Change of Address




I began this renewed blog project in the Summer of 2014 as an outlet  to share random ideas and occasional deep thoughts relevant to the pastry kitchen.  Released with little fanfare,  I am happy to say that in its short life it has built a dedicated following of cooks who appreciate digging a little deeper to explore the misunderstood aspects of  the sweet arts and sciences.

I’m also happy to report that future minor works will be now be published under the umbrella of the amazing Lucky Peach magazine as a monthly column on their new site. In addition to new posts that loosely fit within a monthly theme (for January, I break down alkaline ramen noodles), the complete archive of  Opsuculum will find its home there too.

Thanks to the Lucky Peach team for bringing me aboard among so much great content  – the archives will remain here for a bit, but please head on over  to for more reading!


Maillard Reactions


How cool it must have been to be present – a fly on the cave wall – at the precise moment humans first applied fire to food and the realization that the transformations due to heat  made their hunted and gathered fare taste better. Millennia of food science and cooking technology sought to understand and refine those efforts, but it wasn’t until Louis-Camille Maillard and his discovery just a century ago that we fully began to understand the delicious mechanism of color and flavor of cooked food.

In order to understand Maillard reactions, it’s best to compare its effects with that of a commonly used (though often incorrect) term to describe browning in cooked foods. Though we usually refer to browning and flavor as the result of ‘caramelization’ – that term is actually quite narrowly defined as the pyrolysis, or decomposition, of sugar. Lucky for us, the pleasing amber color of decomposing sugar tastes good! At temperatures exceeding 150˚C/300˚F sucrose undergoes an irreversible, simultaneous change of chemical composition and physical phase. Caramelization remains an active area of study as it is a complex and poorly understood process that produces hundreds of chemical products.

In contrast, Maillard reactions produce browning and chemical compounds under a different set of circumstances. Much of what we refer to as ‘caramelized’ is actually ‘Maillard-ized’ – browned onions, the browning of seared meats, or the golden crust of a freshly baked loaf of bread. Maillard reactions also play important roles in the production of chocolate, coffee, maple syrup, and the malting process used in beer.

Pain au Chocolat

The reaction was named after French chemist Louis-Camille Maillard (1878-1936) following research he conducted in 1912 and published in the Comptes Rendus, in an attempt to figure out how amino acids linked up to form proteins. He discovered that when he heated sugars and amino acids together, the mixture slowly turned brown. But it was not until the 1940s that people noticed a connection between the browning reaction and flavor. World War II soldiers were complaining about their powdered eggs turning brown and developing unappealing flavors. After many studies done in laboratories, scientists figured out that the unappetizing tastes were coming from the browning reaction. Even though the eggs were stored at room temperature, the concentration of amino acids and sugars in the dehydrated mix was high enough to produce a reaction.

Most of the research done in the 1940s and 1950s centered on preventing this reaction within the context of the growing processed food industry. Eventually, however, scientists discovered the role the Maillard reaction plays in creating flavors and aromas. A landmark paper in 1953 by John E. Hodge – a chemist working at the USDA field office in Peoria, Illinois – further fleshed out Maillard’s earlier work and led to the elegant ‘Hodge Scheme’. His ‘Chemistry of browning reactions in models systems’ stands as the most cited article in the Journal of Agricultural and Food Chemistry to this day. There remains much to learn, and the study of Maillard reactions has broad interest in fields far beyond cooking.

Lous Camille Maillard

If you read French and feel like looking at Maillard’s full 1913 thesis, download: Maillard Thesis

Unfortunately, Hodge’s paper is not readily available in full, here is a recollection: Hodge, Citation Classic 1979

The Maillard reaction is, to put it simply, a series of reactions between an amino acid (the building block of protein) and a sugar. This phenomenon is responsible for brown colors and countless aromas and flavors in cooked foods. The basic necessities for Maillard reactions are:


Reducing sugar (such as glucose, fructose, lactose, maltose, or ribose)

Increase of temperature

Removal of water

And though not necessary, an increase of pH also promotes Maillard reactions

A great example of the similarities and differences between caramelization and Maillard can be seen by comparing a conventional caramel sauce with dulce de leche. With the former, one cooks sucrose beyond the threshold of caramelization and finishes with dairy, such as cream and butter. The latter can be easily made by boiling a can of condensed milk for a few hours. The presence of milk proteins from the beginning of the process and the lower temperature of the dulce de leche demonstrate the difference, as does the flavor of each when tasted side-by-side. Similarly, a typical soft caramel could be prepared using both approaches – either caramelizing the sucrose first, or by combining the sucrose with dairy at the outset – in both cases, color and flavor from Maillard reactions will begin to occur just before the typical final cooking temperatures, in the neighborhood of 118˚C/245˚F.

It is true that browning can be the result of both caramelization and Maillard. It is also important to note that sucrose is not a reducing sugar, and thus will not directly promote Maillard reactions. However, in special cases where some of the sucrose is inverted during the cooking process (cleaving its glucose and fructose molecules) Maillard reactions can occur without the addition of other reducing sugars.

With protein, sugar, and heat (and the removal of water due to evaporation) we pass a miraculous threshold of flavor. Why remove water? It’s not until water is removed that we are able to exceed temperatures well above 100˚C/212˚F. If we are patient, however, we may get Maillard products at far lower temperatures –even in aged cheeses or Champagne, which can develop interesting “toasty” flavor profiles. The slow rate of Maillard reactions below 100˚C explains why we are compelled to pre-sear (or post-sear) meats when they are cooked sous vide. We already know that the fermentation of cocoa beans greatly affects the flavor of chocolate – an interesting possibility is that fermentation could increase the types of amine-containing molecules that may be available to participate in Maillard reactions.

Pain Seigle

The following five examples can demonstrate how Maillard reactions differ from caramelization:

‘Caramelized’ Onions – While onions do contain some naturally occurring sucrose that contribute to the golden color and sweet flavor of browned onions, that browning is due primarily to the heating of proteins and simple reducing sugars like glucose and fructose – remember, sucrose does not promote Maillard reactions. As the sliced onions are heated, moisture at the surface begins to evaporate. As temperatures climb well above the boiling point and more water evaporates, brown colors develop and hundreds of different flavor compounds are created. These compounds, in turn, break down to form yet more new flavor compounds, and so on. Each type of food has a very distinctive set of flavor compounds determined by the quantity and proportion of different amino acids. In the case of the onions, the browning reactions can also be promoted by the addition of a small amount of baking soda, which will raise the pH.

Seared Meat – While it may seem obvious that meat contains protein, it may not seem so obvious that sugars are present in meat as well. Ribose, for instance, is a simple sugar found in beef, pork, salmon, chicken, and some mushrooms. When a piece of meat is set into a hot pan, the immediate sizzle we hear is the sound of water leaving the surface of the meat and evaporating on the surface of the hot pan. The combination of protein, sugar, heat, and removal of water all contribute to the flavorful brown crust that results. If the same piece of meat was dropped into boiling water, it would not achieve the same browning and complex flavor, because the temperature of the boiling water (100˚C/212˚F) is not hot enough to trigger the Maillard reactions, not to mention that there would be no loss of water.

