paulraphael

I've started work on a whole series of ice cream posts on the underbelly blog. Just the beginnings right now, but over the next few weeks I'd like to get into the nitty gritty of recipe design, including proportions of sugars and designing a stabilizer blend. If you catch any mistakes or if anything's unclear, I welcome your feedback.

In my experience, if you oil a pan before preheating it, you'll fill the house with smoke and end end up with a coating of polymerized oil ... which will encourage food to stick and be very hard to get off.

All the clad copper pans I've seen use three rivets. Not that two wouldn't be just fine, but that's an oddball claim nevertheless. I don't know exactly how Falk phrases their preheating recommendations. For any kind of high heat searing you'll burn the oil if you don't preheat the pan dry (then add the oil, and then add the meat quickly. You've got just a few seconds before the oil smokes and breaks down.) This is just standard searing technique that you'll see used in any restaurant. The very high heat is especially important if the meat is thin, or if you've cooked it sous-vide and don't want to create a gradient. I also don't know the maximum temperature that's safe for clad copper pans, but I've used them in a 550°F oven, and know that they can take the full output of a 17,000 btu/hr range indefinitely. In either case you'll probably see some blue oxidative discoloring on the stainless, but this wipes off easily with some barkeeper's friend. I've never seen this material warp or delaminate. FWIW, I got rid of my tin-lined copper because the tin melted on the fry pans even under what I consider moderate heat (on a not very powerful range) and I wore through the tin in the saucepan with my whisk. I had no interest in retinning this stuff that couldn't handle the techniques I'd bought them for.

It's not such a big deal. I'm sure the consumer brands are fine for most purposes. If you're working with laboratory precision, and demand perfect repeatability from batch to batch, that's another story. I haven't used any commercial stabilizer blends. Every pastry chef I know uses them, so clearly they're good. I mix my own partly as a vestige of my darkroom days, when I learned that dependence on a commercial product sets you up for disappointment when that product gets discontinued or "improved." And also, I wanted to learn about these magic ingredients in a way that I couldn't if someone else was doing the experimenting and keeping the results secret. Because of this, I'm starting to feel like I'm in control enough to vary my blend for different flavors, which feels pretty cool.

Like most gums, guar and xanthan are synergistic, which means they're more powerful together than separate. In other words, you'll get a stronger effect from 1g of each mixed together than from 2g of either of them used separately. This is a feature, not a bug, as long as you're aware of it. It's one reason you usually see two or more gums used together in ice cream or anything else.

Absolutely. This is why chefs often specify a brand. So far I haven't noticed the differences, but I haven't been using gums for the kinds of things that would be sensitive to the most subtle changes. And I haven't done strict a/b tests on them. Right now I have a couple of brands of locust bean gum on the shelf, and in ice cream they seem about the same. For critical stuff, you can also probably expect better consistency from one batch to the next if you buy from the more technical companies, like CP Kelko and TIC gums. These guys publish elaborate specifications and quality guarantees. I bought my last bach of xanthan at the supermarket ... Bob's Red Mill. I would not expect this stuff to be so tightly controlled. Which isn't to say I've ever had a problem with it. I definitely appreciate being able to buy it right down the block.

I hear you. I've got the cool oven blues also. My dough is the best I've ever made, and I love the big hunk of steel, but have reached the limits of my dumb little oven. I'm tempted to burn a pile of wood in there just to see what happens ...

It works fine, but it's not the best. Other gums and combinations suppress ice crystals better, and offer better flexibility with textures. Manufacturers dislike xanthan because it's relatively expensive. A big problem I have with xanthan is that when used in combination with locust bean gum (which is the best among natural gums for suppressing ice crystals) it forms a gel. So after aging, the mix is as thick as pudding. You have to blend it to break the gel and get it thin enough to spin in an ice cream machine, and even then it's thicker than what's ideal.

Manufacturers will give a hydration temperature for their gums that may differ from the generic temperatures you'll read elsewhere. Gum molecules are heterogenous; from any source there will be different versions of the molecules that will hydrate at different temperatures. The gum companies make their formulas by carefully selecting the source and refining for a particular range of molecules for each product. But the hydration temperature they give is still most likely a recommendation. The locust bean gum I use is supposed to hydrate at 80°C. When I asked the manufacturer what would happen if I only cook it to 75°C°, the rep said that probably 90% of it would hydrate. Which is easy (and maybe even unnecessary) to compensate for. But probably at 50°C it would be completely inneffective.

putting it in the fridge for 20 minutes or so should work better than softening it on the counter. Less will melt on the outside. Starches and gums don't suppress the freezing point of the water or effect the hardening properties of the nut oils. These things have to be addressed in other ways.

