How is rosé made?

Good rosé comes from much more than following a list of winemaking techniques. The terroir, climate, grape varieties and vintage are just as important, especially at the top end. Many new rosé producers around the world follow Provence-style tech-sheets as if they were instruction manuals, when in reality wine and rosémaking is so much more complicated. This is even more true during difficult vintages. Perceptions of rosé as a ‘technical wine’, whose quality depends on the rigour of the winemaking techniques or the quality of the equipment and oenological inputs, are outdated.

In the early days of rosé’s success – there was a focus on the simplicity of rosé and the lack of requirement to know about grapes and terroir – these details were never mentioned. Even today, most websites explain in detail the terroir in relation to their red and white wines, but rarely for their rosé. This encourages the image that rosé is nothing more than a technical wine – chill-press-chill-ferment-bottle. This causes no end of frustration when trying to understand why one rosé is so different from another. Some producers even feel that acknowledgement of the terroir, as with the Provence dénominations de terroir, is overly confusing rather than adding premium value. 

A major factor in the growth of rosé over the past thirty years has been the enormous focus on improved quality, to make clean, fresh wines. Faulty wines are rare.

Almost all rosés (and many or most whites) now follow variations on the following procedure:

  • Night or early-morning harvest
  • Direct press
  • Cold settling and/or cold stabulation
  • Cold, temperature-controlled fermentation in steel
  • Short lees ageing in steel or (for premium wines) oak

Although much emphasis for making quality rosé is put on the techniques required after harvest, any winemaker will say that a great wine starts in the vineyard, not the cellar. In this respect, rosé is no rosé, and a great rosé begins in the vineyard. Grapes destined for rosé have very different requirements from either white or red, and ideally the decision as to what to produce is made years ahead.

As with reds, it is widely believed that low yields produce wines of greater concentration. This can be translated either as fewer bunches, with more but smaller berries (therefore emphasising skin and pip production, which are good for reds), or more bunches of fewer but larger berries (emphasising juice and increasing the juice:skin ratio, good for pale rosés). Careful pruning can achieve either of these – but the decision must be made during the winter preceding the harvest, or before planting, as variety, rootstocks and clonal selection can make all the difference.

In the vineyard: harvest

Higher yields with more dilute juice can be partially compensated by leaching as much structure and flavour out of the skins, at the ‘expense’ of colour. This is not an option for any wine for which pale colour is desirable, and therefore low yield is significantly more important for pale direct press rosés. In our experience, one cannot achieve flavourful, structural direct-press rosés at yields over 40hl/ha. Many winemakers still claim rosés can be made from significantly higher yields than equivalent-quality reds, and that these higher yields prevent the rosés from being too heavy. We do not agree, as these rosés tend to be watery and dilute.

Multiple harvest dates to obtain freshness and fruit is more likely in premium rosé, with early-picked grapes providing acidity and later-picked ones providing the fruit. Saigné rosés, where the rosé juice comes from the grapes being used for red wine, are harvested later, at the same time as the reds. For logistical as well as winemaking reasons, the direct press rosé harvest is usually over before the reds begin. See Vintage character.

Many, but not all, harvest in the early hours of the day when the grapes are still cool, to reduce the amount of chilling needed before processing. Some use machine harvesting, others manual harvesting. Mechanical harvesting almost always involves destemming the bunches, unlike hand harvesting which allows the preservation of the bunches.

Whole bunches are better protected from oxygen than individual berries, and the stems create small gaps in the press which allow the juice to flow more freely. It is therefore possible to make paler rosés in a more reductive style with whole bunches and hand harvesting. Other than the lower cost of mechanical harvesting, there is also a clear logistical advantage: machines do not need as much light for night harvesting as teams of people with secateurs and baskets.

Night, or at least very early morning, harvesting is increasingly common, to the point of being almost universal in recent hot vintages. The grapes are naturally colder, reducing risk of oxidation on the way to the cellar, and the energy saved by not needing to artificially cool them make a non-trivial contribution to reducing carbon footprint. Numerous studies have also shown that night-harvested grapes have greater levels of thiol-precursor chemicals, although the direct benefits of this remain unproven.

