Excerpt, 'The Best Light Recipe'

From "The Best Light Recipe"

Everyday Macaroni and Cheese

Weighing in at about 650 calories and 40 grams of fat per serving, a bowl of homemade mac and cheese should really be a treat every once in awhile, like a slice of cheesecake. The truth, however, is that it winds up on the dinner table much more often because it's easy to prepare and kids will eat it without complaint. We wanted to develop a macaroni and cheese recipe that could be used as a weekly workhorse meal with reasonable amounts of calories and fat, but without losing too much of the cheesy flavor or creamy texture that make it such a perennial favorite.

Starting off in our cookbook library, we found dozens of recipes for low-fat mac and cheese. Making a few of these lightened recipes to get the lay of the land, we wondered if we had bitten off more than we could chew. Some of the recipes were downright awful, producing flavorless, rubbery mixtures due to large amounts of nonfat cheese, while other recipes turned out versions with a grainy texture because they included ricotta or cottage cheese. Obviously, we were on our own here. Heading into the kitchen with the motto "Make the fat count," our philosophy was that every calorie and gram of fat needed to work for us and that none would slip by unaccounted for.

With lots of ideas to test, we decided to focus on the cooking method first, then tweak the ingredients and flavors. The most common cooking method for mac and cheese is to make a béchamel (a milk sauce thickened with butter and flour) and then stir in the cheese and cooked macaroni. Another popular Cook's Illustrated method involves slowly cooking a mixture of milk, eggs, cheese, butter, and cooked macaroni over low heat until it thickens and becomes creamy. Yet a third, and somewhat unusual, method calls for boiling the pasta in a small amount of milk (the pasta starch thickens the milk to a sauce-like consistency) then stirring in cheese to finish. Giving these three methods a whirl, we were able to eliminate two right off the bat. The popular method of cooking a mixture of milk, eggs, cheese, butter, and cooked macaroni over low heat yields great--tasting mac and cheese but it simply doesn't work without the eggs and lots of cheese, both of which are very fatty; this approach was out. Also, cooking the pasta right in the milk was an interesting and naturally low-fat technique; however, tasters found the resulting flavor to be one-dimensional and the texture of the pasta to be gummy. Making a béchamel was the winning approach so far, but we still had a lot of calories and fat to trim.

Using 2 percent, 1 percent, or skim milk instead of whole milk in the béchamel was an obvious way to reduce more fat and calories. Testing them side by side in batches of mac and cheese, tasters didn't like the sauces made with 1 percent and skim milk because they tasted too thin and didn't coat the pasta well. The sauce made with 2 percent milk, however, was acceptable and helped to trim about 2 grams of fat per serving.

Wanting to reduce the fat in the sauce further, we took a closer look at the roux (butter and flour mixture) used to thicken the milk into a béchamel. Cutting back on the butter as far as we could from the original 3 tablespoons, we found we needed at least 1 tablespoon of butter to make a roux with the flour. But even 1 tablespoon of butter adds a fair amount of fat and we wondered if we could lose it all together. Making two sauces without any butter, we tried thickening the milk with either a flour slurry (flour dissolved in a liquid) or a cornstarch slurry (cornstarch dissolved in a liquid). Both slurries were able to thicken the milk to an appropriate sauce consistency, but tasters described the sauce thickened with flour as tasting grainy and pasty, while the sauce thickened with cornstarch had a smooth, silky texture. Using cornstarch instead of a traditional roux had saved us another 10 or so grams of fat per serving.

The cheese is obviously one of the heavier ingredients in the dish, and using low-fat cheese was an easy way to trim even more off the calorie and fat counts. Most mac and cheese recipes use about 12 ounces of cheddar per H pound pasta (serving 4 to 5 people), and we made batches of mac and cheese using incrementally less cheese until the tasters cried uncle. Eight ounces of cheese turned out to be the breaking point, beyond which tasters thought the mac and cheese tasted too bland. We then tested several batches of mac and cheese substituting other types of cheese (including Parmesan, Monterey Jack, and Gouda) for some of the cheddar, but tasters preferred the flavor of the cheddar alone. Finally, we made two more batches pitting nonfat cheddar against low-fat cheddar, and the tasters unanimously hated the rubbery texture and sweet flavor of the nonfat cheddar. Using 8 ounces of low-fat cheddar in place of the 12 ounces of regular cheddar, however, had already saved us a whopping 72 grams of fat in the overall recipe.