It is, but there's a lot of room for confusion, because pastry people use glucose in different forms: glucose syrup, atomized glucose (basically the dried powder form of the syrup) and anhydrous glucose (the powder form that's usually sold as dextrose). These all behave differently and so are not interchangeable. I usually specify dextrose, because that's most likely to be understood as the anydrous powder that I prefer in ice cream.

Oat flour does taste like ... oats. Not powerfully so, but it's perhaps most useful in something that has a lot of flavor, or something where the oats are welcome.

In his Pizza Blog, Jeff Verasano says he has to use foil to keep his stone from getting too hot relative to the top of the oven. I think he says puts the foil loosely on the stone, shiny side-up, until a few minutes before popping in the pizza. He uses a stone instead of a steel, since he's rigged his oven to heat to 900. But I imagine a similar tactic would work with a steel at a lower temperature. One question: how much room is there for air flow around the edges of your steel? Is it possible that it's blocking the flow to the roof of the oven?

That's not as bad as some of KA's suggestions. It's about 1300 grams, 57% hydration (taking into account the malt powder). A lot of bagel dough is 50% hydration; a recipe this size would break any small planetary mixer. But at 57% it should be more pliable. A workout for the machine, but it probably shouldn't overheat it or break it. My guess is that you're kneading it much more than necessary. That recipe uses the old-school approach of throwing all the ingredients into the mixer and letting it work until the dough is perfectly formed. This requires many times more work than more modern approaches. If you're committed to using this method, I'd suggest cutting the recipe by about a third. If you want to make it this size, try the following: 1. Add an autolyse step. This means, when you first mix the ingredients together, leave out about a third of the flour. Mix it together with your flat beater, and only until it forms a loose dough. Cover it and let it sit for at least 20 minute. Preferably 40. This allows the starches and proteins in the dough to hydrate, and allows the gluten to start forming. These processes take time whether the dough is being mechanically mixed or not. There aren't advantages to working the dough before it's hydrated. So you and your mixer can just kick back and watch a movie. 2. Wet mix. Still with the flat beater, work that soupy dough on low speed for a bit. It will be very wet and, and will quickly get sticky. You don't want to mix too long, or the gluten will start to break down. I'm guessing that with this dough a minute will be enough. If you see it starting to look less elastic, stop. The dough is so wet that even with the flat beater, this is very easy on the mixer. But it works the bejeezus out of the flour proteins on the molecular level. A minute of this is like 10 or 20 minutes of dry kneading. 3. Switch to the dough Mix in the rest of the flour, only until it's well incorporated. You'll probably have to stop a couple of times to scrape the sides of the bowl. cover the bowl and let it sit again for another 20 minutes, so the new flour can hydrate. Share a cocktail with your mixer. 4. Knead with the dough hook. At this point, 90% of the work is done. Use speed 2, and watch closely. If the dough is really stiff, the bowl can be forced off the supports. Mix only until the dough looks smooth and homogenous. This should take at most 4 minutes. The mixer should barely get warm. 5. Lightly flour the counter and finish off with a couple of hand kneads, just to make sure the texture is right. This method adds time, but greatly reduces the work for you and the machine. It will generally give better results, because you're not overheating or overoxygenating the dough with too much mechanical mixing. It's not KA's fault that these stiff doughs are hard on the mixer. It's KA's fault for pretending otherwise. For reference, here's Hobart's capacity chart. The N50, on the far left, is their small countertop mixer. It costs over $2000, and is powerful enough to puree a Kitchenaid mixer. Notice that for "Heavy Bread Dough—55%" and for pasta dough, and pizza doughs of any hydration, they don't recommend the mixer at all. They're saying, if you want to do this tough stuff, get a 12 or 20 quart mixer and pay up! I think Hobart's being overly conservative. But then again, you never hear the complaints about Hobart that you do about KA. You're getting what you pay for from each company. But one of them promises too much. FWIW, you can overheat or break a Kenwood or Delonghi or Cuisinart as easily, if not more so. They're all wimps compared with the Hobart.

