Caramels for Christmas

12 12 2010

I recently took a great food writing class through UCSD Extension with Deborah Schneider, longtime San Diego chef, author of Cooking with the Seasons at Rancho La Puerta, and owner of Sol Cocina, a modern Mexican restaurant in Newport Beach.

My final project was inspired by caramels that we made as gifts for Christmas this year. We made fleur de sel, chocolate brownie, apple pie, and Marcona almond flavors.

Christmas Caramels

Caramelized sugar is one of the few ingredients that have the power to evoke both warm feelings of home and deep terror. Standing over the pot in the kitchen with slightly spitting boiling sugar on the verge of burning, I kept telling myself that the warm happy part with soft, buttery, chewy caramel candy was coming soon. This was last Christmas when I was making a collection of sweets for holiday gifts and time was running short on getting the caramels done. People loved the flavor, but I never did get the texture right. This year, I was determined to solve the texture and do the whole process without giving myself an ulcer. I think I’ve found a good method.

Chewy caramel candies are actually an American invention that originated in the 1800’s. They were made with sugar, milk, butter, and molasses or syrup that was mixed together and boiled until the right consistency was reached. Before this, caramel was the stage in the process of boiling sugar when all the water had boiled away and the molten sugar began to color. This use lives on in flan or crème caramel, filigree spun sugar, and has been incorporated into yes, many caramel candy recipes.

The caramel recipe here starts with boiling sugar until it changes color; then you add butter, cream, and milk and cook it until it further browns and reaches the right texture. Why bother browning the sugar separately and then the milk, you ask? Well, browning sugar and browning milk proteins are actually two very different chemical processes: caramelization and the Maillard reaction. Each contributes different flavors to the overall taste profile and the combination makes for a complex, layered experience when you chew the caramel.

The Science of Caramel Sidebar
Caramelization is the process by which sugar alone turns brown. Sugar molecules start breaking down and reacting with each other higgledy piggledy. New molecules like diactyl, the molecule that makes butter and microwave popcorn taste like butter, get formed. The darker you go, the more the reaction continues. Take it to coffee color and you’ll have a smoking, bitter concoction called caramel color that is used to add color and flavor to colas. Depending on the sugar, caramelization occurs at different temperatures. Fructose caramelizes at lower temperatures (220 F), with glucose higher (300 F), and sucrose (table sugar) even higher (340 F).

When sugars are heated in the presence of proteins, a different reaction occurs between some types of sugar and the amino acids of the protein. Called the Maillard reaction, after Louis Camille Maillard, the French scientist that described it, this is the same process as browning meat in a pan. In caramel candies, the lactose from milk as well as glucose and fructose, if present, react with the amino acids in the milk and butter proteins. Table sugar, or sucrose, oddly doesn’t play a role here. Sucrose is made up of glucose and fructose, but they are bound together at the same part of the molecule that is required to react with amino acids. So, as long as sucrose stays sucrose, it doesn’t make Maillard products. Maillard reactions can occur at much lower temperatures than caramelization and we’ll see most of ours around 225-250 F.

So, why is caramelizing sugar so stressful? Well, the classic method requires split-second timing. Sugar and a little lemon juice or corn syrup are mixed together until it looks like wet sand. The lemon juice breaks down sucrose to glucose and fructose, effectively making a version of corn syrup in the pan. We’re using corn syrup because it is easier to control. The sugar is heated until it begins to melt. Stirring the sugar can make it crystallize into a giant globular mess, so it is swirled gently to mix the ingredients. The melting is usually uneven, though: some parts of the sugar may melt and darkly caramelize, while others are still crystallized. And you want to stop it before it burns. But the sugar needs to melt before adding the liquids or else crystals can form in the finished candy. And then the phone starts ringing and the kids start yelling. You get the picture.

An alternative method is to dissolve the sugar in water first. The water is then boiled off and molten sugar is left. The crystals are fully dissolved and the caramelization of the sugar is even, resulting in much more control over when we want to add the liquids. We still swirl, not stir, but we’re swirling a liquid, not a slushy mix.