By now our recipe had been substantially reduced to just 422 calories and 13 grams of fat per serving, but we realized that somewhere along the line, we had lost the creamy, velvety texture of the original. Had we gone too far? Looking into how we could add back some of that silky texture, we landed on the idea of evaporated milk (an ingredient used in the popular Cook's Illustrated method which we had dismissed earlier). Substituting a can of evaporated milk for some of the milk in the sauce, we hit the jackpot. The evaporated milk rounded out the texture of the sauce and even fooled some of the tasters into thinking that they were eating the real deal—full-fat mac and cheese. Testing the difference between whole evaporated milk, 2 percent evaporated milk, and skim evaporated milk in the béchamel, we found that tasters preferred the texture, flavor, and fat content of the 2 percent evaporated milk. Using low-fat (2 percent) evaporated milk not only helped the mac and cheese's texture, but it actually reduced its fat and calorie count even further, to just 360 calories and 10 grams of fat per serving.Everyday Macaroni and Cheese
Serves 5

Don't be tempted to use either preshredded or nonfat cheddar cheese in this dish—the texture and flavor of the mac and cheese will suffer substantially. For best results, choose a low-fat cheddar cheese that is sold in block form and has roughly 50 percent of the fat and calories of regular cheese (we like Cabot brand).

Salt
1/2 pound elbow macaroni (about 2 cups)
1 (12-ounce) can reduced-fat evaporated milk
3/4 cup 2 percent milk
1/4 teaspoon dry mustard
1/8 teaspoon garlic powder or celery salt (optional)
Pinch cayenne
2 teaspoons cornstarch
8 ounces 50 percent light cheddar cheese, grated (about 2 cups)

1. Bring 2 1/2 quarts water to boil in a large saucepan. Stir in 2 teaspoons salt and the macaroni; cook until the pasta is completely cooked and tender, about 5 minutes. Drain the pasta and leave it in the colander; set aside.
2. Add the evaporated milk, 1/2 cup of the 2 percent milk, mustard, garlic powder (if using), cayenne, and 1/2 teaspoon salt to the now-empty saucepan. Bring the mixture to a boil, then reduce to a simmer. Whisk the cornstarch and remaining 1/4 cup milk together, then whisk it into the simmering mixture. Continue to simmer, whisking constantly, until the sauce has thickened and is smooth, about 2 minutes.
3. Off the heat, gradually whisk in the cheddar until melted and smooth. Stir in the macaroni, and let the macaroni and cheese sit off the heat until the sauce has thickened slightly, 2 to 5 minutes, before serving.

Per serving: Cal 360; Fat 10 g; Sat fat 6 g; Chol 40 mg; Carb 45 g; Protein 24 g; Fiber 1 g; Sodium 720 mg

Variation: Everyday Macaroni and Cheese with Ham and Peas

Follow the recipe for Everyday Macaroni and Cheese, adding 2 ounces deli-style baked ham, cut into 1/3-inch pieces, and 3/4 cup frozen peas with the macaroni in step 3.

Per serving: Cal 390; Fat 10 g; Sat fat 6 g; Chol 45 mg; Carb 49 g; Protein 28 g; Fiber 3 g; Sodium 880 mg.Chocolate Bundt Cake

A Bundt cake is baking at its simplest. With their decorative shape, Bundt cakes don't require frosting or fussy finishing techniques. In fact, we think they are just great served plain. This is especially true of a chocolate Bundt cake, which should taste every bit as good as it looks, with a fine crumb, moist texture, and rich chocolate flavor. Unfortunately when it comes to low-fat chocolate Bundt cakes, this is usually not the case. Despite their tantalizing looks, most of these cakes deliver a minimum of chocolate flavor, and many are dry, rubbery, and devoid of any flavor whatsoever. We were determined to improve upon these low-fat disasters and create a chocolate Bundt cake that tastes as good as it looks.