Browned Crust of Bread – As a loaf of bread bakes in a hot oven, complex reactions are taking place; the gas produced by the yeast expands and the proteins and starches set and gelatinize to form the final structure of the loaf. On the exterior surface of the loaf, proteins and sugars (during the fermentation process some the flour’s starch converts to maltose) react with the intense heat of the oven (usually well above 204˚C/400˚F); and with rapid evaporation of water, the crust develops and browns. This browning and complex flavor does not occur within the center of the loaf because the interior temperature may only reach 93˚C/200˚ – far too low for Maillard reactions – and because desirable moisture is trapped within the loaf.

Brown Butter – As discussed below, brown butter is prepared by heating butter and gently cooking until the solids begin to brown. This browning occurs after all of the water (16% of the butter) has evaporated and the temperature increases; and of course because the remaining nonfat milk solids are made up of proteins and lactose. This immediately demonstrates for the pastry chef how dairy products added to almost any preparation may, under the right circumstances produce these complex Maillard browning and flavor reactions in baked goods (like the classic financier) and confectionery items (soft caramels and more). The dulce de leche described above is another example of how milk solids (of which sweetened condensed milk contains roughly 23%) help create flavor.

Pretzels – The characteristic flavor and brown color of a pretzel is similar to that of the loaf of bread described above, but with an additional treatment with a highly alkaline solution. Before baking, pretzels are dipped into, brushed, or sprayed with a mixture of water and an alkaline – either weak, like baking soda (sodium bicarbonate), or very strong, like lye (sodium hydroxide). This rise in pH greatly exaggerates the browning reactions, even more so than in the browned onions example above.


Novel Applications

It was my foray into dairy and ice cream science that helped me to connect the Maillard dots in my everyday cooking. Another significant ‘a-ha’ moment came soon after with the discovery of what a few chefs were doing with white chocolate – counter-intuitively ‘roasting’ it to add much-needed character. Such thinking shows how a better understanding of composition and function lead us in creative directions. My frustration in perfecting a brown butter ice cream a decade ago was another exercise with inventive results.

The original formula I had been using for the brown butter ice cream, nicked from a Michelin-starred chef, was delicious, but fickle due to its high fat content – often resulting in a coarse, grainy texture. A few ill-informed attempts to adjust the recipe still didn’t yield favorable results. However, once armed with the data and the knowledge of how to balance the water, solids, and fat of an ice cream, it seemed logical to focus on the browned butter solids that make the flavor of the ice cream so special.

Brown Butter Ice Cream

Good quality butter will have a fat content of at least 82% by weight; that leaves 18% and the water in butter accounts for most of it, about 16% of the total. Nearly all of that water cooks off or evaporates in the process of browning butter. So really, the most important constituent of butter is everything that’s left: the 2% that is comprised of nonfat milk solids. When clarifying butter, these proteins and sugars are the foamy scum that we carefully skim off; for brown butter, we carefully allow those solids to brown. As the 16% fraction of water has cooked off, we’re left with just the fat and the solids. Once that water has evaporated, the fat can finally exceed that temperature barrier of water’s boiling point where the Maillard magic happens.

We can make our own butter by over-whipping cream to the point where the fat molecules jam themselves together, ‘squeezing’ out a good deal of water. And in that water- the original buttermilk- remain the majority of its milk solids. So if we were to compare equal measurements of butter and cream, we’d discover that the cream contains 6% milk solids by weight, or three times more of the solids obtainable from the butter. It’s the same process as for the butter, but as one would imagine, it just takes longer. As more and more water evaporates, the natural emulsion of the cream breaks up, causing the non-fat solids to separate from the fat. Whereas the tiny brown particles from butter are difficult to extract, these non-fat solids tend to clump together. When carefully strained we have a sort of brown butter ‘powder’ that can be added to just about anything. There remains some amount of fat among the solids, but little enough to easily adjust for in our recipe. There is however a very simple alternative solution to increase the non-fat solids, one that I initially overlooked: one can simply increase the solids by adding extra non-fat milk powder. This eliminates the necessity to stand over a pot of reducing cream or butter, and also greatly boosts the yield of the non-fat solids.

A great paper detailing Maillard with dairy products: Dairy Maillard. Food & Function 2012

Stages Brown Milk Solids

To see how Maillard works in its most simplistic form, we can try an experiment found in one of my favorite books, Harold McGee’s The Curious Cook. By heating some corn syrup (glucose) and amino acid supplements (such as lysine, cysteine, or glutamine) we should be able to see the Maillard browning and perhaps even isolate some of the specific aromas that might result from different amino acids. And since we mentioned caramelization, read Harold’s excellent post on the subject.

Below, a few of my favorite applications high-lighting Maillard reactions:

Brown Butter Ice Cream

Roasted White Chocolate, Pistachio, Raspberry


Big thanks to NYU Chemistry Professor Kent Kirshenbaum, my partner in two Maillard-related presentations for the World Science Festival

For further research:

The Curious Cook: More Kitchen Science and Lore, by Harold McGee 1992

Catching Fire: How Cooking Made Us Human, by Richard Wrangham 2009

Dulcey Cremeux, L2O 2013

Porcini, Bacon, Fig, Financier

Pâte de Fruit

Grapefruit-Campari Pate de Fruit

Pâte de fruit is boring.

Maybe I should qualify that by saying instead that there is a reason why it ends up on most restaurant petit four programs. It’s fairly easy to prepare in bulk, it has a long shelf life when stored properly, and it’s relatively inexpensive in terms of ingredients and labor. There is a reason why an assortment of petits fours usually represents a round-up of the usual suspects (along with marshmallows, caramels, macarons, etc.). Rarely can one devote unlimited resources (time, space, money) to upgrade these tiny bites to the next level of complexity and deliciousness. That’s why – at least for pastry cooks – they tend to get boring, even with the occasional variation.

That’s a harsh stance, but I do feel room for innovation remains. As with most preparations, possibilities expand exponentially when afforded greater insight into their inner working. Below, a few fast facts regarding pectin gels such as pâte de fruit

Low Sugar Carrot Pate de Fruit


The broad range of available pectin and their properties present an understandable degree of confusion, compounded by the fact that in many cases our access to that spectrum is actually limited to vaguely defined products provided by a handful of purveyors.

Pectin exists in nature in most plant material as a high molecular weight polysaccharide ‘glue’ that holds its cell walls together. Some plants and their fruit contain more pectin than others, and its solubility is affected by factors such as ripeness of the fruit. Historically commercial pectin has been extracted from citrus peels and apples through a process of hydrolysis – using heat, water, and acid. Factors ranging from the type of raw material to the processing method will affect a pectin’s properties, primarily its DE, or ‘degree of esterification’, which determines its methoxyl content. While I don’t think pastry chefs necessarily need to fully understand the precise chemistry involved (I certainly don’t), what is important to understand are the key differences in properties of either high- or low-methoxyl pectin.  Further, low-methoxyl pectin can be classified by DA, or ‘degree of amidation’, which also affects its performance. When discussing pâte de fruit, we will concern ourselves primarily with high-methoxyl pectin.