Although there is a growing amount of rosé made using organic, sustainable, regenerative, and biodynamic viticulture, rosé as a whole is viewed as the antithesis of sustainability. High volume, large scale production and energy-intensive chilling are the most widely accepted impressions. 
However, the hot, dry climate in the south of France is highly conducive to organic viticulture as fungal diseases caused by humidity are relatively rare and the south has a high percentage of organic viticulture. Rosé, as a consumer-driven category, is often way ahead of this – Tavel, for instance, has the highest proportion of organic vineyards of any Rhone appellation. This is only a start, and we would agree that far more needs to be done to move the monoculture of vineyards into a more sustainable and biodiverse environment. From ground cover, hedges, vine-training, recycling water, solar panels, insulation, recyclable glass, bag-in-box, cans… progress is slow, but it is happening.

In the vineyard

Recent vintages have been increasingly dry, with 2022 set to be one of the driest on record throughout the south. Irrigation is not widespread, although it is gaining significant traction with every new planting. Opinion is divided: some see it as a necessity to maintain volumes, others say it eliminates terroir character. Permission is still required on an annual basis to irrigate AOP vines, although in practice it has been allowed in most recent vintages. In biodynamic vineyards it is not allowed at all.

The consequences of drought and hydric stress are severe and can easily be detected in wines, especially rosé. As the vine struggles for water, it prioritises survival over grape production, known in French as blocage de maturité. Phenolic ripening is halted, acidity levels drop, and the grapes remain small and concentrated. Sugar levels do continue to rise, albeit more slowly. Winemakers can either choose to wait out until the autumn rains, or harvest early while some acidity still remains. Neither option produces particularly good wine. In recent vintages early harvest has been the preferred route. This results in pale wines with very little fruit character, and a green unripe austere acidity. Alcohol levels are typically 1-2% lower than might be expected. Traditionally, ‘good’ terroir cannot protect from this, as in the past the best sites were the most exposed, hottest and driest. With climate change this becomes a liability as heat increases and water availability decreases.

Direct press

Upon arriving at the cellar, the grapes are sent to the press, where they are pressed. The juice can then be then chilled or not, then put in temperature-controlled tanks, often stainless steel. This method is known as direct press. Variations exist within the direct press method. Sometimes the grapes are destemmed before being put in the press (always the case if mechanically harvested), and sometimes, albeit rarely, crushed as well. Both of these extra steps add a little colour, but can contribute phenolics and weight to the wine.

Crushing stretches the limits of ‘direct press’, as the grapes will inevitably macerate for up to a few hours in the press while it fills up. Crushing is the iconic stamping-with-your-feet step that transforms intact grapes into a mushy, liquid pulp. Pressing is then required to separate the solids from the juice and turn the crushed grapes into grape juice. The timing of the pressing step is one of the major differences between rosés and red wines. For red wines, the pressing occurs after the crushed grapes have fermented. For rosés, it is the pressed juice that ferments.

When making premium rosés, many producers only use the very first, free-run juice (typically up to 30%) to come out of the press, and can include a slight pressing (up to 200 or 300 millibars). It is the palest juice – this is a similar technique to how blanc de noirs Champagnes are made.

The next juice to come out of the press is the jus de presse and has lower acidity levels, but can be more aromatic and fruity, and is therefore often better for easy-drinking young rosés. Press pressure is often a good indicator of the philosophy and style profile of a rosé, although it is rare to find it given on a tech sheet. Lower pressures indicate more modern wines. 

Press size is rarely discussed, but is one of the most important factors. Larger presses will take longer to fill, and the juice will be in contact with more skin as it trickles out, resulting in juice with more colour. Small, usually pneumatic presses, were one of the first major investments in Provence twenty years ago and still convey an immense advantage today. 

Equally rarely discussed is the widespread use of pectolytic enzymes. These are added to the grapes as they are placed in the press, and break down the skins to facilitate juice extraction. Free-run juice volumes can be increased by up to 15% without requiring prolonged or high-pressure press cycles, therefore keeping phenolic and colour extraction low.

Direct press, in one form or another, is used for the overwhelming majority of rosés across the south of France, and indeed the world, today.