A traditional chocolate Bundt cake calls for 6 to 8 ounces of bittersweet chocolate (plus cocoa powder), but this amount of chocolate was far too much for a cake we were trying to lighten. Our first thought was to remove the melted chocolate altogether and replace it with cocoa powder, which is much lower in fat. The resulting cake tasted bitter and its texture was chalky—reminiscent of third-rate brownies. We then tried adding the melted chocolate back an ounce at a time. With 3 ounces of chocolate and I cup of cocoa, we found a good balance. The cake had a more robust chocolate flavor and a lot less fat. We then came upon a trick to develop the chocolate flavor even more. Several recipes we had seen called for mixing hot water with the cocoa powder before adding it to the cake batter. So we poured boiling water over the cocoa and chocolate to dissolve them, and found that this step not only dispersed the cocoa particles throughout the batter but also helped to bloom the flavor. In addition, we dissolved a small amount of espresso powder along with the chocolate and cocoa and added a healthy tablespoon of vanilla extract. Both flavors -complemented the floral nuances of the chocolate. We now had the best chocolate flavor possible.

The next fat-laden culprit we turned our attention to was the butter, since most cakes have 12 tablespoons or more. With other low-fat chocolate cakes (such as our Chocolate Sheet Cake, page 335) we had discovered that using oil in place of butter yielded a cake that had a much more intense chocolate flavor (the butter masked the chocolate flavor). We were curious if that would be the case with this cake as well, and indeed it was. The cake made with 1/2 cup of oil had a rich, intense chocolate flavor. Another source of fat in the cake was the eggs. In order to achieve a solid, dense structure that will hold it shape, most Bundt cakes use four or five eggs. After some tinkering, we found that we could reduce the eggs to two and increase the leavener slightly and still achieve a statuesque cake with a fine, tender crumb (and save ourselves 15 grams of fat).

Now that we had a great, complex chocolate flavor and had reduced a significant portion of the fat, we could focus on the texture. Although our cake had great flavor and a nice crumb, it still lacked moistness. We tried decreasing the flour but this compromised the structure of the cake. We considered increasing the oil or adding an egg, but these alternatives seemed counterproductive to our core mission of developing a lower fat cake. Finally, we switched from granulated to light brown sugar, which not only added moistness but also improved the flavor.
But even after making this change to the recipe we were still falling short of our goal of a really moist cake. In our initial tests we used buttermilk as the liquid base for the cake. We liked the tangy flavor the buttermilk added to the cake, but we began to wonder if the thickness of the buttermilk was causing the cake to seem dry. We tried making the cake using low-fat milk. This cake was an improvement, but we were now curious about whether the dairy was masking the chocolate flavor (just as the butter did). So we made the cake again but this time we omitted the dairy and used water. It gave us the results we wanted: The batter was looser, the finished cake was significantly more moist, and the chocolate flavor intensified.

Up until now, we had simply been whisking everything together by hand, but it wasn't easy. The batter is heavy, there are a lot of ingredients to incorporate, and it takes a lot of batter to fill a Bundt pan—stirring the ingredients turned out to be more like arm wrestling. Looking for an easier solution, we tried both an electric mixer and a food processor. The electric mixer didn't work as well as we had hoped because the batter became overworked before all the ingredients had been incorporated—the result was a tougher, denser cake. We then turned to the food processor. We pureed all the liquid ingredients and melted chocolate together to a uniform, emulsified consistency (which also helped ensure that the chocolate was thoroughly melted and dispersed throughout the batter). But when we added the dry ingredients, we hit the same problem that we had with the electric mixer—overworked batter. We found an easy fix—a dual mixing method. After pureeing all the liquid ingredients together, we poured the liquid into a large bowl and gently whisked in the dry ingredients by hand.
Sprinkled with confectioners' sugar, our cake was almost perfect. But some of us felt that it needed something else to finish it, especially if we were going to serve it to company. Frostings and whipped cream concoctions were out of the question, but an intensely flavored glaze was just the ticket. While a glaze added calories, we found that if it was strongly flavored, we didn't have to use a lot of it. And considering we had cut out 100 calories per serving and significantly reduced the fat, we think this small splurge is worth it.Chocolate Bundt Cake
Serves 16

Although you can substitute natural cocoa for Dutch-processed, the cake won't rise as high. If you don't have baking spray with flour, mix 1 tablespoon butter with 1 tablespoon cocoa powder into a paste and brush inside the pan.