The perfect caramelization point depends on the other flavors in the recipe. Light fruit flavors may benefit from a lighter caramel, while toasted nut caramels may like a darker flavor. The first time you caramelize sugar, take a sample at various points as the sugar goes through lemon, gold, amber, and coffee colors. Take some with a spoon and let cool on a lightly oiled heat-proof surface. When cool, you can taste these stages and decide what will be best for your flavors. The recipe here uses a gold colored caramel, which is reached at about 340 F. As the color gets close, take the pan off the heat. It will continue to brown, just more slowly. If you want it darker, put it back on the heat.

Adding the liquids can be exciting, but you can take it slow. Since the sugar is much hotter than the boiling point of water, all the water in the butter, milk, and cream will start to boil when you add it to the molten sugar, making for a bit of a bubble show. Start adding just a bit, stir, and wait for the bubbles to quiet down before adding more.

Choosing the time to take it off the heat and pour it into a lightly oiled container is another critical stress point. Hardness is determined by the temperature when you take it off the heat and slight variation in temperature can be important. I had been under the impression that if you cooked it too much, there was nothing you could do. You just had to deal with hard caramel. But that isn’t true. You can just add liquid and it will soften. It was like a weight had been lifted. As temperature reaches close to the end, just take the pan off the heat and test the hardness by drizzling a small amount in ice water. As soon as it goes from soft to strongly resisting, pour it out. Too hard? Just add a spoonful of milk or water and try again.

The last point is to let it rest after pouring for at least 12 hours. Touch it any sooner and the oil will start to break and ooze out of the candy. After 12 hours, you can cut easily with a chef’s knife or a pizza cutter. Wrap individually in plastic candy wrappers or waxed paper, put in a pretty box, and you’ve got an enviable holiday gift that was faster and less stressful than parking at the mall.

Recipe: Salted Caramels

The best things in life are often the simplest. Caramelized sugar, salt, and butter are a holy trinity of flavor.

220g (1 ¼ c.) sugar
30g (2 T.) corn syrup
55g (¼ c.) water

90g (6 1/2 T.) unsalted butter (splurge on the European style, here)
220 g (1 c.) half-and-half or:
160g (½ c. + 3 T.) milk
60g (¼ c. + 1 T.) heavy cream
1 t. sea or flaky kosher salt (a good sea salt like fleur de sel is great here)

¼ t. sea or flaky kosher salt

Mix sugar, corn syrup, and water in a heavy 4 qt. saucepan and stir until evenly distributed. Cover with a lid and set to high heat. The steam as it comes to a rolling boil will dissolve any stray sugar crystals on the side of the pan. When boiling hard, take off the lid and set a candy thermometer in the liquid such that the bulb is submerged, but is not touching the bottom of the pan to get an accurate temperature reading.

Heat the half and half or milk/cream mixture in a separate saucepan with 1 t. salt. Stir until the salt is dissolved. Set aside.

Boil at high heat without stirring until the temperature reaches 300 F. Then, turn the heat down to medium and monitor the solution for color change. It should turn a lemon yellow before turning golden. If the color appears uneven, swirl the pan. As the color changes, moderate the heat by taking the pan off the burner as needed. When the solution is golden, take the pan off the heat and add the butter. Stir with a wooden spoon until melted. Then, slowly add the half-and-half or milk/cream. Let bubble and stir.

Place saucepan on medium to medium-high heat and bring to a boil. Turn the temperature down if the caramel is in danger of boiling over. Boil, stirring often to constantly, until the solution reaches the desired consistency. This should be about 248 F. To test for consistency, drizzle a small amount into ice water, let cool, and taste. The candy should give resistance when bitten, and feel hard when ice cold, warming to soft at body temperature.

Pour into an 8×8 lightly greased heat-proof container, like pyrex. Sprinkle with remaining salt. Let cool at room temperature at least 12 hours. Remove from container and cut into 1×1 inch squares. Wrap in plastic or waxed paper.

Makes 64 pieces. Keeps well at room temperature for 2 months.

Yuzu Macarons

27 11 2010

Macarons filled with yuzu curd and candied yuzu peel. They are tart, sweet, and a touch of heady citron bitter on top of being the most cheerful looking cookies we’ve made.

