adrianvm

To investigate the effect of batch size on overrun I made a full batch of the same recipe I tested previously, 2 cups cream, 1 cup water, 16 g whey protein, 6 egg yolks, 1/4 cup erythritol, 2 T glycerin, 3/32 tsp KAL stevia powder. I cooked it, cooled it to 40 F, and churned it in my machine. Residence time was 18 minutes, so three times longer than a half-batch. Overrun was 38%, so double a half batch. The pre-churn density was 8.4 oz / cup, which is higher than my measurement last time of 8 oz / cup the last time. Not sure what to make of that. The resulting ice cream is very soft. It also melts very quickly. Another observation is that it seems to have a icy / watery feel...though it doesn't seem to be coarse. I'm not sure how to describe it. I'm wonder this is a result of decreased total solids because of the sugar reduction. Because it is softer than necessary I think the amount of glycerin needs to be reduced.

I revisited the recipe that had the whipped cream. It called for 3/8 cup half and half, 3 egg yolks, 25 g erythritol, 25 g polydextrose, cooked into a custard. Then you fold in 3 whipped egg whites and 1.5 cups of cream, whipped. I tried to determine the overrun, and had some difficulty. I estimated the density of my initial mix at 8.9 oz / cup, which seems too high. The measured density after churning was 4.6 oz / cup. If I trust those numbers the overrun is 93%. If I assume it was more like 8.2 oz / cup before churning then the overrun is 78%. The resulting ice cream is easily scoopable out of the freezer, but not as soft as the ice cream containing the glycerin from the previous test. It also has an odd hard feel in the mouth despite being scoopable. So it seems that I can increase overrun quite high through this method of incorporating whipped eggs and cream, but this alone doesn't give a satisfactory texture.

I think the guy at icecreamscience does say that longer residence time gives higher overrun. I hadn't thought that just using a mixer would create a stable enough foam to keep air incorporated. I didn't think you could get a stable foam with a fat content lower than heavy cream (36%). Does anybody have insight? I wonder if the stabilizers would help the mix retain air. I made a recipe that was prepared in a blender, with the blender running for several minutes, and it didn't seem to have a lot of incorporated air---but it was a weird recipe with butter instead of cream. I was thinking about folding in whipped cream---but not whipping all of the cream. I want to revisit the recipe I made before with the whipped cream---figure out what its overrun is and decide what I think of its texture. Of course, another phenomenon there was that the whipping of the cream made the volume of the mix much larger and may have increased residence time. (Air is an insulator.) I was looking that recipe over and noticed that it calls for 1 tsp low-sugar pectin. What, if anything, might this do? It's a custard style mix that gets heated to 180. (Normally pi I sampled the ice creams after about 24 hours of hardening in the freezer. The ice cream with the glycerin was soft and scoopable straight from the freezer. The ice cream with the cremodan was much harder. The ice cream was in a thin layer and I was able to break off some pieces with a spoon and after ten minutes at room temperature it was soft enough to eat. This made me wish I had a control with no special ingredients added, because I feel like even that result was somewhat softer than I would have gotten with no additives. Certainly in the past I've had batches (with different formulation) that needed to warm up for 30 minutes so I could cut them with a chef's knife into portions. The need to heat the ice cream mix to 186 seems like a significant limitation of the cremodan. I detected a slight iciness in the cremodan treated product as well. I suspect this may have been due to the time I spent measuring overrun (after churning and before hardening) during which the ice cream melted slightly. (However, it doesn't demonstrate great powers of ice crystal suppression from the stabilizers, which are supposed to keep ice crystals from growing during such abuse in the frost-free freezer.) So glycerin seems like a great option assuming that its metabolic impact is acceptable (which I still don't know). Cremodan may have helped, but without a control I can't tell for sure, and it didn't help enough by itself. But perhaps combined with increased overrun the result might be good. Because overrun wasn't controlled I can't isolate the effect of the additives in my study---was the glycerin treated product softer because of the glycerin, or because it had 3.6 times more air incorporated? I was wondering what the right amount of overrun is so I took a look at the Cooks Illustrated review of commercial ice creams where they list overrun, and found that their two top picks (Ben and Jerry's and Haagen Dazs) have about 25% overrun. It's interesting to note that almost every other product on their list has 95% overrun---and I think I read somewhere that the legal maximum is 100%, so clearly the other brands are all pushing the limit.

That's an interesting result. I'll have to try the icecreamscience recipe. Can you elaborate on the difference in flavor you noticed between the two recipes? What does "too tame" mean?