Nonstick baking spray with flour
3 ounces bittersweet chocolate, chopped fine
3/4 cup (2 1/4 ounces) Dutch-processed cocoa powder
1 teaspoon espresso powder
1 cup boiling water
1 3/4 cups (8 3/4 ounces) unbleached all-purpose flour
1 teaspoon salt
1 teaspoon baking soda
2 cups packed (14 ounces) light brown sugar
1/2 cup vegetable oil
2 large eggs
1 tablespoon vanilla extract
Confectioners' sugar, for dusting (optional), or a glaze (optional), recipes follow

1. Adjust an oven rack to the lower-middle position and heat the oven to 350 degrees. Lightly spray the inside of a standard 12-cup bundt pan with the nonstick baking spray.
2. Combine the chocolate, cocoa, and espresso powder in a large bowl. Pour the boiling water over the chocolate mixture, cover, and let sit for 5 minutes to melt the chocolate. Whisk the chocolate mixture until smooth, then set aside to cool slightly, about 2 minutes. In a separate bowl, whisk together the flour, salt, and baking soda.
3. Process the melted chocolate mixture, sugar, oil, eggs, and vanilla together in a food processor until smooth, about 1 minute. Transfer the batter to a large bowl. Sift half of the flour mixture over the batter and gently whisk in. Repeat with the remaining flour mixture and continue to whisk the batter gently until most of the lumps are gone (do not overmix).
4. Pour the batter into the prepared Bundt pan and smooth the top. Wipe any drops of batter off the sides of the pan. Bake until a toothpick inserted into the center comes out with a few moist crumbs attached, about 50 to 55 minutes, rotating the pan halfway through baking (do not overbake).
5. Transfer the cake to a wire rack and cool in the pan for 10 minutes. Invert the cake onto the rack and cool completely, 1 to 2 hours. When cool, dust with confectioners' sugar (if using), or drizzle with one of the glazes below and let the glaze set before serving, about 15 minutes.

Per serving: Cal 320; Fat 11 g; Sat fat 2 g; Chol 25 mg; Carb 51 g; Protein 6 g; Fiber 5 g; Sodium 170 mg
Per serving with glaze: Cal 350; Fat 11 g; Sat fat 2 g; Chol 25 mg; Carb 58 g; Protein 6 g; Fiber 5 g; Sodium 180 mg.New York Cheesecake

Of all the desserts that people long for in a low-fat form, cheesecake is probably the most popular—but it's also the most difficult to lighten. Since just one modest slice has more fat than most people should consume in a day, never mind in one dessert, cheesecake would require a serious makeover, and we had our doubts that it would be a successful one. Truthfully, we'd rather eat cheesecake once a year and really enjoy it than settle for something that didn't deliver its trademark tang and creamy texture. But we were determined to give it a shot. Low-fat recipes in hand, we headed into the kitchen to get a handle on the landscape of low-fat alternatives.

Just about every low-fat cookbook has tried to create some version of a reduced-fat cheesecake. The result is usually a rubbery, gummy mess, chock full of artificial and off flavors. And it's no wonder: Cheesecake is pretty much 100 percent fat in a graham cracker crust. Removing the fat would take away everything we love about the ideal New York cheesecake: its thick, smooth, satiny, creamy core, a velvety exterior, and sweet and tangy richness throughout. Or would it?

A standard New York cheesecake contains 580 calories and 43 grams of fat per slice. These jaw dropping numbers mostly come from the whopping 2H pounds of cream cheese that make up the base. In order to trim some fat it was obvious we were going to have to find an alternative to the cream cheese. We had hoped that we could simply substitute nonfat cream cheese for the full-fat cheese, but we quickly realized this was not the case. Cheesecakes made this way had an unsettling gummy mouth feel and a dry, chalky texture. Using all Neufchâtel reduced-fat cream cheese still gave us a cheesecake with a high amount of fat. Light cream cheese was a better alternative but the flavor was still off. Instead, we knew we'd need to cut one of these lower fat cream cheeses with something else. We tried tofu, ricotta, and low-fat mayonnaise. In each case the cakes were beyond disappointing. Either they tasted terrible or they had a texture that bore a striking resemblance to linoleum. That was until we tried cottage cheese. We had seen several recipes with cottage cheese, but had dismissed this ingredient because we were concerned it would make the filling grainy. But then we wondered, what would happen if we processed the cottage cheese? We put a pound of 1 percent milkfat cottage cheese in the food processor and let it run until it was perfectly smooth. The processed cottage cheese with a pound of light cream cheese was a great combination. The filling was smooth and rich with just the right density. There was one drawback, however. It was a little loose from the liquid in the cottage cheese. We simply spooned the cottage cheese into a bowl lined with paper towels and allowed them to draw off some of the moisture while we prepared the other ingredients in the cheesecake. This worked perfectly, wicking away the majority of the moisture.

