Copigmentation — stacking of anthocyanins with colourless phenolic cofactors — is estimated to contribute about 30 to 50% of the colour of young red wines and typically moves the absorption maximum 5 to 20 nm towards longer wavelengths.
Caveat: An estimate quoted by a review, not a measured property of any particular wine; the share varies by grape and vintage.
Source: Anthocyanins and Their Variation in Red Wines I. Monomeric Anthocyanins and Their Color Expression
FactModerate evidence
About 25% of anthocyanins are estimated to have polymerised with other flavonoids by the end of alcoholic fermentation and more than 40% after one year of ageing; polymerisation protects the chromophore from bleaching by sulphur dioxide.
Caveat: A literature estimate; the rate depends on tannin content, oxygen and temperature.
FactLimited evidence
In a comparison on 112 wines, the three-wavelength absorbance indices and CIELAB agreed only for dark red wines, and CIELAB reflected sensory perception better, though neither described every aspect of wine colour.
Measured: 112 commercial wines (56 red, 56 white) of mixed variety, origin and vintage.
Caveat: Reported in a conference abstract without the underlying statistics; treat as an indication, not a settled result.
Colourwise analysisModerate evidence
Across the eight wines in the OIV method's interlaboratory validation, mean lightness runs from 98.0 to 17.6, and the chroma Colourwise calculates from the published a* and b* means rises from about 9 to about 69 before falling to about 56 for the darkest wine.
Based on: Chroma computed as the square root of a*² + b*² from the mean coordinates in section 2 of the method; the lightness figures are transcribed.
Caveat: Eight wines chosen to span a range, identified only by letter; not a survey of wine colour.
Colourwise interpretationModerate evidence
Because shade is the yellow-brown reading divided by the red reading, both the loss of red anthocyanin colour and the growth of orange-brown pigments raise it, which is why shade climbs as a red wine ages.
Based on: Follows arithmetically from the OIV definition of shade combined with the pigment changes described in the cited reviews; Colourwise has not measured wines.
Reviewed 1 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.