Working on a badge, we’ve gotten a ton of inspiration and info from Kitchen Musings. The recipe is basically from Pierre Hermé, and is similar to this one.

Calissons in San Diego

8 10 2010

Ah, finally we get to try traditional French calissons. Maison en Provence, a cute shop in Mission Hills carrying specialties of the Provence region, recently started carrying calissons. They are from Chabert et Guillot in Montelimar, who appear to be particularly known for their nougat.

You can see in the picture how it breaks with a coarse crumble, but is moistly holding together. This version seems to have more water/sugar syrup than ours, is brighter orange in color, and is not ground quite as finely. I think the slightly coarser grind may help in bringing out more melon flavor — the very fine grind seems to swamp out the floral notes. The flavor reminds me of fruitcake with the almonds, candied fruit, and moist cake-like texture.

Green curry with fish dumplings and eggplant

30 09 2010

Continuing with green curry, we have David Thompson’s green curry with fish dumplings and eggplant from Thai Food. The fish is chopped finely, flavored with cilantro root, garlic, ginger, white pepper, fish sauce, and a touch of sugar before being formed into little balls that get poached. They are spiced to the level of wonton filling and are tender when broken. The fish is white sea bass from our favorite fish supplier Catalina Offshore. They’re a wholesaler, so they have very fresh fish, and for whatever reason they actually sell it to the public.

This coconut milk-based curry uses essentially the same curry paste that we used for our green curry with beef with a touch of krachai and tumeric. Krachai, also known as fingeroot and lesser ginger, is strong like ginger, but with a different flavor. It tastes almost medicinal and seems to give strength to fish and vegetable curries. We’ve gotten ours frozen at Minh Huang.

Huitlacoche with corn pots de creme

28 09 2010

Huitlacoche (wheet-la-co-che), also known as corn smut, is a wonderful delicacy, especially if you can find it fresh. It is hard to describe, but has a earthy mushroom flavor and lends a light savoriness to foods. It goes very well with corn, cream, eggs, light white cheeses (like mozzarella or Oaxacan cheese) and light amino acid rich seafood like scallops. This time of the year, Chino Farm occasionally has an ear or two and we snatch it up whenever we see it.

This preparation worked out really well. It is a savory pots de creme using corn juice instead of milk with dashes of salt, honey and butter. The huitlacoche is sauteed with shallots, butter and corn, and placed over strips of roasted red anaheim peppers. A bit of epazote is on top. It can be easy to drown out huitlacoche, but in this, it was highlighted with the roasted flavors from the peppers and the sweet creaminess of the custard.

Green Curry with Beef

26 09 2010

We’ve been working through Thai Food by David Thompson and recently got his new book Thai Street Food. This is our first foray into Street Food: a spiced green curry with beef, served with roti.

The roti needs some work and the brisket was too tough, but the curry was bright, rich, and complex with green chiles, lemongrass, shallots, galangal, coconut milk and more. The recipes are tough to get right, but feel much more authentic than any other cookbook we’ve seen on the subject.


1 08 2010

Our latest candy obsession has been calissons.

They are similar to marzipan, but have ground candied fruit, particularly melon and orange, in addition to almonds. The paste rests on wafer paper and is covered by royal icing.

Calissons have quite the history and lore regarding their name and origins. They are a specialty of Aix en Provence where they have been made in a semi-industrialized process since the 1800’s. This region of Provence was well known for it’s almonds until the early 1900’s and melons are still celebrated with an annual festival in the nearby Cavaillon. Prior to the Aix factories, they date to the middle ages, possibly coming from Italian monasteries, although they are often attributed as being first made for King Rene’s (the last king of Provence) second wedding in 1454. They retained a religious significance for some time and were said to protect from the plague.

It seems like a simple candy, but finding a recipe for these things has proved challenging. In seemingly reputable descriptions (like here, here, and here), they indicate ground almonds, ground candied fruit, and sugar or syrup from the candied fruit, but fail to provide actual directions. The ground paste is molded on top of wafer paper and then topped with royal icing as you can see in this video. The result is an almond shaped confection.