Maceration

The major alternative to direct press is maceration. Following harvest, the grapes are crushed and left to macerate in a tank or in the press. This contact between juice and skin confers a lot of colour as well as flavour and structure. It can last anywhere from a few minutes to several days. Colour is a good indicator of this length, but an imperfect one, as the must colour depends on many factors, including temperature (cold temperatures will slow colouration), use of enzymes, SO2 levels (high SO2 will make the juice look initially pale), and variety. Grenache and Cinsault are particularly pale, whereas Syrah will look dark within a few hours.

Following this maceration, the tank is emptied of its free-run juice, and, in some cases, the remaining must will be pressed and the jus de presse will be used as well. Taking only the free-run juice results in a significantly paler wine, although often without the tannins and phenolics that are part of the appeal of this style. Chilling the grapes significantly reduces colour transfer, with some producers going as far as almost freezing them. Tavel is a maceration rosé, typically with 24-48 hours or longer on the skins, and typically uses both the free-run juice from the tank and the juice from the press. Château Gasqui’s Silice is an example of a maceration rosé using only free-run juice. Free-run juice after a short maceration of 3-4 hours at low temperature is almost indistinguishable from equivalent direct press juice.

A major drawback of maceration rosés is the drop in acidity levels of the must. This is caused by increased levels of potassium (more present in the skin than the pulp) causing a precipitation of tartaric acid, and resulting in a considerable loss in acidity.

Saigné

A subset of maceration rosés is the saigné technique. The major difference historically was that the grapes were harvested as if for red wines with the winemaker looking for dark colour and ripe tannins rather than light colour and fresh acidity. The grapes would macerate and then some of the juice would be bled off (saigné) to further concentrate the red wine. This juice would traditionally produce dark, rather heavy rosé and is the cause for the poor reputation for rosé, as the winemaker’s only priority was usually the red.

However, if done well, the results are excellent. Red wine producers who harvest in several passes may only use the juice bled off the red wine from the earliest harvest; or juice may be bled off almost immediately and a small amount added to classic direct press rosé to give some extra fruit. The length of time the juice spends in contact with the skins before bled off into its own tank can vary from a few minutes to several hours. Since the volumes bled off are often quite low, rarely exceeding 15% of any given tank, winemakers traditionally took off a small amount from several tanks of red, resulting in an erratic blend based off of whatever reds needed concentrating.

Winemakers’ tech sheets often wrongly describe saigné wines (as in a by-product of red wine) as maceration rosés. This misdescription causes extreme confusion for consumers, which is a major barrier towards these wines’ quality development. The term “saigné” should only be used when the remaining juice is vinified as a red wine, since the term refers to the red wine that is ‘bled’ rather than the rosé bled off. Less egregiously, tech sheets will sometimes describe a rosé as “maceration followed by direct press”, which usually indicates short maceration and a desire to be perceived as ‘modern’.

This confusion mostly comes from translation issues between English and French winemaking terms, but lack of rosé education in winemaking schools plays a large part too. Robert Parker himself in “The Wines of the Rhône Valley” (1987 and 1997) incorrectly and confusingly said that Tavel could be made in two ways: saigné (bleeding juice from grapes being macerated normally for red wine) and ‘méthode Taveloise’ (bleeding juice from grapes being macerated and then pressing those grapes, specifically to make Tavel rosé) – further muddying the waters.

A very small but growing number of producers are returning to traditional saigné rosés. Thanks to greater understanding of this method and technology, this has not meant the return of heavy, tannic rosés, but wines with a touch more colour and structure, which are usually more gastronomic in style.

A more recent development, still fairly rare, is the opposite of saigné, and involves adding a portion of the earlier-harvested rosé (usually the last juice from the press) to the reds to add acidity and lighten the red wine. What a reversal!

Semi-carbonic maceration

A technique with particular promise is semi-carbonic maceration, which results in a dark, extremely fruity wine that, unsurprisingly, is very reminiscent of a light Beaujolais. This involves adding whole bunches of un-crushed grapes to a tank and (sometimes) topping it up with CO2. An enzymatic intracellular fermentation then begins within the intact berries, converting sugar to alcohol and leaching polyphenols (colour and tannins) from the skin into the juice.