Low-methoxyl pectin does open up a world of possibility; it gels in the presence of calcium and within a broad range of solids content and pH. Generally, LM pectin is also thermo-reversible where HM pectin is not; one cannot ‘melt down’ an HM pectin gel like pâte de fruit and recast it. A common version of LM pectin available to pastry chefs is known as NH pectin, which is often used in preparations like neutral glaze.

High-methoxyl, or HM, pectins are those with a DM exceeding 50%, and their primary function is gelation. Pectins gel under conditions of reduced water activity – or a high degree of soluble solids such as sugars – and low pH. Generally an HM pectin will gel when the soluble solid content is between 55%-85% and the pH is in the range of 2.0 to 3.8. Generally the higher DM of a pectin, the easier the gel forms; a pectin of very high DM, for example 72%, will gel at a very high temperature and thus sets very quickly and is referred to as ‘rapid set’ pectin. ‘Slow set’ pectin may have a DM of 60% and gel at a lower temperature. Though the attributes of a gel are due to many factors including the specific recipe and its ingredients, often pectin is chosen based on processing needs. With pâte de fruit, I may wish to seek out the flexibility of slow set pectin if I am looking to deposit individual pieces, as opposed to casting a large slab to be cut later. Rapid or slow set can also play a role in aiding the suspension of solid pieces (rapid) or reduction of air bubbles (slow). If using a vacuum cooker (see below), the much lower cooking temperature would require a slow set pectin.

Final strength tends to be standardized by the industry (typically referred to as SAG 150±5), thus setting time and temperature is a common  way of distinguishing one pectin from another; below is a general guideline of such ranges, though an increase in amount of solids or a lower pH can also affect the setting conditions :

Pectin DE and Setting Temperature

The pectin we may have available to us does not usually offer its DE, let alone whether or not it is of a decidedly rapid, medium, or slow set variety. When possible I use a ‘yellow’ or ‘ruban jaune’ pectin for pâte de fruit, designed to offer a fairly slow set with a solids content greater than 75% and within a range of 3.2-3.5 pH. Otherwise, what often is labeled ‘apple pectin’ does the job well enough. At the end of the day, choosing one pectin over another comes down to what is available, and perhaps a bit of trial and error to find what works best on a consistent basis.

Dosage is most often determined by the amount of naturally occurring pectin present in the base fruit puree. Added pectin will generally be about 1% (or slightly higher) of the initial weight of ingredients – more for fruits that contain little usable pectin of their own. This is why it is common to fortify pectin-weak fruits with additional purees such as apricot, apple, or pear to offset this deficiency; indeed I come across many pâte de fruit recipes that call for apple juice as a consistent practice, no matter the base fruit being used.

Pectin is prone to clumping upon contact with liquid, so it’s vital to dry blend the pectin with some of the sugar used – typically about 5 times the weight of the pectin – to thoroughly disperse. I also like to heat the base puree to at least 40˚C/104˚F before adding the pectin, and most importantly, maintaining a temperature above 60˚C/140˚F during the addition the remaining sugars by adding them gradually – the most common fail with pâte de fruit occurs when all of the sugars are added at once, dropping the temperature and resulting in a weak final set.

 Balsamic Pate de Fruit


Sucrose is overwhelmingly the bulk sweetener of choice, with the addition of some glucose to both inhibit crystallization and slightly reduce overall sweetness. I occasionally see invert sugar used in small amounts as well. Added sugars will also fluctuate slightly based upon the amount of sugar supplied by the fruit itself. Part of what makes pâte de fruit boring is that because of the need for high solids content (sugar), most of the potential flavor is obscured by sweetness. To this end, I’ve been playing around with sugar alcohols and even low DE maltodextrin (here, meaning ‘dextrose equivalence’) to reduce sweetness. With the end goal of perhaps approaching what one might even call a ‘savory’ pâte de fruit. There may be limits to how much sucrose can be replaced with other sweeteners, but I hope to discover just what those limits are. Below a general reference chart for sweetener composition and properties:



The addition of acid is also critical in dropping the pH, allowing for the pectin to reach its final setting conditions. We most often see citric or tartaric acid used in pâte de fruit, but malic acid also works just fine. Something as simple as lemon juice will work, but is perhaps difficult to control. Given the circumstances I do not personally feel that one can taste any difference when one acid is used in place of another. I usually do not find it a problem to add the acid in dry form directly to the cooked mass, but it is far safer to dissolve the acid in an equal weight of water to help disperse it into the molten mix, especially when working with a rapid set pectin.


As previously mentioned, maintaining high temperatures during the addition of the sugars is crucial, but so is cooking to the correct temperature. Final cooking temperature is a fairly standard 106-107˚C/222-224˚F. This temperature also correlates to the final solids content of a water/sugar mixture – boiling point elevation as a function of dissolved solids is what we call a colligative property (as is the depression of the freezing point). Part of the reason it seems that confectioners use temperature to calculate solids content is in part out of ease, but also because it is an imperfect science:  it is difficult to precisely measure solids/water content under the extreme conditions of heat (see the Hartel article below).

Values vary widely by source, but the chart below begins to offer a rough idea on the correlation between temperature and the solution of sucrose in water (alongside the old school visual test method).  Of course in the case of pâte de fruit, the addition of other sugars such as glucose would cause these figures to change ever so slightly:

Sucrose Solution Temperature and Solids Content

The above figures of course hold up under normal atmospheric pressure, but I’ve been recently turned on to the possibilities of vacuum cooking, which allows for moisture reduction at far lower temperatures. For example, if cooking a sugar syrup to 150˚C/300˚F to achieve a solids content in the neighborhood of 2-3%, pulling a vacuum to 25inHg (inches of mercury) would give the same results cooked to the far lower temperature closer to 93˚C/200˚F. Not only is this far lower than the caramelization point, but it is also below the threshold where Maillard reactions begin to take place – allowing for ‘white’ caramels.  So what would this mean if applied to pâte de fruit? I’m not sure yet, but the final cooking temperature could be reduced to 81˚C/178˚F (from 106˚C/223˚F using an open system). Perhaps when cooked under a vacuum it would retain better color and flavor and result in less inversion of the sucrose and less degradation of the pectin. This small batch vacuum cooker below is currently on my wish list should be arriving to a shiny new ICE confectionery lab in a few months:

Vacuum Cooker

Back to the cooking process before moving on, and a point that I cannot stress enough: outside of inaccurate scaling or undercooking, from my observation the most common cause of improperly set pâte de fruit is the sudden drop in temperature that occurs when the sugars are added too quickly. Slow and steady wins the race. And of course once the acid is added setting begins at a rapid pace, so that’s when one must act quickly and decisively. Recently wrapping my head around the idea of how to use the pasteurizer portion of our Bravo Trittico machine to cook pâte de fruit, it became more a question of how to extract it. The solution was a simple one: cook the mixture to the necessary temperature and extract from the machine (while still quite fluid) into a bowl before adding the acid.