There is another version that contains egg yolk and heavy cream that seems like it just can’t be right, and others that seem close, but simplified, with ground almonds and jam. MFK Fisher even has a contribution in her book Map of Another Town. She gives the recipe (you can search for it in the Amazon book) from a factory that she never seems to get around to visiting — a pint of blanched almonds, a pound of sugar, and a few tablespoons of fruit syrup — no mention of candied fruit itself and no candied melon.

We decided that we needed to re-engineer the recipe. We based it on Greweling’s marzipan recipe, which is 50% almonds, but decreased the almonds and added candied fruit and syrup. We changed the recipe to about 35% almonds, 30% candied Cavaillon melon, 5% candied Seville orange rind, and 30% melon syrup + a little honey.

Usually, the candied Cavaillon melon is the most difficult ingredient to source. You can find it online, but it is prohibitively expensive for candy making. You can make it yourself if you can find a really good cantaloupe-ish melon. Cavaillon melons are a type a cantaloupe with a strong, honeyed flavor that is just wonderful. We get ours at the Chino Farm, but they seem like they are becoming more common at farmer’s markets and the like. It can be candied like any other fruit by letting it sit in progressively stronger sugar solutions until it is translucent and the solution is about 75% sugar. It can be dried to look more like candied fruit or drained briefly and used for cooking, as in this recipe.

The almonds we get at Terra Bella Ranch, which have the benefit of being very fresh with a clear almond flavor. Unlike European almonds, though, these do not have any additional flavor from the occasional bitter almond. So, we add a little pure almond extract to compensate.

Here’s our current calisson recipe:

350g almonds
300g drained candied Cavillon melon
50g drained candied Seville orange peel
350g syrup from melons
60g lavender honey (not the infused kind — the kind from bees feeding on lavender)
1/4 t. pure almond extract

The syrup is 75% sugar, but gets cooked to ~88% sugar, so it is >30% to compensate. The honey is in place of glucose in the recipe and is characteristic of Provence, which is renowned for their lavender honey. Historically, fruits were often candied in honey, so this is probably consistent with some version in the past.

To make the paste, we use the same method as Greweling’s marzipan. We blanched the almonds in water for 5 min, removed the skins, lightly ground them with the fruit in a food processor. This gives a roughly chopped mixture:

We boiled the syrup to 244F, poured over the fruit/almonds, lightly stirred and let cool on a marble slab.

This recipe is a little too wet, I think and the above image is probably too moist. We actually microwaved this for 1-2 min to reduce some of the water, let cool, and then proceeded with grinding. We ground the mixture to a paste in 2 batches with the almond extract in the food processor for about 8 minutes per batch. Greweling specifies a commercial processor, but our home 11-cup version works just fine. You just wouldn’t want to make too much, as the motor gets pretty hot. This resulted in a fine, smooth paste that was sticky when warm, but smooth when cool.

After this cooled, we rolled it out onto wafer paper (rice paper, but made of potato starch, available here) using .375 inch aluminum bars as guides.

If you try to cut this immediately, it is pretty sticky. But, if you leave it overnight, it hardens on the outside and becomes much easier to cut. We cut them with a slightly oiled knife, using the aluminum bar as a guide.

Although these are usually molded into tapered almond-shaped diamonds, candies like this have historically been cut into diamonds for centuries. Since we don’t have a mold, we have opted for the cut diamonds. After cutting, they are dipped into royal icing and set out overnight to harden.

These are a pretty subtle, but quite addictive candy. They are not terribly sweet and are floral from the melon and orange. I felt the honey and the almond extract really helped to bring all the flavors together. It is much like marzipan, but with a lot more character — in a light, fruity, flowery sort of way.

We’ll have to run these by the French folks in our lives to verify that they are close to the real deal (we haven’t tried the real deal yet!), but we’re quite happy with their flavor. I think a badge will be in our not too distant future!

Pâtes de fruit

5 07 2010

Well, it seems we are on the path to French candy land. Another day, I will tell you about our caramel adventures. For today, we have a pâte de fruit story. This summer, we have been exploring how to make the fruit jellies called pâte de fruit (pronounced paht de fwee, translated as fruit paste). When done right, they have bright, concentrated fruit flavors. When done wrong, well they are just nasty. My only experience with fruit jellies as a kid was spitting them out. In France, however, they are held in high regard and can often be quite expensive. So, we figured with really good fruit and a solid recipe, we could probably make something quite good.