When alcohol levels reach about 2% abv (usually in a few days, but up to a week), intracellular fermentation stops, the grapes are pressed and the juice is then fermented as any other rosé or white. Maverick producers who are experimenting with semi-carbonic include Domaine de l’Odylée in the Côtes du Rhône and Jeff Carrel in the Roussillon. Some producers in Tavel claim that this is the most authentic, traditional method of making Tavel rosés – Domaine de l’Anglore most notably, taking the Burgundian approach one step further and fermenting in ‘cuves tronconiques’ – conical wooden vats. The results are excellent, and there are good chances this category will expand significantly in the coming years – if buyers can get over the dark colour.

For all rosés, following pressing, the must is cooled in temperature-controlled tanks or even whilst in the pipes on the way to the tanks. Exposure to oxygen is kept to a minimum, for which dry ice is often added and to cool the juice. Although winemaking in general can lie anywhere on the reductive-oxidative spectrum, most southern rosés are firmly at the reductive extreme; that is to say with minimal oxygen contact with the wine. If not carefully monitored, this can result in excessive reduction, or stinky, sulphurous rotten cabbage notes, in the bottle. The other extreme, oxidative rosés, can taste flat and have aromas of cooked apples, quince and nuts.

Chilling whole grapes or bunches before pressing, rather than chilling the juice after pressing, is more common in pale, high-end wines, as the temperature in the press strongly impacts colour extraction: the lower the temperature, the lighter the colour. The energy cost is higher. Some producers claim that it is easier to produce pale rosés from destemmed, crushed and very chilled grapes, as there is more free-run juice and therefore less need to press, and only minimal colour is taken on if they are sufficiently chilled. This is, as yet, unproven.

Before fermentation, most rosés undergo clarification, racking and settling (débourbage) procedures, and in some cases cold stabulation as well. Pea and potato protein are usually used to fine and clarify the must. As well as removing small solids, this makes the rosé significantly paler. Carbon filtration is only permissible in very small quantities in some appellations. Taken to excess, this can also remove structure and aromatics – winemakers must juggle their desire to produce an aromatically interesting wine with the need to satisfy customers’ demands for increasingly and unrealistically pale rosés.

Stabulation or cold stabulation (stabulation à froid) is a relatively new technique, more common in Côtes de Provence and for premium rosés than in the Rhône or the Languedoc, but catching on. It involves chilling the freshly-pressed juice to around 0°C for 10-14 days, while stirring a few times a day, allowing longer contact between juice and solids before settling. This is still a relatively experimental technique whose major merit is supposedly an increase in thiol precursor chemicals and therefore thiol aromatics (grapefruit, pineapple) in the finished wine.

Fermentation

Alcoholic fermentation then proceeds in much the same way as it would for a white wine, with either cultured or indigenous yeast. If white grapes are included in the blend, they must be mixed with red grapes before the fermentation is finished.

A wide range of vessels is used for fermentation: stainless steel, fibreglass or concrete tanks; jarres or amphora; and wooden barrels (mainly oak) of various sizes. Stainless steel tanks are the most frequent, especially in Provence. More unusually, Domaine Lafond Roc-Epine in Tavel have been experimenting with a stone basin quarried from a local quarry and barrels made from local trees. Temperature control during fermentation is almost universal. This is much easier in stainless steel tanks than in barrels, which require complex and expensive equipment, as pioneered by Château d’Esclans with Garrus.

Fermentation temperatures can give a variety of profiles. Colder temperatures (10-12°C) keep the rosé fresh and fruity for summer drinking. Low to middle temperatures (12-16°C) offer more evident fruit characteristics, as well as greater levels of citrus, tropical and floral notes, developing stronger thiol notes (see below) when fermentation goes up to 20°C. However, over 16°C can also lead to more complex characters and, in terms of ecology, lower energy costs. High temperatures over 20°C are rare, but can quickly take on the slightly jammier notes of light reds. Although much is made of fermentation temperatures, in our experience yeast and the initial flavour profile of the juice has a more evident impact on the final wine.