With the cottage cheese–light cream cheese combo we had an ideal base and knew we were on a roll, but we also knew it was missing that trademark tang that is essential to any great New York cheesecake. Most cheesecakes use sour cream to provide this tang, but when we tried to substitute no-fat and low-fat sour cream we found that the cakes had a tacky texture and odd aftertaste. We then tried yogurt, and while it certainly provided the tang we sought, the yogurt was too watery and made a cake that was too runny. Someone in the test kitchen then suggested yogurt cheese, also called labne, which is essentially plain yogurt drained of its whey (see page 8 for more information). It was the perfect solution. It offered a touch of tartness and helped give the cheesecake a smoother, creamier texture without making it too wet. We decided to make our own yogurt cheese using low-fat plain yogurt (see page 8).

Eggs help bind a cheesecake, giving it structure and a creamy texture; they also contribute a fair amount of fat. While some cheesecakes use a combination of whole egg and yolks, we decided to stick with just whole eggs in order to get the most structure with the least fat. We tried as few as one and as many as six whole eggs, and found that three eggs were the right number. Three eggs helped build a firm, rigid structure and gave our cheesecake the requisite dense, velvety texture without adding an excessive amount of fat. Our cheesecake was now on its way and suddenly there was renewed interest among the once-doubtful test kitchen staff: Volunteer tasters were beginning to crawl out of the woodwork.

Perfecting the flavor of the cheesecake was easy. Lemon juice is an ingredient commonly found in New York cheesecake and we liked the way it brightened the other flavors in our cheesecake, but found it clashed slightly with the already tangy filling. For us, lemon zest was a better route. It provided a floral note and subtle tang without being too aggressive. Just a bit of salt (cream cheese already contains a good dose of sodium) and a couple teaspoons of vanilla extract rounded out the flavors. Simple was decidedly better—everyone in the test kitchen appreciated this minimalist approach.

With the filling behind us we could now turn our attention to the all-important graham cracker crust. In order to keep fat and calories to a minimum we considered a simple dusting of crumbs on the bottom of the cheesecake, but tasters were unanimous in their protest. They wanted a crust with more presence. We tried making a thicker crust with low-fat graham crackers, a tablespoon of butter, and no sugar. But this crust was very wet. We tried brushing the crust with egg whites in order to form a protective barrier, but this gave the crust an odd, spongy texture. It was obvious we needed the butter to provide a moisture barrier, and we would have to use more of it. We increased the butter a tablespoon at a time, and settled on 4 tablespoons of butter. This amount gave the crust a rich, toasty flavor and crisp texture. Although this was more fat than we ideally wanted in the crust, we would rather have a little bit more fat than a soggy bottom.

When it came to baking the cheesecake we preferred the standard New York method—500 degrees for about 10 minutes, then 200 degrees for about an hour and a half. This method yielded the nut-brown surface that is a distinguishing mark of an exemplary New York cheesecake. It also produced a lovely graded texture, soft and creamy at the center and firm and dry at the periphery. To ensure a properly baked cheesecake, we found it best to bake cake to an internal temperature of 150 degrees. Although it may seem unnecessary, an instant-read thermometer inserted into the cake is the most reliable means of judging the doneness of the cheesecake.
After the cheesecake had chilled, we knew we had achieved perfection. It sliced into a neat slab with a cleanly set center texture—not a wet, sloppy one. Each slice kept its shape, and each bite felt satiny on the tongue—cheesecake heaven.New York Cheesecake
Serves 12

Be sure to use light cream cheese in this recipe—it is most commonly sold in tubs, not blocks (see page 339). You can buy low-fat yogurt cheese (also called labne) or make your own with low-fat yogurt—allow at least 12 hours for the yogurt to drain (see page 9 for instructions). For a more lemony flavor, use the higher amount of lemon zest.