Fruit pastes are some of the oldest sorts of candies. They are basically stiff jam that has been cut into pieces. The ingredients are the same: fruit, sugar, pectin, and some acid, but it is cooked longer. Pâte de fruit recipes often have corn syrup/glucose to prevent crystallization, too. Quince fruit was historically used as it contains plentiful pectin and tastes wonderful after a long cooking.

Now, when we started this project, we thought that pectin was pretty much the same wherever you went. Turns out we were wrong — there are a bunch of different kinds. Generally speaking, there are two types: high methoxyl and low methoxyl. Pectin is a polymer and the number of methoxyl (-OCH3) groups that are attached influence how the pectin sets. High methoxyl requires low pH and high dissolved solids (mostly sugar, about 55%) to set well. Low methoxyl requires less sugar, but needs calcium to set — it is what is used in low sugar jellies. Within low and high, there are additional ranges, with some setting rapidly and other setting slowly. Slower setting pectins are better for clear jellies and faster setting better for things with inclusions. Pectin is unfortunately not labeled very well and it can sometimes be hard to figure out what you are buying. Pectins used for making jams and jellies in the US are not pure pectin and are calibrated to certain recipes. Pâte de fruit recipes aren’t calibrated for these types of pectin. They assume pure pectin and call for things like apple pectin or yellow pectin. Apple pectin and yellow pectin are not usually found in stores. You can find them at L’Epicerie (yellow and apple) and Le Sanctuaire (yellow). If you find other sources, please post a link.

This is our starting formula for pâte de fruit, although the specifics will vary depending on your fruit.

for ~50 small pieces (can be doubled or quadrupled easily)
250g fruit puree (225g fruit puree + 25g sugar)
250g sugar
50g corn syrup
50g sugar + 7.5g pectin, whisked together
5mL lemon juice

This is pretty much the recipe that comes on the apple pectin package from L’Epicerie and is also based on the recipes for Boiron purees, which are high quality purees that are usually used for this sort of thing. We use lemon juice not because we can’t get tartaric/citric/malic acid (check your local brew supply store, here’s ours), but because when we used it, the candy was really strongly acid flavored. We feel that good fruit will have high acid content and if not, lemon juice seems to do the trick for flavor and final gelling of the pectin. We use the acids when we are not using fruit (e.g. vanilla-mint). The amount of pectin that you will use will vary depending on how much is in the fruit. We have found 7.5g/250g puree to be a good starting point.

The Fruit
Now, we finally get to the most important part — the fruit. Make sure to get the best fruit that you can. It doesn’t need to be completely ripe, but it does need to be fruit that would taste amazing when ripe. We find our fruit at Chino Farm, where they consistently produce amazing fruits and veggies. We did a few test runs with fruit we got at a less expensive market and most people had trouble identifying the flavor. By spending $3-5/50 pieces of candy more on fruit, we went from a nondescript sweet treat to a deep fruit experience.

For this summer’s project, we attempted to make 7 kinds and succeeded at 5: strawberry, vanilla-mint, beet, nectarine, and rose geranium. We failed moderately or pretty badly at two: watermelon and corn.

First off, a success:

Beet pâte de fruit:

Now, beets are not normally used for pâte de fruit, but they are naturally very sweet and the Chinos grow wonderful ones. We began with about 5 medium-large beets, peeled & chopped:

And juiced:

Since you will be stirring nonstop, it is important to have everything in place before you start. Set up the sugar, corn syrup, lemon juice, and sugar/pectin mixture in bowls.

Also, get ready whatever you are going to use to form the candies. You can use a rectangular container, lined with plastic wrap. We have these aluminum bars that we got from Online Metals that work as caramel bars. We set them up on a silpat on top of a marble board to insulate the heat from the table. There is no need to oil them as the candy will pull away cleanly.