How big is your batch size (by weight)? (I'm asking because I suspect many—maybe most—of the problems with these mixers are the result of KA's insanely optimistic recommended capacities)

I’ve only experimented a bit with nut ice creams. My first attempt used nut butters (100% nuts, pureed in a high-powered blender—basically, nut paste without the added sugar). The trouble with this is that in order to get a vibrant nut flavor, you have to add so much nut oil and solids that the texture suffers. My ice cream was almost peanut-buttery. In the future I plan to use much less nut butter, and to make up the rest by infusing nuts into the milk (steeping crushed nuts for 30 minutes or so at around 85°C). Pistachio paste is the traditional method for gelato, but I question if it’s still the best. The old reason for using it is that it was once the only way to get the best pistachios, at least outside the Mediterranean and the Middle East. There were one or two brands coveted by every pastry chef. But I’ve read that these brands have gone downhill. Meanwhile, you can get better raw ingredients almost everywhere now. I’d explore getting your hands on the best possible nuts first, and then work on getting the best flavor extraction.

A portion of liquid sugars will help. So will higher protein flours that absorb more moisture. Bread flour can work, if your method minimizes gluten formation. I make chocolate chip cookies with a significant portion of oat flour, which is high-protein and non gluten-forming. Even without a liquid sugar component, and with tons of butter, these stay soft and chewy almost indefinitely.

This is right; the trouble with ice cream that lacks enough dissolved solids (with sugars being the most effective) is that at -12C, they're still hard as a rock. To get them soft enough, you have to warm them to a temperature that feels unnaturally warm for ice cream.

You're covering a lot of territory! You can think of the milk fat percentage as a mostly separate issue from the others. It doesn't have a significant effect on freezing point (am I right in understanding that this is the biggest problem you're tackling?) I'd start with a milk fat percentage that you like. More than 16% you risk an ice cream that mutes flavors and that leaves a greasy film your mouth; lower than 10% you risk one that has no creaminess and tastes more like sherbet. Context is also important here. Some pastry chefs go for a low fat level because the ice cream will be part of a big plated dessert served after a rich meal. Flavor also matters. Fat-soluble flavors (like spices) come through well in a high-fat bass. Water-soluble flavors (like fruits) don't. If you want to avoid milk sugars, I suppose you could start with lactose-reduced milk (sold for lactose-intolerant people) and experiment with building up the body with some whey and/or casein. I have no experience with these ingredients, so you'd have to do the research and experimenting. Luckily, you said you don't mind lots of egg yolks. Making a rich, French-style base with lots of custard takes care of a lot of the problems. You get stabilization, body, and smoothness. I don't see a need for added emulsifier if you're using 4 yolks or more per liter of mix. I use just 2 and don't add emulsifying ingredients. You shouldn't have to whip the cream before spinning the ice cream, unless your ice cream maker is specifically designed to make very low overrun ice cream. Whipping is one of the things that happens naturally during the process. Are you aging your mix overnight in the fridge before spinning it? It needs to chill at least a few hours for the fats to crystalize and become whipable. Stabilizers also improve whipability. You can buy an off-the-shelf ice cream stabilizer blend, or I can recommend a DIY recipe (requires a scale that reads to 0.01g and a blender).

I wasn't responding to you with my annoyance on demonizing sugars. You didn't make any sweeping statements about sugar's universal evil ... Note that erythritol is less likely to cause digestive problems, but the other sugar alcohols, and the polydextrose are all likely to cause such issues. It may be less likely than the others (not sure) but they all work in essentially the same way. I would expect people's tolerance for sugar alcohols to be quite variable. My advice (to anyone) is to go easy until you know how much you can tolerate. Just to avoid unpleasant surprises. Does milk help primary because of its sugars or because of the proteins? Both. The sugars probably suppress the freezing point more; the proteins offer other textural advantages. The proteins include molecules that work as emulsifiers (which in ice cream is really a de-emulsifier, but that's a different topic). This is especially helpful if you're using fewer egg yolks. Would egg white protein be expected to lower freezing point significantly? I've seen several recipes that contain egg whites, though a few egg whites doesn't give you all that much protein. I've never experimented with this, because all the literature says flat out that egg whites are bad for texture in ice cream Do you have any idea what goes wrong if you add too much whey (or egg white) protein to an ice cream? Too much whey results in a problem called "whey-off" or whey separation, by which the whey proteins coagulate and come out of solution, wrecking the texture. Whey is a common filler in economy ice cream ... it adds body in very low-fat ice creams that don't have other sources of milk solids. I have never used it, and don't know what advantages (if any) it would have over nonfat dry milk). Small quantities of carrageenan help prevent whey-off. Whey can also lead to off-flavors (people describe it as graham cracker flavor, or just a lack of freshness).