The typical finishing of pâte de fruit is sort of like adding insult to injury – sugar on top of sugar. Whenever practical, I like to blend 1-2% of an acid into the sucrose used for dusting, or a dry spice, or any dried and powdered flavoring. Such a simple detail of added flavor can do a lot to elevate the ho-hum status of pâte de fruit  to above-average.

So, with better insight into the mechanics of how pâte de fruit really works, what do we do with it? Well, for me, even the most boring tasks become exciting again once I better grasp the science, so there’s that. Though I had already begun playing around with savory flavors and alternative sugars, I was inspired to go back to square one and make pâte de fruit with nothing but water; a ‘pure’ starting point from which to remove conventional associations and move in new directions. If through baby steps I arrive today at a less sweet carrot, red pepper, or balsamic vinegar pâte de fruit, perhaps tomorrow might bring jellies flavored with mushroom, artichoke, white bean, etc. I realize that there are several methods beyond pectin that we could use to gel a savory liquid. But none would give me just that texture a conventional pâte de fruit provides. In the end maybe these flavors and that texture just aren’t compatible (or palatable).

If the journey is in fact more important than the destination, I’m more than happy to take this ride to see what happens along the way.

 Water Pate de Fruit

Below, the formula for a recent favorite of mine, a duo of grapefruit and Campari:

Grapefruit-Campari Pâte de Fruit – Opusculum


An interesting (and free) paper by confections experts Richard Hartel, Roja Ergun, and Sarah Vogel that in part addresses the difficulty in measuring solids content of high temperature syrups:

Phase-State Transitions of Confectionery… Hartel, 2010

A great way to begin understanding pâte de fruit formulas is to reference the parametric charts published by companies like Boiron; their latest edition below:

Boiron Chocolate and Confections 2013


Chocolates and Confections: Formula, Theory, and Technique for the Artisan Confectioner, 2012

by Peter P. Greweling

Science of Sugar Confectionery, 2000 by William P. Edwards

Alternative Sweeteners., 2001 by Lyn O’Brien-Nabors.2001

And did you know that Kitchen Arts and Letters in NYC now sells books online? Awesome!


Milk Chocolate Praline Croustade

The light and flaky pâte croustade long ago entered my repertoire as a go-to dough preparation for savory tarts. Beyond that, I didn’t think of it much until it resurfaced as a solution to a three-fold ‘problem’, a moment of inspiration:

1. In general, I’m always thinking about reducing sugar where appropriate – not out of altruistic concerns about health and wellness, but rather in service of flavor. While sugar can be a great vehicle for flavor, sweetness can also obscure it.

2. More recently, I’ve also been thinking a lot about ‘reconstruction’. The dominant style of plated desserts these days still seems stuck in the rut of ‘piles of stuff’. It’s a pleasant aesthetic, but I’m starting to feel that it’s too easy. Where can we take it next? How can we put things back together in a way that ‘eats’ better, but also challenges the chef to move beyond this stylistic crutch? How do we clean up our plates and still manage to arrange these flavors and textures in an interesting way?

3. One of my favorite pastry chefs to keep tabs on is Christophe Michalak – like Pierre Hermé, he represents a great balance of contemporary and classic with work that is clean, precise, and to-the-point. About a year back I saw something he referred to as ‘tartelette slim‘ – a seemingly impossible, near-cylindrical tart shell. Here was an idea that both changed the notion of what form a ‘tart’ could be, but also presented an interesting vehicle to re-assemble components in a more ‘constructed’ way.

I never came across a formula for the specific dough Michalak used in his version (to be honest, I didn’t look all that hard), but I knew the challenges would be in finding a way to form the cylindrical shape, and then adapting a typical pâte sucrée/brisée  that would hold its shape in the baking process. I immediately ruled out anything with a substantial amount of sucrose – the appealing tenderness it provides in every other dough would be a serious structural disadvantage here. And then I remembered that  pâte croustade  – it’s capable of being rolled extremely thin, it can be formed into various shapes without shrinking or falling apart, and it’s lack of sweetness also appealed to my flavor forward sensibilities.

Croustade 1

The dough is a bit strange – whereas most flaky doughs require cold temperatures to keep small discrete pockets of fat from coalescing, this dough is assembled with melted butter and warm water. Indeed, the end result of mixing often resembles a broken – if not outright greasy – mass. But it is that very ‘broken quality that creates a fine flaky texture after chilling and sheeting. After a couple of years of not making it, the croustade has re-emerged as one of my favorite doughs to work with.

We used to fashion tiny croustade ‘tacos’ as a canapé at Le Bernardin – draped over stainless steel tubes (cannoli forms, actually) and baked – so the problem of form was also solved. When looking to achieve such geometric precision, the smallest things can throw that precision off; I’ve found that cold working temperature and ample resting between the steps of sheeting, cutting, and baking to be crucial. In addition to exact measurements, another challenge was keeping the cylinders stationary during baking; a tight wrap in foil not only holds the dough in place, but foil cut to a specific length can allow for the slack to rest under the narrow gap between the edges of the dough – this prevents the foil from unraveling and the tube from rolling around in general.

Croustade 2

The first iteration of this ‘tart’ appeared with my collaboration in the kitchen at L2O, with wintertime flavors of citrus and fennel (below). A more recent version forms a vehicle for milk chocolate and hazelnut (above).

Below, the recipe for the base dough, and the general dimensions necessary to produce the thin 3/4 cylinder shells:

Pâte Croustade – Opusculum

Courtesy of L2O

Image Courtesy of L2O



Gelatin – few products are as ubiquitous in the kitchen yet so wrought with confusion and potential for misuse. Whether discussing gelatin in front of students who have had little or no exposure to it, or with seasoned pros who use it every day, I’ve come to realize that a lot of cooks see gelatin – as I did for many years – through a lens of  vague, misleading, or just-plain-wrong information, however well-intended its source. Gelatin serves as a prime example of how subtle nuance with regard to ingredients and how we handle them can either lock us into a state of perpetual frustration or liberate us with better control, refinement, and creativity.  What I will attempt to do here is quickly cover some of the basics gleaned from practical experience, some experimental testing, and a lot of technical reading. As with many aspects of cooking, there aren’t always hard and fast rules or optimal ways of doing things; there is always room here for personal preference. Hopefully with a bit of guided navigation through the complex world of gelatin others can benefit as I have, with a better informed launching point from which to make those choices and explore new territory…