Cultured yeasts are favoured for consistency, impact on colour, alcohol and fruit profiles including strawberry, floral, and exotic fruit characters. Indigeneous yeasts, usually left to ferment at higher temperatures, tend to give the most interesting wines, but are unreliable if consistency over large volumes is required. Many producers employ cultured yeasts for their rosés, but make the effort to use indigeneous yeasts for their reds and/or whites which can lead to cellar contamination issues with the reds and whites actually using the yeast chosen for rosé. An increasing number of producers are choosing a middle path of selecting their own indigenous yeast and creating their own cultivated yeast unique to their property.

For many producers, the pressure to have their rosés ready in bottle by January or February, to reach the market by Easter prevents either longer ageing or slower fermentation. This is a major reason why many producers, even when they use ambient yeast for their red and white wines, avoid them for their rosé.

The grapefruit aromatics found in many rosés come from a family of chemicals called thiols, also often found in Sauvignon Blanc. Thiols provide grassiness, grapefruit, boxwood (privet), and even gooseberry notes. The popularity of this style of rosé is unquestionable – as seen by rosé pamplemousse, or rosé mixed with grapefruit juice, available in supermarkets all summer long in France. Thiol characteristics can be likened to other divisive wine characteristics such as jammy fruit or lots of new oak – appealing to some, but not conducive to quality or ageworthiness, especially as they typically disappear within a year or two of fermentation. Although we appreciate the delicate fresh fruity aromatics they can provide, we strongly dislike the one-dimensional profile they produce at high concentrations. This is partly a matter of taste, and thiols are not necessarily an indicator of quality. 

There are many ways to increase thiol levels. Yeast strains can be purchased expressly to enhance grapefruit characteristics, (a good reason to look out for rosés made with spontaneous fermentation!), or at least those using very neutral yeasts. Thiol-producing yeasts express themselves differently at various temperatures, varying from grassy boxwood and grapefruit at low temperatures (14-16°C) to exotic fruit at slightly higher (16-19°C). Cold stabulation is another technique that can increase thiol aromatics, as can excess nitrogen during fermentation and use of pectolytic enzymes during maceration or pressing. Certain varieties are naturally more likely to have thiol precursor chemicals in the juice. Ageing rosés on Sauvignon Blanc lees is one way to give that extra thiol character, and Syrah also has a tendency towards thiol notes.

Malolactic fermentation is uncommon, although not unheard of, especially for rosés that are subsequently aged in oak and premium cuvées that rely on richness rather than freshness. One Tavel winemaker described malolactic as ‘the soul of Tavel’, and it is seeing a comeback among some producers there. Eric Asimov claims that before the 1980s, it was widespread in Bandol too. If there is sufficient acidity in the wine, the gentle creaminess brought by malolactic fermentation produces a shockingly rounded body that plays off rather well against ripe, slightly jammy red fruit and is starkly opposed to rosés that rely on crispness and austerity. It has been described as having the same impact as an additional year of age. Good examples that showcase the strength of malo are Château Léoube’s Léoube La Londe, Régine Sumeire’s Château Barbeyrolles and Foncalieu’s Nuvoté. Many winemakers are not putting sufficient thought into whether their rosé undergoes malo: best summarised by one winemaker who told us “well, we don’t actively block it from happening, so if it happens it happens…”. Malo in a rosé is usually a very good sign. With increasingly hot summers, retaining acidity is more important, and producers are less likely to encourage malo.

Other than a few exceptions (see Fortified and Sweet rosés), all wines across the south are fermented totally dry (4g/l residual sugar is the general maximum for a dry rosé), with most having no more than a gram or two of sugar left over. Alcohol levels vary between 11.5% and 14.5%, with some strong regional patterns emerging. Tavel stands out way at the top, with most wines at 14% or 14.5% abv. Pays d’Oc is at the other extreme, often skirting either side of 12. Much of Provence and the Rhône lies in the 12.5-13.5% range, with market pressure keeping levels lower. Although a vast oversimplification, alcohol levels are still a good indicator of ripeness, concentration and phenolic maturity, and most top cuvées are not afraid of reaching 14.5 or even 15%. These levels no longer raise eyebrows, but do require that the acidity hold up. Not all high-alcohol rosés are good, and not all low-alcohol wines are bad – but it is still an excellent indicator of whether a winemaker has gone for a fashionably ‘light drink’ or a ‘wine’. In some cases, a dark colour can also indicate that the grapes were harvested for the reds and juice subsequently bled off. Anything under 12% instantly raises suspicions of early harvesting and unripe fruit.