To cook the candy, we have seen a number of different methods. You can bring the fruit puree to a boil, then add sugar/glucose, bring to a boil again, and then add pectin. When it reaches 224F, you add the acid and pour into the mold. You can also make a syrup with the sugar/glucose and a little water, boil to ~230, add half the fruit, boil, add the rest of the fruit and the pectin, boil to 224F, and add acid and pour. You always add the acid at the end. Adding fruit later I think helps the flavor, but I am not sure. For this version, we added the fruit later. Both ways work. Here’s a pic of the beet with everything but the lemon juice added:

And poured into the mold after reaching 224F and adding lemon juice:

After about an hour, the candy is cooled, ready for cutting, and tossing in sugar. We cut with a chef’s knife into 3/4 in x 3/4 in pieces. We use normal C&H sugar for this, but you can get fancier with larger crystaled sanding sugar. Here’s nectarine, beet, vanilla-mint, and strawberry.

Nectarine pâte de fruit
5g pectin

Strawberry pâte de fruit
5g pectin

Vanilla-mint pâte de fruit
enough water to make a syrup (1/2 c)
1 Tahitian vanilla pod, scrapped into syrup while cooking
bring to a boil & then add pectin

1/2 bunch fresh mint, chopped, added with acid
4g citric acid in a little water

Rose geranium pâte de fruit
1c water + 25g sugar
1 small bunch (7-8 flower clusters) rose geranium, with flower petals separated and reserved.
Bring to boil water, 25g sugar and rose geranium leaves. Add 250g sugar, and corn syrup. Bring to a boil and add 7.5g pectin.
4g acid (half citric, half malic), in a little water
When solution reaches 224F, remove from heat and add acid. Stir in acid and boil for a minute. Add reserved flower petals and pour into mold.

For watermelon, we reduced the juice about 2.5x before making it with 7.5g pectin. The result was sticky and gooey and never really set well. We’re wondering if there wasn’t too much glucose/fructose in the reduced fruit juice. We’ll try leaving out the corn syrup next time and see if that helps.

For corn, we weren’t sure what would happen. We used less pectin (5g) because we thought the natural cornstarch would add some thickening power. We also doubled the lemon juice (10 ml) because the corn doesn’t have that much acid. It didn’t set up and was super sticky. After sitting a day, it developed a crystallized crust.


The flavor, however, was unreal and I suspect a quest is in our future. We’ll try going heavy on the pectin and acid next time.

Well, did we badge? We’ve sourced pectin and fruit/veggies well. We’ve researched the science and history to some extent, and we’ve made the recipe a number of times until we usually get it right. I’d say that qualifies as a badge. However, I think the key to pâte de fruit is being able to optimize to the particular ingredients that you have. I see an advanced badge in the future about optimizing the recipe to new ingredients. Not that we need to get the corn right, mind you, but figuring out a harder one like watermelon would be nice!

eGullet Pâte de fruit
Chocolates and Confections by Peter Greweling
Hungry in Hogtown’s description of El Bulli’s Vanilla pâte de fruit
Sweets: A History of Candy by Tim Richardson

Butter Badge: Making Cultured Butter

2 05 2010

As usual, we were getting an ingredient from a good supplier when all of a sudden they stop carrying it. So then what? We have to figure out how to do it ourselves.

This is the story of how we lost butter and found it again, picking up creme fraiche/crema, marscapone, and buttermilk along the way.

We had been buying Plugra from Trader Joe’s, a higher milkfat butter that I had thought was cultured to give it a more buttery flavor. It was reasonably priced at about $4/lb and better than a lot of other butters we had tried at the time. But one day, it was just gone. When we checked, the Trader Joe’s in San Diego wasn’t intending to carry it again.

In trying to find a replacement and searching for cultured butter, we found a number of sites with directions on how to make it yourself. Here are some good ones: Positron, Cook Like your Grandmother, What Geeks Eat, Traveler’s Lunchbox, and more recently, Michael Ruhlman.

Cultured butter is a pretty straightforward process. You take heavy cream, inoculate it with bacteria that make lactic acid, let it thicken overnight, and then beat/churn it until the fat congeals together and the whey/buttermilk is expelled. The first half is very similar to making yogurt, but you start with cream instead of milk and the specific cultures that you use can be different. The butter culture bacteria are mesophilic (medium temp) and thrive at cooler temperatures (64°- 77° F) whereas the yogurt cultures are often thermophilic (warm temp) and grow closer to 110°.