1. The word gelatin has its roots in the Latin verb gelāre, to freeze, but  eventually began to refer to any general transformation of a liquid into a solid – gelāta. It’s easy to see how it evolved over the years into terms we use today, such as congeal, heladogelato, and specifically gelée and jelly. It’s believed that the earliest ancient forms of gelatin were not put to use in food per se, but rather as crude glues and adhesives. And though its culinary value became better understood and exploited over the centuries, the commercial gelatin we take for granted today did not begin to appear until the early 19th century. Indeed, most gelled dishes up to that point – savory and sweet – began with calves feet, much like this recipe for blancmange from the classic  1832 cookbook by Eliza Leslie, Seventy-Five Receipts for Pastry, Cakes, and Sweetmeats:

Blancmange - Eliza Leslie, 1832

2. Gelatin is mostly protein, which is derived from collagen, which only comes from animal sources. It’s dry composition is roughly 87-92% protein, 7-12% water, and about 1% ash. The great majority of gelatin is produced from porcine (pork) or bovine (beef) sources – predominately from skins and hides, but also from bones (ossein). A lesser amount is derived from piscine sources (fish skin) – both cold and warm water species – and an even smaller quantity from avian (poultry) sources. Porcine gelatin still tends to dominate much of the worldwide production, especially in Europe, but elsewhere bovine gelatin is increasingly common; piscine and avian sources account for less than 2% of total production. The source of any given gelatin is important for those with dietary restrictions, but uncovering the source can be difficult. Package labeling can be frustratingly minimal, though a bit of internet searching can usually help. Lacking info directly from the manufacturer, we can make a few broad assumptions:

– If one sees ‘Kosher’ or ‘Halal’ labeling, one can assume it is of bovine origin (though I have seen some implausible internet rants that claim otherwise).

– If the gelatin is of European origin, there is good chance it is from a porcine source – strictly based on the numbers and the brands readily available to cooks.

– We can also use the two primary types of gelatin – A and B – in determining the source. Before the extraction process, all raw material for gelatin is treated with either an acid (Type A) or an alkaline (Type B). Most porcine sources undergo acid-treatment (A) and bovine sources with alkaline (B), though that isn’t absolute. Adding to the confusion, sometimes ‘Type A’ is used to imply ‘food grade’, but from a manufacturing perspective, both A and B types can be food grade. Labeling by type is not necessarily definitive of overall quality – beyond this initial acid/alkaline pre-treatment, the production of both types are generally the same, though the properties of Type A and B gelatin can differ slightly in the final product.

– Gelatin can be of mixed sources, as is the case with the widely available Knox brand – its parent company Kraft states that it is a blend of both porcine and/or bovine sources.

– For those looking to bypass mammalian sources altogether, gelatin from warm-water fish  – sometimes referred to as isinglass – can approximate the properties of bovine and porcine of similar strength, though its setting and melting temperatures are generally lower, by a difference of up to 5°C/9°F. Cold-water species of fish gelatins form very weak gels by comparison and on the whole are not considered commercially viable.

– Though perhaps obvious to most, anything labeled ‘vegetable gelatin’ is not made from collagen, but rather one or likely a blend of plant-based hydrocolloids aiming to replicate the unique properties of gelatin.

Grea tLakes Gelatin

3. The most common identifier of gelatin is by its ‘strength’ – a grading system known as the ‘Bloom Scale’. This rating is named for Oscar Bloom, inventor of the gelometer patented in 1925 – an instrument that measured the amount of force needed to deform a gel to a specified degree. Though adapted over time, the same basic mechanism is still used as an industry standard in measuring gel strength: a 6.67% solution of gelatin in water is heated and set at specific temperatures for specific periods of time. The resulting gel is pressed with a cylindrical plunger to a depth of 4mm – and that force, measured in grams, is its Bloom rating. Thus, a gelatin whose tested gel requires 200 grams of force has a strength of 200 Bloom. All commercial gelatin lies within a range of 50 to 300 Bloom, however the range accessible to most cooks is generally narrowed from 125 to 250.

A higher Bloom value generally represents a stronger gel, a higher gelling temperature, a lighter color, and neutral odor/taste – but Bloom strength is not necessarily a complete means of determining a gelatin’s specific properties or purity. But overall quality is also a function of the extraction stage from which the the base gelatin is produced – early extractions of the raw material tend to have higher Bloom ratings than subsequent extractions – which number from three to six – requiring increasing amounts of heat and water to access the remaining protein. I like to think of it like olive oil – the first pressings at lowest temperatures offer the highest quality.

Regrettably, as with animal source, most retail gelatin labeling omits this very important Bloom number.  In its place we see instead a grading system using the names of precious metals (bronze, silver, gold, and platinum) in designating gelatin sheets of different strengths (more on this below). The aforementioned Knox powder does not list a Bloom number on its packaging, however my inquiry to Kraft revealed a ‘target’ Bloom strength of 235 (most sources list it as 225 – close enough – so I will stick to that ’rounded’ figure).

Before moving further – the use of the word ‘bloom’ can be confusing as it is used to refer to two things – a gelatin’s strength, and the initial process of hydrating or soaking in water.

4. Although sheet gelatin of different grades have different weights and different bloom strengths, they are theoretically interchangeable.

To quote Schrieber and Gareis (Gelatine Handbook):

In the case of leaf gelatine, the leaf thickness and hence the weight of the individual leaves is set according to the type of gelatine being processed. Thus, any particular leaf of gelatine dissolved in a given amount of fluid results in the same gelling power, independent of whether ‘‘high-Bloom’’ or ‘‘low-Bloom’’ gelatine is used in the leaf manufacturing process. This principle is valid for both major worldwide manufacturers and hence for all the brands available on the market.

The table below offers an average Bloom strength for each grade of sheet gelatin, its weight, and number of sheets in 1000g, alongside a common 225 Bloom powdered gelatin:

Gelatin Grades

The grade differentiates the gelling ability on a per gram basis, so bronze gelatin actually weighs more than gold yet achieves the same “set”.  For example, in a recipe calling for 2 sheets of silver gelatin, the same gel strength should be achieved by substituting 2 sheets of either bronze or gold grades. Recipes that call for a weight of sheet gelatin without also stating the Bloom strength can cause obvious confusion, as the weight of each sheet changes with an increase in Bloom strength. In such situations, while we might assume a silver grade will produce the desired result, it is often best to prepare small test batches in order to fine tune the formula.

            Converting between sheets and powdered gelatin can be more problematic. Several sources cite the powdered gelatin equivalent of one sheet to be approximately 1 teaspoon; however, volume measurements are not as consistently accurate as measuring by weight, and as we can assume from the table above, 2.3g of 225 Bloom powdered gelatin would most certainly result in a stronger gel than one sheet or 2g of 200 Bloom gold grade gelatin.

            The following conversion formula allows us to use the figures above to more accurately move back and forth between different gelatin grades, or between sheet and powdered gelatin:

The weight of the known gelatin x square root (known gelatin bloom strength/unknown gelatin bloom strength) = weight of unknown gelatin.