Élevage: Oak, Tank, Cement, Jarre or Amphora

Following fermentation, some degree of ageing is universal for quality rosés, although most super-entry-level wines are bottled only a few weeks after alcoholic fermentation is completed. 

Lees ageing and bâtonnage, often in steel tanks, is the most common. It is often barely noticeable. In some cases biscuity, toasty lees characters are evident, as is a creamy, fuller body. The amount of lees ageing and oak can sometimes give a rosé a more structured tannic character which is not dissimilar to an orange wine. Lees ageing can be used to reduce the colour of a rosé, although the effect is minimal.

Oak is a hotly contested topic amongst producers. If rosé’s recent phenomenal success comes from its easy, fresh, fruity, approachable style, then surely making more complex, oaky rosés with structure and ageing potential, is contradictory?

Initially a number of producers responded to market demands for a ‘gastronomic rosé’, that is a wine which could stand up to and even support a wider range of food, and ageing their rosé in oak was seen as the solution. Sometimes these wines were exactly that – a rosé wine with oak. Some managed to achieve an oaked rosé with ageing potential, such Guy Negrel of Mas de Cadenet’s Grande Garde created in 1989. There was plenty of criticism about the betrayal of rosé’s true character, and, until ten years ago, very few producers made an oaked rosé. This started to change following the success of Garrus in 2007-8, and a growing number of estates started to produce oaked rosé, either fermented or aged in barrel. Today, oaked rosé is still a minority style, reserved for premium or ‘gastronomic’ rosés, and it seems that most serious estates, at least in Provence, now produce an oaked cuvée, with varying degrees of success. As a rule of thumb, the higher the price bracket of a rosé, the more likely it is to have spent time in oak. At the very top end, oak is almost universal, and represents a decent chunk of the €25+ segment.

In some cases it does seem that leaving the wine in barriques for a few months is seen as a shortcut to a premium rosé. Oak is a bit too often seen as a replacement for structure and fruit, rather than a way to amplify them, and we sometimes feel that some of these oaked rosés need more weight and fruit structure to carry the oak.

Oak-ageing is most common in ultra-premium Côtes de Provence and Coteaux Varois, but it is increasingly common in the Rhône too, specifically around the left-bank Dentelles de Montmirail area. It is not uncommon in traditional Tavel, and is increasingly common for high-end rosés across the Languedoc. Bandol, despite its reputation for long oak-ageing for reds and Mourvèdre’s propensity for oxidative ageing, has almost none.

In order to understand why the rosé was oaked and how that oak-ageing will affect the rosé, important questions to ask are: the origin and type of oak, the original purpose of the barrels, the age of barrels, the size of barrel, and whether it is used for fermentation and/or ageing.


Some oaky rosés are fermented in wood, while others fermented in tanks and subsequently aged in barrels. Some shout oak, for others the wood is barely noticeable. Oak can provide aromatics and flavour, but is usually used for its contribution to a wine’s structure, by imparting tannins and providing micro-oxygenation during the ageing process. Use of oak requires enough fruit and concentration to support the extra weight and structure that oak provides.

While most rosé producers in southern France use French oak, a significant number use Austrian oak (usually Stockinger), appreciated for its neutrality of flavour and focusing more on providing invisible tannic structure and non-reductive winemaking..

Although oak is the most common, chestnut and acacia can occasionally be found. Acacia especially lends wines a subtle floral note, and is a choice driven more by aromatics than structure. A good example of a partially acacia-aged rosé is Gérard Bertrand’s Clos du Temple. Chestnut is very rarely used for rosé as it is usually found to impart too much tannin to the wine.