Stabilizers are great, but they don't have any effect at all on freezing point suppression. The quantities are too minute for this. All else being equal, they increase the whipability of the ice cream so you'll get more overrun. This makes the ice cream softer ... but you may not want more overrun. If you're going to use stabilizers for the more traditional purposes (reducing iciness, improving freezer life, and adjusting texture) xanthan works pretty well, but after a few years' experimentation I've found it to be one of the less effective stabilizers. Locust bean gum is just about the best of the commonly available natural ones. In combination with guar and lambda carrageenan it works wonderfully.

If you're making these with high proportions of sugar alcohols (xylitol, erythritol, etc.) you probably want to restrict the total amount of ice cream you eat, or at least experiment slowly with what you tolerate. We don't count these as sugars because we don't digest them, but our gut bacteria do, and this can cause gas and bloating and other kinds of gastric upset. Any dissolved solids in the ice cream will suppress the freezing point. The lower the molecular weight of substance in solution (the smaller the molecule), the greater the freezing point suppression. This is why monosaccharides like glucose work better than disaccharides like table sugar. And why salt works better than both (not recommended ...) Nonfat dry milk will help suppress freezing point and will also give your ice cream more body and better smoothness. Almost every pastry chef uses this. I don't like to use alcohol (unless it's a booze flavored ice cream, because it contributes nothing positive to texture besides freezing point suppression. Milk, of course, is full of sugars (particularly lactose). But stop demonizing sugar. It is not "bad" for you. That's b.s., unsupported by any science. We evolved to eat sugar. Just not nearly as much sugar as the modern diet often entails. If you want eat a sugar-restricted diet, great, but please don't spread internet health memes as if they're factual. We can all go to Facebook for that. If you don't tolerate foods well that have a high glycemic load, you can control the effects of this by what you eat with or before the sweet food. Your body only responds to the glycemic index of the mix of foods in your stomach. If you put prosciutto on a piece of french bed, the effective glycemic index drops from very high to very low.

I think they're mostly repeating lore when it comes to flavor. It's the exact same strain of saccharomyces cerevisiae regardless of what form it's in. I can imagine a flavor difference if you used a huge amount of cake yeast, and did a very short proof. But that would really just be about adding the flavor of lots of dead yeast organisms.

If you're smelling ammonia, that's from spoilage bacteria. Which (probably) won't hurt you. But the conditions that allowed the stinky organisms to grow also allow the infectious and toxic organisms to grow, and these are not polite enough to warn you with smells and flavors. So yeah, you dodged a bullet. I too question the idea of spending as little as possible on your protein. Stop and consider for a minute how it's possible to raise a mature pig for $1.63 a pound. Most farms can't grow lettuce for that little. Is this something you want to support? To eat? I'd rather see all of us eating better-raised animals, and doing it less often.

When I call yeast flavor a defect, I don't mean to do so authoritatively. One person's bug is another's feature. That just seems to be the convention in most of the writing I see on artisan bread. If you do like the flavor /aroma of yeast, my suggestion of getting it through brewer's yeast is for the purpose of getting more predictable and consistent results. I haven't tried this, but am imagining it would work. I agree with Jo that instant is superior, at least to 'active dry' yeast in every way. I'm not sure why active dry is still sold. It's the same yeast organism as instant, but with no way of knowing how much of it is dead or alive. Maybe it's for people who like surprises. As Franci says, in Italy they traditionally use the yeast very low quantities, and grow it through long fermentation times. Because of this, you're not going to be loading the bread with dead yeast, so you wouldn't expect to taste it. I believe the Italians prefer cake yeast the reason they prefer many things: unquestioning devotion to tradition. Bakers didn't start using cultivated yeast until around 150 years ago (they got it from brewers; it's the same basic strain we buy today). But that's about the age of the Kingdom of Italy.