Crust:
9 whole graham crackers (5 ounces), broken into rough pieces and processed in a food processor to fine even crumbs (about 11/4 cups)
4 tablespoons (1/2 stick) unsalted butter, melted
1 tablespoon sugar

Filling:

1 pound 1 percent cottage cheese
1 pound light cream cheese, at room -temperature
8 ounces (1 cup) low-fat yogurt cheese (see note)
1 1/2 cups (101/2 ounces) sugar
1/4 teaspoon salt
1/2–1 teaspoon grated zest from 1 lemon
1 tablespoon vanilla extract
3 large eggs, at room temperature
Vegetable oil spray
Fresh Strawberry Topping (optional), recipe follows

1. For the crust: Adjust an oven rack to the middle position and heat the oven to 325 degrees. In a medium bowl, stir together the graham cracker crumbs, melted butter, and sugar until combined. Transfer the mixture to a 9-inch springform pan and press evenly into the pan bottom. Bake the crust until fragrant and beginning to brown, 10 to 15 minutes. Cool on a wire rack. Increase the temperature to 500 degrees.
2. For the filling: Meanwhile, line a medium bowl with a clean dish towel or several layers of paper towels. Spoon the cottage cheese into the bowl and let drain for 30 minutes.
3. Process the drained cottage cheese in a food processor until smooth and no visible lumps remain, about 1 minute, scraping down the work bowl as needed. Add the cream cheese and yogurt cheese and continue to process until smooth, 1 to 2 minutes, scraping down the sides of the bowl as needed. Add the sugar, salt, lemon zest, and vanilla and continue to process until smooth, about 1 minute, scraping down the sides of the bowl as needed. With the processor running, add the eggs one at a time and continue to process until smooth.
4. Being careful not to disturb the baked crust, spray the insides of the springform pan with vegetable oil spray. Set the springform pan on a rimmed baking sheet. Pour the processed cheese mixture into the cooled crust.
5. Bake for 10 minutes. Without opening the oven door, reduce the oven temperature to 200 degrees and continue to bake until an instant-read thermometer inserted into the center of the cheesecake reads 150 degrees, about 11/2 hours.
6. Transfer the cake to a wire rack and run a paring knife around the edge of the cake to loosen. Cool the cake at room temperature until barely warm, 21/2 to 3 hours, running a paring knife around the edge of the cake every hour or so. Wrap the pan tightly in plastic wrap and refrigerate until cold, at least 3 hours.
7. To unmold the cheesecake, wrap a hot kitchen towel around the springform pan and let stand for 1 minute. Remove the sides of the pan and blot any excess moisture from the top of the cheesecake with paper towels. Let the cheesecake stand at room temperature about 30 minutes before slicing.

Per serving: Cal 340; Fat 13 g; Sat fat 8 g; Chol 85 mg; Carb 41 g; Protein 11 g; Fiber 0 g; Sodium 530 mg
Per serving with topping: Cal 400; Fat 13 g; Sat fat 8 g; Chol 85 mg; Carb 57 g; Protein 11 g; Fiber 1 g; Sodium 540 mg

Fresh Strawberry Topping
Makes about 3 cups

A ruby-colored, glazed strawberry topping is the classic accompaniment to New York cheesecake. This topping is best served the same day it is made.

1 pound strawberries, hulled and cut lengthwise into 1/4- to 3/8-inch wedges
1/4 cup sugar
Pinch salt
1/2 cup strawberry jam
1 tablespoon juice from 1 lemon

1. Toss the berries, sugar, and salt in a medium bowl and let stand until the berries have released some juice and the sugar has dissolved, about 30 minutes, tossing occasionally to combine.
2. Process the jam in a food processor until smooth, about 8 seconds. Transfer the jam to a small saucepan and bring to a simmer over medium-high heat. Simmer, stirring frequently, until dark and no longer frothy, about 3 minutes. Stir in the lemon juice, then gently stir the warm jam into the strawberries. Cover with plastic wrap and refrigerate until cold, at least 2 hours or up to 12. To serve, spoon a portion of sauce over individual slices of cheesecake.

Per 1/4-cup serving: Cal 60; Fat 0 g; Sat Fat 0 g; Chol 0 mg; Carb 16 g;
Protein 0 g; Fiber 1 g; Sodium 10 mg

Reprinted with permission from The Best Light Recipe by the Editors of Cook's Illustrated magazine. Recipes, subscription information and cookbooks are available online at www.CooksIllustrated.com