            For example, a formula calls for 2 sheets of silver grade gelatin (160 Bloom). At 2.5g per sheet, the total weight of the known gelatin is 5g. To substitute powdered gelatin in place of the silver grade sheets:

Weight of known gelatin = 5g

Bloom strength of known gelatin = 160 (silver)

Bloom strength of unknown gelatin weight = 225 (powdered)

160 divided by 225 = 0.71

The square root of 0.71 = 0.84

Multiplying our known weight of silver gelatin by a factor of 0.84 will give us our equivalent weight of powdered gelatin.

5 x 0.84 = 4.2

Thus, 4.2g of 225 Bloom powdered gelatin has the same gelling strength as 2 sheets or 5g of 160 Bloom silver gelatin, or for every sheet of silver, substitute 2.1g of powdered gelatin.

            Or you can save time and use the table below, which calculates this basic formula for substituting between different Bloom strengths by weight from commonly used 140 to 225 Bloom:

Gelatin Conversion

So, given all these options, which gelatin is ‘better’? Some feel that powder is somehow less ‘pure’ than sheets (in reality, all sheets are made from powdered gelatin). Some prefer the convenience of counting sheets over weighing of powder. Some use a specific gelatin for different applications. Some choose based on price (though for a true comparison, it’s best to compare the price per sheet between, say, silver and gold – a 1K box of silver may be cheaper overall, but remember that there are a hundred more sheets in that box of gold). And then some are limited by what is available from their local purveyors, though I’d argue that should no longer be an obstacle to sourcing anything in the age of internet.

5. Gelatin will absorb a fairly constant amount of water – about five times its own weight. So why do we tend to use large amounts of excess water for hydration? For example, one sheet of silver gelatin that weighs 2.5g will absorb about 10 to 12g of water. Because it can be difficult for such a small amount of water to cover the surface area of one gelatin sheet, it makes sense to hydrate the gelatin in a larger amount of water to completely submerge the sheets, and then gently squeeze out any excess water. This practice is perfectly acceptable, though care must be taken in handling the gelatin so that it loses none of its gelling power in the process; common loss can occur when the gelatin breaks into smaller pieces, is hydrated in warm water, or when it is squeezed with warm hands.


Ideally gelatin sheets are bloomed in just enough ice water to cover, followed by draining through a sieve with gentle pressure to remove excess water. Excess water that is added to a recipe along with the gelatin will in effect dilute its gelling power. As a measure towards consistency, I have always preferred hydrate gelatin in a measured amount of water (about 5x)  for 10-15 minutes and simply add any excess with the gelatin, having accounted for the water within the recipe. A third method employed by some chefs (and one that I’m playing around with) is to prepare a ‘gelatin mass’ in bulk by hydrating a large amount of gelatin in a precise measurement of water and heating the whole until the gelatin is dissolved; this is further allowed to set, and these chefs’ recipes will then call for a measured weight of the set gelatin mass rather than a number of sheets. Converting recipes that use gelatin mass – and I see more and more of them – can be a pain if the variable includes gelatin of unknown Bloom strength.

Gelatin Mass

Powdered gelatin is hydrated in a similar manner; the gelatin is sprinkled over a small bowl containing 5-10 times its weight of cold water. Sprinkling the gelatin on top of the water will allow it to disperse and absorb water at an even rate – pouring the water over the gelatin will often result in lumps of dry granules that lose access to the water because the surrounding gelatin swells in size as water is absorbed. It is common practice to use water when hydrating either powdered or sheet gelatin, but most liquids are acceptable, as long as they are cold. When converting between sheets and powder, it is important to know that gelatin absorbs about five times its weight in water. That is, 5g of gelatin will absorb 25g of water. While this water is always listed in formulas using powdered gelatin, it is not listed in formulas where sheets are placed in excess water. This difference in water should be considered when converting between sheets and powder.

A couple more general reference points on the behavior of gelatin gels:

– In typical dosages, gelatin begins setting in the range of 15-20°C/60-70°F and melts at just about body temperature, 35°C/95°F. This narrow gap between setting and melting is referred to as hysteresis. Though visible signs of gelling occur rather quickly under refrigeration, it’s important to realize that the complete gel structure  and thus full strength may take up to 12 hours to set.

– Excess heat over extended periods of time can damage gelatin and produce a weaker gel; ideally, gelatin should never exceed temperatures 60°C/140°F.

6. Consider how other ingredients affect gel properties. Let’s take a step back to remember that gelatin is a hydrocolloid – it gels water. Although we set complex mixtures that also contain fats, sugars, and other ingredients, we should be mindful in determining gelatin dosage based on water content. For example, In setting a creme anglaise into a simple cremeux, we should look first at the weight of water, excluding the fat and nonfat solids; further fine-tuning of the gelatin ratio can then be done as one assess how these other factors may affect the gel’s properties, if at all.

A very worthwhile exercise is to set a range of plain water gels of various gelatin types and concentrations –  it’s a valuable exercise to try with all hydrocolloids – one can learn a lot about the subtle differences in products before other factors like fat, solutes, and pH are introduced…

Water Gels

Factors that can affect the properties of gelatin:

– High levels of dissolved sugars or salt can slow the hydration and dissolution of gelatin; because they bind water, the gelatin must compete for available water. However, sucrose and sugar alcohols like sorbitol help stabilize gelatin gels, increasing both the setting time and the melting temperature.

– Fats can soften or ‘plasticize’ gels, which may require a higher dosage of gelatin. On the other hand, similar items that contain solid fats like cocoa butter may require a lower dosage of gelatin, as those fats provide structure of their own.

– Gelatin is stable within a pH range of 5-9 – increasing the acidity can weaken gel strength. For gels with a lower pH, one can simply add more gelatin to compensate, or use a buffering salt such as sodium citrate at roughly 1% of the weight of gelatin.

– Some fruits – such as pineapple, papaya, kiwi, and fig – contain proteolytic enzymes (papain, bromelain) that inhibit proper gelling; using these fruits with gelatin requires prior heating to destroy these enzymes.

– Alcohol will also inhibit gel formation and requires higher concentrations of gelatin to achieve gels comparable to water alone. An intriguing formula and good starting point, for edible cocktails proposed by Martin Lersch of gelatin to add = (% alcohol in final mix x 0.1) + 2

– The synergistic effects of gelatin combined with other hydrocolloids are many; perhaps the best overview of gelling possibilities can be found in Volume Four of Modernist Cuisine.

I think that sets us up for some more in-depth topics in the future, as this only begins to scratch at the surface of gelatin. Below, of many variations of gummy candy that I’ve tried, my current favorite – adapted from a formula taught by Jean-Marie Auboine:

Passion Fruit Gummy Candy – Opusculum

Some invaluable references below on the subject of gelatin:

Gelatine Handbook: Theory and Industrial Practice, by Reinhard Schrieber and Herbert Gareis 2007

Handbook of Food Proteins, edited by G.O. Phillips, P.A. Williams 2011

Handbook of Hydrocolloids, edited by G. O. Phillips and P. A. Williams 2000

Modernist Cuisine, by Nathan Myhrvold, Chris Young, and Maxime Bilet 2011

How Baking Works, 2nd Edition, by Paula Figoni 2010

Gelatin Handbook,Gelatin Manufacturers Institute of America 2012

Chef Steps

And did you know that Kitchen Arts and Letters in NYC now sells books online? Awesome!