Butter flavor interlude
The flavor of butter is dominated by a couple related chemicals called diacetyl and acetoin. They are made during fermentation — usually by bacteria or yeast. In wines, like chardonnay, wine makers can set up the conditions just right for yeast to make a lot of diacetyl, giving the wine a rich mouthfeel and buttery flavor. In butter, the compounds are made by bacteria (including this guy:
Lactococcus lactis subsp. lactis biovar diacetylactis). Or, if you get the cheap stuff (or butter-flavor popcorn), they just add it as a natural flavoring.

We have tried using yogurt as a culture starter (Fage and Trader Joe’s Greek Yogurt) before upgrading to a mail order butter culture from Dairy Connection. The first attempts with yogurt worked, but the flavor really doesn’t get to buttery without the diacetyl/acetoin producing bacteria. We use the Flora Danica culture — for $10 + shipping, we can culture 50 gallons of cream.

This culture has 4 bacteria friends — 2 which seem to primarily drive the acidification (Lactococcus lactis subsp. lactis, Lactococcus lactis subsp. cremoris) and 2 which drive the flavor production (Lactococcus lactis subsp. lactis biovar diacetylactis, Leuconostoc mesenteroides subsp. cremoris).

The cream that you use is surely important. And as soon as we find a good creamery in San Diego, we’ll tell you just how important. For the time being, we’re relying on heavy cream from Costco (Swiss Dairy, which seems to be based in Riverside and is part of a conglomerate Dean Foods, which also makes Alta Dena and Berkeley Farms, among others).

Like most creams on the general market, this cream is ultra-high pasteurized, and has some additives like mono and diglycerides. The price is cheaper than anywhere else we can find at about $3.10/quart. We have had very good results with it, though. The main thing to remember is always use the freshest cream possible to start. Don’t let it sit in your fridge for weeks and don’t use it if it has been sitting open in your fridge for long (a week was too long for us). It may taste ok as cream, but the bacteria somehow find a way to amplify slightly off flavors into straight yuckiness.

To culture the cream, we start by placing Flora Danica culture (1/8 t. per gallon of cream) in a clean bowl and add about a tablespoon of cream. We make sure to clean the bowl, but we don’t sterilize or go to great lengths to purify our tools. As long as the cream is fresh, we haven’t had a problem.

We let the culture thaw and then work it with a spatula to resuspend the culture in the cream, resulting in a paste.

We heat the rest of the cream to 77 F (25 C) and pour in, mixing well. Then, it gets loosely covered with plastic wrap and set on the kitchen table for 18-24 hrs until thick like sour cream.

Then, in a very boring, but very important step, the cultured cream gets refrigerated for 24 hours. If you don’t do this, the final butter texture is kind of flaky, not smooth and creamy like good butter. It is the kind of butter that breaks when you slice off a piece instead of smoothly cutting through. The flavor also seems to develop some.

After a day in the fridge, the cultured cream can be used as creme fraiche or crema. If you strained it like Greek yogurt, you’d end up with marscapone. It will keep at least 1-2 weeks in the fridge. We usually make a gallon, keeping a quart for creme fraiche and using 3 quarts for butter.

To make butter, you begin by warming the butter to about 50-60 F, which I think is supposed to help it separate when you are beating it later. We do this by placing the bowl in hot water in the sink and stirring until the cream doesn’t feel cold to the touch anymore.

Then, it is off to the mixer. We use a 6 qt. Kitchen Aid stand mixer. Because the beating incorporates air, I would suggest that you make no more than 1/2 the volume of the bowl in cream. We use the paddle attachment because whisking, although effective, seemed to incorporate more milk solids (proteins, etc.) into the butter. When we spread it on bread, the butter melted with a punctate pattern of milk solids. With the paddle attachment, it just melts.

Beat the cream on medium speed (4-5). It will take about 15 minutes or so to turn into butter. First, the cream whips like whipping cream, incorporating air to about double its volume.

Then, like over-whipped cream, it takes on a granular structure. This is the milk fat globules starting to clump together.