The initial purpose of the oak is important, whether the producer is looking for overt oak character, or more neutrality. Many coopers now offer barrels designed specially for rosé. These usually are almost untoasted and aim to impart as few tannins as possible. Barrels destined for whites are the most common, adding both structure and oak aromatics. Barrels used or designed for reds are also common, and are particularly helpful for darker rosés with more body as they provide tannins and controlled oxidation to the wines. Some producers opt for the same barrels for all of their wines – reds, whites and rosés – which results in a signature profile across their range. This allows other factors such as terroir and vintage to stand out instead.

Second-use, third-use or older barrels give the best results, providing gentle structure without any aromatics. New barrels often dominate gentle wines with toasty and vanilla notes that take years to wear off, by which time the fruit might well have disappeared too. Whilst we are huge, huge fans of oaked rosés, raw new oak can detract from what should be a particularly exciting category. It is mostly a problem in estates without a long history of making rosé, and hopefully will die down in a few years. There are, of course, plenty of wines aged in new oak which work extremely well, most notably Clos Cibonne’s Cuvée Caroline which uses new oak exclusively.

Size of barrels is important, as producers are generally not searching for the tannic structure given by small 225l barriques. Medium-sized vessels (demi-muids, 400l and higher) have less surface area to the volume, which allows for less contact and therefore influence and oxygenation. These are increasingly common, and seem almost universal at the very top end. Large and very large foudres, starting at 8hl and going all the way to 60hl are very traditional throughout the south, although less common than they once were. Maby’s Libiamo Tavel is an excellent example of the beauty of demi-muids; despite using only new barrels, the oak is barely evident.

The time spent in barrels can vary, with anything from a few weeks (during fermentation only, for example Domaine Gueissard’s Marcel, leaving imperceptible oak but a little structure) to a year or more (Clos Cibonne’s Cuvée Tradition spends 12-18 months ageing). Five to six months is the most common, as this allows a wine to be bottled and released by March or April, and therefore sold during its first summer.

Fermenting in oak is seen as a way to impart oak structure and tannins without extensive ageing and the risk of oxidation. Barrels are significantly harder to temperature-control throughout fermentation than stainless steel, and many producers opt to let the fermentation proceed at naturally high, uncontrolled temperatures. This is often part of the appeal, letting more secondary and tertiary aromas shine, as well as slightly jammier fruit. Château d’Esclans pioneered a high-tech barrel-cooling system for Garrus, combining the benefits of barrel-fermentation and ageing with the cool temperatures of steel fermentation.

Partly as a result of the trend for over-oaked wines, concrete and ceramic eggs and jarres are increasingly common. These are less porous and breathable than oak, resulting in a crisper, fresher wine with less oxidative fruit character. Château Pibarnon’s Nuances (also part-aged in Stockinger barrel in some years), Domaine Alône, Château de Milles, Domaines Ott’s Etoile and Muse de Miraval are examples of concrete egg-fermented and/or aged rosé. Unlined concrete tanks are also used.

Clay can be thought of as a middle ground between steel and oak. Stainless steel allows for an oxygen-free environment and does not impart any flavours into the wine. Like oak, clay is porous, so it allows for some oxygen, giving the wine a deep and rich texture, but like steel it is a neutral material that will not impart additional flavours.

Amphora-like vessels have many names – amphorae, jars/jarres, jugs, qvevri (Georgia) and tinajas (Spain). The most commonly used material for wine amphoras is clay, including terracotta. Other materials used are sandstone (grès) and concrete, but those are not usually referred to as amphoras. Domaine Lafond Roc-Épine have even experimented with a fermentation tank sculpted into a block of Tavel limestone quarried near their vines. Results were promising, but given the immense cost, it is unlikely to develop further.

Natural rosé is no longer an oxymoron, and often captures the attention. Despite some producers’ cult status and growing volumes, natural rosé is, however, extremely difficult to find, and even harder to talk about. We cover as many natural producers as we can in their respective regional sections – but not in the following chapter on winemaking.

In this competitive market, producers are following two paths to sell their rosé. The first and most popular is to follow the pack. Seeing Provence as the market leader, a vast number of producers around the world (far too many in our opinion) The second path is to attempt to stand out from the crowd by making a very different product.

image/svg+xml