Though perhaps impractical on a large scale, making butter is a fun process through which we can better understand the wonders of milk. In speaking of butter, it helps to place it context within a broad view of milk and its derivations. All mammals produce milk whose purpose is to feed infant mammals of that species. The same is true of humans, although we didn’t always seem to be hard-wired for consuming milk as adults; in childhood our bodies are supposed to stop producing the enzyme needed to break down lactose – the sugar found in milk. Indeed it is lactose that the dairy intolerant have problems with. Depending on whom you ask, it is believed that an ancient genetic mutation or biological imperative (perhaps vitamin deficiencies due to human migration further away from the equator) led to a growing toleration of lactose well into adulthood, leading to widespread fresh milk consumption in the neighborhood of about 7,500 years ago, if not earlier.

Long before, however, domestication of animals for meat as well as dairy had led to the development of fermented milk products of many kinds – from yogurt to hard cheeses – which were not only easily digestible (bacterial fermentation ‘eats up’ most of the lactose), but nutritious, portable, and stable for longer periods of time. Think about it: fresh milk is highly perishable and made up of 88% water; if you’re a nomadic herder constantly on the move, it makes sense to remove some of that heavy water and concentrate milk’s energy and nutrients, in turn increasing its shelf-life, and in many cases expanding its flavor profile.

Below, an approximate composition for cow’s milk and another that compares it against other mammalian species:

 Composition of Cow's Milk

 Comparison of Mammalian Milk Composition

Not only is the composition important in determining the properties of milk, but looking at the physical structure is also helpful. Milk in its raw form is relatively unstable, in that the milk fat has a tendency to flocculate – to cluster together and rise to the surface. Because of milk fat’s natural tendency to form this ‘cream layer’, most commercial milk products are homogenized in order to separate the fat into small, stable droplets evenly dispersed throughout the product. In a process that has remained fundamentally unchanged for over a century, the milk is pumped through a screen of tiny openings at high pressure and velocity, effectively breaking the individual fat droplets into smaller ones. Homogenization is typically carried out immediately after pasteurization – the process of heat plus time used to destroy harmful microbes.

The structure of milk can be expressed in three different ways:

– An oil-in-water emulsion with the fat globules dispersed in the continuous water phase. An emulsion is an evenly dispersed mixture of two substances that do not readily mix – in this case the milk fat and water. Homogenization of the milk creates this emulsion.

– A colloidal suspension of protein particles; a suspension involves solid particles that are, in effect, ‘floating’ within in the milk.

– A solution of lactose, soluble proteins, minerals, vitamins other components; a solution refers to all of the substances dissolved within the milk.

World Science Festival Butter Lab, 2014

Generally speaking, both the price and perceived level of quality of any dairy product is proportional to the percentage of milk fat it contains. The fat component in milk adds richness, aroma, and flavor to pastry preparations, and can contribute a smooth texture and creamy body, adding a sensation of moisture and lubrication to the palate as it is consumed. In baked goods, milk fat shortens, or tenderizes the finished product. The moisture in butter can create leavening as the tiny droplets turns to steam in the oven. Milk fat can be provided by whole milk, but also from concentrated forms, such as heavy cream and butter. With regard to whipped cream and ice cream, it is the fat that provides their primary solid structure, a ‘scaffolding’ of sorts surrounding air that has been whipped into it.

Below, a table comparing the composition of the most commonly used products derived from milk:

Dairy Composition

On our way to butter, we must also discuss cream – that portion of milk that naturally clusters and rises before homogenization. As mentioned, traditionally this is referred to as the ‘cream layer’ skimmed off of fresh milk. Today, a range of high fat products are produced, tailored for specific uses. Like milk, creams are categorized according to their fat percentage. Below, a table that compares common consumer and commercial products by composition:

Comparison of Cream Product Composition 


Heavy cream or heavy whipping cream contains between 35% and 40% fat and is most commonly used in those two forms, providing flavor and texture to pastry preparations. The high fat content provides desirable boiling and whipping properties; the reason why boiled milk forms a ‘skin’ of protein and cream does not is because the extra fat in cream buffers the casein proteins, preventing coagulation when the cream is boiled (there’s also, of course, less protein cream to begin with). The fat also stabilizes whipped cream by forming a continuous, partially coalesced network around air that is whipped in. Because even a 5% difference in fat can adversely affect some delicate preparations, recognizing fat content and being able to adjust a formula is important.

An intermediate along the spectrum of cream and butter is clotted cream – also known as Devon or Cornish cream – a unique high fat product. Clotted cream is cream which has had a portion of its water cooked off, to produce a minimum fat content of 55%, though it can be higher.

           Whipping Cream

As with ice cream, we can better understand butter by first looking at liquid cream’s ability to whip up into a solid foam – an everyday example of shear-thickening. Technically speaking, all foams can be described as colloidal, two-phase systems in which air forms the dispersed phase and water forms the surface phase. In the case of whipped cream, the surface phase contains casein proteins that complex with the partially coalesced fat -destabilized by the physical force of whipping –  forming that ‘scaffolding’ surrounding the air cells that are incorporated during the whipping process.

The two primary factors responsible for cream’s ability to whip are fat content and temperature.A minimum 25% is necessary to create a solid structure. In order for the fat to properly adhere or coalesce, the milk fat within the cream must also be partially solid, or crystalline. Success also depends on a minimum of 5˚C/40˚F. Half and half, for example will never whip up to a solid because it lacks sufficient fat. Likewise, cream that is room temperature or warmer will fail to form a stable structure because a large portion of its milk fat will have begun to liquefy, simply forming larger fat globules rather than a network of small globules that retain some of their own identity.

Counter-intuitively perhaps, whipping cream to ‘stiff’ or firm’ peaks does not necessarily provide for maximum stability and volume of whipped cream. A firm textured whipped cream is a result of too much coalesced milk fat, which will eventually ‘squeeze out’ some of the air and water phases; an unstable, over-whipped cream will likely have a slightly grainy texture, will ‘weep’ water over time, and will be difficult to incorporate into another mixture, such as when using cream to lighten a mousse. Fat coalescence in whipped cream is irreversible, meaning that once adhered, the milk fat cannot be separated without heating and chilling the cream to liquefy and recrystallize the fat. Thus, over-whipped cream can rarely be ‘rescued’; it is often better to use the over-whipped cream for another use and start over, whipping the cream to a ‘soft’ peak for maximum volume and stability.