The globules keep clumping together and it starts to look like butter, but it isn’t there yet. You may want to reduce the speed at this point so it doesn’t splash out of the bowl.

Finally, the buttermilk separates and the pieces of nascent butter form together.

At this point, you can take it off the mixer. You’ll work it with a spatula to get the butter to become one big clump and then pour off the buttermilk.

The milk is thick and rich. If you haven’t had buttermilk from fresh butter, this will likely be a treat. It is actually drinkable and great in other recipes, which is much more than I can say of the kind in the store.

Then, you wash the residual buttermilk from butter by adding cool water and working it with a spatula. After 3-4 additions of water, the water is not milky anymore and mostly clear.

Then, pour off the last of water, work the butter some to expel the residual water, and you are ready to salt or wrap. We don’t normally salt, but this time we tried 1 teaspoon/lb of salt. Salting seems to vary between .5% and 2% salt (3-9g/lb). It sounds like you need less salt for cultured butter and this time was a little salty. Next time, we’ll do 1/2 t. per lb. To salt, you just work in the salt with the spatula. Then, wrap up for storage. We wrap 8 oz. in plastic wrap.

Three quarts makes about 2 1/2 lbs. of butter. This will keep 3 weeks or so in the fridge and gets better with time. If you are going to keep it longer than that, just put it in the freezer, preferably vacuum packed. It easily takes on fridge/freezer flavors if left too long without good wrapping.

And was it worth it? Well, although we’ve been making the stuff, I think we’re even more likely to buy fancy butter now than before. In comparisons, we think this butter is better than any uncultured butter we’ve had — including Strauss European Unsalted. It is not as good as Echire or Le Gall Beurre de Baratte. If we had to price it, we’d put it at $8-10/lb — as tasty as anything we’ll usually find in the specialty butter section of Whole Foods, but not elite. And as for Plugra? Well, it turns out it isn’t cultured at all, but has “natural flavorings” added (i.e. diacetyl). It now has a harsh artificial taste like microwave popcorn. Ah well.

Crispy Skin Duck III: steam/roasting with a pressure cooker

24 04 2010

Today brings us a new tech for crispy duck skin. We’re going to pressure cook the duck partially and finish with roasting. This is a variant of the steam/roast method that gets a lot of airplay with the advantage that pressure cooking is faster, more convenient, and might be more effective at melting fat.

The idea behind steam roasting is that the wet steam effectively melts the fat better than dry heat in the oven. Roasting a fat-reduced bird should then produce a crisp skin. Deep frying is also an option here, but we went with roasting.

We began with the usual: separate the skin from the meat with a bicycle pump. We’re getting a lot better at this and it only takes a couple minutes now. Then, we dried the bird on a vertical roaster, uncovered, and with good air flow in the fridge for about 24 hours. Here it is afterward. The skin is often described as “like parchment” and you can see the light, dry texture.

The skin under the wings doesn’t dry well, so here you can see the contrast between dried and non-dried skin. The skin is originally thick, white, and moist. The dry skin is darker, thinner, and smooth.

We seasoned the inside of the duck before steaming with a half teaspoon of salt. Salting is pretty essential here and you need to do it before cooking or it won’t permeate the meat as well. We also slashed the skin to let the fat drain more effectively.

We pressure cooked the duck for 15 min at 15 PSI (2nd ring) in our pressure cooker. We reasoned that the fat seems to melt well at 250, which is about the temperature inside the pressure cooker at 15 PSI, and that the moist steam would help it render. We placed him on top of the steaming plate with a half cup of water. When he came out, he was glistening with fat droplets. We got about 3/4 c. of fat that had rendered into the water.

After we steamed him, we poured the light glaze (2T honey in 1c. water) over him and then let him dry in the fridge for 6 hours or so.

Roasting was at 350 for about 30 min. The skin was a bit uneven here, but much was a nice color.

Flavor was excellent — this duck really doesn’t need much but a little salt. The overall fat content was still pretty rich and only just lighter than drying and roasting alone. The meat was juicy, firm, and tender — just what we want. So, so far, not a major advance with the pressure cooker, but a very nice duck. We served him with a carrot & lemon vinaigrette salad and sauteed spinach with almonds.