As noted above, cream can be described as an oil-in-water emulsion with milk fat globules dispersed in a continuous water phase. In whipped cream, that milk fat has partially coalesced to form a semi-solid structure. When the cream is whipped or beaten further, those milk fat globules continue to coalesce – they irreversibly increase in size, expelling much of the air and water phase to eventually form a water-in-oil emulsion where the remaining water is dispersed in a continuous fat phase. This simple inversion of the original cream’s original structure is how butter is produced.

Below, various stages of whipped cream stable soft peak to finished butter.

Whipped Cream, Soft Peak

Whipped Cream, Firm Peak

Whipped Cream, Stiff Peak

Whipped Cream, Broken

Whipped Cream - Coalescence

Early Stage of Butter

Final Stage of Butter

Drained Butter

As more and more of the fat becomes tightly packed together, it is further separated from much the cream’s original liquid, referred to as buttermilk. This buttermilk (a little less than half the original weight of the cream) contains only a trace of the cream’s original fat, but a majority of its nonfat solids. The buttermilk we produce as a byproduct of making butter ourselves is a bit different than the commercial buttermilk available. Although its name survives from the traditional practice of allowing ‘true’ buttermilk to spontaneously ferment, modern commercial buttermilk is produced by inoculating low-fat or skim milk with a bacterial culture (Lactobacillus bulgaricus or Lactococcus lactis), sometimes supplemented with additional nonfat milk solids to create a thicker texture (as in all cultured dairy products, the acids produced by the bacteria denature the proteins, which gives the product body). Even though it differs from the commercial version, the possible uses of fresh buttermilk are many; I recently made a fairly decent ricotta from it, as there remain a substantial amount of curd-forming proteins.

To salt, or not to salt: up to 1.5% salt can be added to butter (at the final kneading stages, otherwise it will dissolve and drain away with the buttermilk) in part for flavor, but traditionally relied upon as a preservative. It is generally preferable to use an unsalted butter, which allows the chef to better control the salt content of a preparation as well as to ensure a fresher product. If I intend to use salt with butter for bread service, I actually prefer to serve it with a sprinkle of coarse Maldon salt at the table.

Federal law requires that butter contain a minimum 80% milk fat. Most commercial butters are composed of roughly 82% fat, 16% water, and 2% nonfat milk solids. High fat butters, sometimes referred to as European-style, may have a fat content of 83% to 84%, and are desirable for their plasticity, especially in the preparation of laminated dough products. A simple means of determining the fat content is to cut a thin slice from the cold butter and try to bend or fold it – a low fat butter will break and feel brittle, whereas a high fat butter will maintain a greater degree of flexibility, even when cold.

A simple method for making cultured-style butter by hand from heavy cream and crème fraiche is given below. The butter-making process is relatively simple; the cream and crème fraiche are whipped until all of the fat has coalesced. The resulting liquid, or butter milk is drained off and further kneaded from the fat. This method can produce a butter of higher fat content more complex flavor than most commercially available butter:

Cultured Butter – Opusculum

Finished Butter, Washed and Kneaded

A few in-depth resources on milk:

On Food and Cooking, by Harold McGee

Milk and Milk Products: Technology, chemistry and microbiology, by A. Varnam and Jane Sutherland

Dairy Science and Technology, by Pieter Walstra, Jan Wouters, and Tom Geurts

University of Guelph (Ontario), Dairy Education Series

Also of interest, from my dairy and ice cream guru, Cesar Vega:

The Kitchen as Laboratory: Reflections on the Science of Food and Cooking, by César Vega, Job Ubbink and Erik van der Linden

And did you know that Kitchen Arts and Letters in NYC now sells online? Awesome!

Crunchy Choux

Crunchy Choux

I should probably give pâte a choux a rest – I’ve been off and on obsessed with it for over two years – but I feel there is still much more to harness from this understated preparation and more to refine.  When it’s done well, there are few better pastry-based vehicles. But therein lies the problem: often viewed as ‘just a vehicle’ for whatever is inside of it, we don’t always give it the attention it deserves.

It’s true that the basic ratio of ingredients doesn’t vary all that much, and hasn’t really strayed from that of Carême, who is generally regarded as the author of the modern recipe use today. I went back to the only choux recipe I have in my files from Carême – an 1834 English translation of his Le Pâtissier Royal Parisien – though the quantities are surprisingly imprecise, any novice pastry cook would immediately recognize the general proportions and method:

Royal Parisian Pastrycook and Confectioner, Carême 1834

'Almond Choux' - Carême

Because the standard formula of liquid, fat, flour, and eggs is fairly constant, I get the impression that few chefs ever adapt beyond the first version of the recipe they acquire as a student or young cook. That’s a shame, because there is a lot more about the method to understand and subtle tweaks to fine tune in order to raise the bar for choux… Small adjustments in milk fat and nonfat solids can help determine texture, flavor, and color. Sugars – and sometimes salt – are omitted outright. From cake to bread flour, small adjustments in the overall protein content can affect the final structure and exterior appearance. Time and temperature of the preparation matter at each step – how long to cook the roux, at what temperature should the eggs be added…

I can’t say that I’ve worked through every variation of ingredient and procedure, but I’ve continued to improve my choux slowly but surely over time. A huge revelation came with understanding the best technique for applying a crunchy exterior to the finished piece: a sablée of sorts, but more so in proportions similar to a streusel (roughly equal parts of fat, sugar, and flour) that is ‘tender’ enough to expand with the choux, where a conventional dough would set too quickly and restrict the ‘puff’. In fact, perhaps counter-intuitively at first glance, the sablée- draped choux actually rises up to twice as much as an uncovered one – the sablée slows the drying and setting of the choux surface, allowing it to expand that much more.

Pistachio Choux

Always looking to refine and streamline production along with quality, the crunchy choux also offered a unique efficiency – in part thanks to a push from my friends at the PreGel training center, run by Frederic Monti. Understanding that choux can be frozen before baking as well as after, the choux is piped into silicon molds (perfectly consistent size and shape) and frozen; the sablée is sheeted, cut into discs and draped over the frozen choux domes. The choux is then baked comme d’habitude – high heat to begin for maximum oven spring, with a gradual decrease in temperature to dry out the interior.

Choux Method 1  (Photo Credit - Lauren DeFilippo)

Choux Method 2  (Photo Credit - Lauren DeFilippo)

Choux Method 3  (Photo Credit - Lauren DeFilippo)

Below, the full recipe for the choux and  sablée, with one of my favorite fillings, a chartreuse mousseline:

Crunchy Choux – Opusculum

Finding ultra-technical, underlying science of choux is not as easy as one would think. A few of my favorite general baking science references are:

How Baking Works: Exploring the Fundamentals of Baking Science, by Paula Figoni

Understanding Baking: The Art and Science of Baking,
by Joseph Amendola, Nicole Rees, Donald E. Lundberg

Baking Problems Solved, Stanley Cauvain and Linda Young

Eclairs!, by Christophe Adam

And did you know that Kitchen Arts and Letters in NYC now sells online? Awesome!

Crunchy Choux (photo credit Stephen Kenny)