When weighing very small amounts I use weighing papers and very little remains on the paper. http://www.amazon.com/LabExact-Nitrogen-Non-absorbent-Cellulose-Weighing/dp/B00ARK0T2M?ie=UTF8&psc=1&redirect=true&ref_=oh_aui_search_detailpage I checked with my wife about the pistachio gelato and she thought it wasn't particularly soft. Since we actually served it after storing it in a cooler for several hours it was pretty warm at that point, and hence definitely soft. So my memory may be bad. I checked the book and it does say at the end of the recipe that "You may need to temper the gelato at room temperature slightly before serving it."

There's plenty of sugar in that Modernist Cuisine recipe. It seems strange that it froze so hard. I don't remember that being an issue when I made it. I don't think it has to do with your problem, but did you substitute guar gum for xanthan gum for some reason? I seem to recall that they call for xanthan gum. Regarding do-it-yourself pistachio paste, I don't know if you'll get better results with a vitamix or not. I have made nut butters in the food processor. It can take a long time and it looks like nothing is happening, but if you wait and let the machine run for a really long time it eventually works. Adding some pistachio oil may help keep it softer and hence near the blades rather than sticking to the sides, but I would add the minimal amount to keep things moving. I've never gotten anything as smooth as the commercial nut pastes I've bought (nor as smooth as what I've made using my grain mill). Personally I would use roasted pistachios, either buy them roasted or roast them yourselves. If you roast them too much they'll develop a roasted nut flavor that overwhelms the pistachio taste, though.

I refrigerate or freeze all nuts and nut products. Nuts go rancid. In fact, this is such a problem that I gave up on walnuts entirely because they always seem to be already rancid when I buy them. It seems like a lot of people can't identify rancid foods and think walnuts are just supposed to be like that. Walnuts in the shell were a revelation...but too much work. Pistachios aren't as unsaturated as walnuts, so they should be somewhat more stable. I don't think any of these things are going to go bad in a couple days, so if you're going to use it quickly it doesn't matter. But I'd keep in in the fridge. Katie, the Villa Reale product is only 20% pistachios---only 1.56 oz of pistachios in that jar---so it's probably not the best option for baking or cooking. It's also pretty expensive per pound of nuts, about $155 / lb of actual pistachio content at the current amazon price. The fiddyment farms paste and butter, and several of the pastes from L'Epicerie are 100% pistachio---always check so you know what you're getting. There's also a Pistachio Factory "butter" product that is 100% pistachio. I tried this last year when I sampled pistachio pastes and didn't like it as well as the fiddyment farms, but I don't remember why at this point. (It may be coarser.) If you want a spread that's only 20% pistachios you could buy a 100% pistachio paste at $40/lb and then mix up your spread to suit your tastes and it would be a lot cheaper.

Why 186? I looked up the constituents in Ideas in Food. sodium alginate: can gel at any temperature carageenan: hydrates around 175-185 locust bean paste: begins to degrade around 176 guard gum: no temperature given, but "can take 2 hours to hydrate in cold liquids" I figured based on those numbers that I wanted to heat it to 175 or so. My procedure wasn't great: it was 193 when I removed it from the heat. I strained it, divided it, and immediately added the Cremodan, and I neglected to take the temperature at that point, so I don't know how much it cooled. Perhaps the carrageenan wasn't activated. Xanthan gum doesn't seem to be recommended for ice cream. It's not clear to me if that's just due to cost or superiority of other gums for this application. Ideas in Food says that xanthan gum inhibits ice crystal formation and "is used in frozen preparations to ensure a smooth texture and mouthfeel." Locust bean gets mentioned frequently for ice cream as well. Since you sent me the cremodan I figured I would avoid the question of which stabilizer(s) to use and assume that the cremodan is a good combination of stabilizers and I can use it to assess the value and function of stabilizers. The cremodan appears to be overpriced, so if stabilizers appear to be important, I could then try to figure out my own formula. My general impression is that the more non-sugar sweeteners you use the better the result. (I think somebody suggested a combination earlier in this discussion.) In the case of Stevia, the plant produces dozens of sweet compounds. Some vendors sell a refinement containing just the single sweetest compound. I don't think it tastes as good. I stick to erythritol and stevia mainly because I'm less certain about the safety of other sweeteners. (Or in the case of some such as maltitol, they have a high blood sugar impact.) Everybody in the house eats ice cream, though I don't think the kids ever touch the sugar free stuff. They make their own, conventional ice cream.

I started a test today. I prepared a mix from 2 cups heavy cream, 1 cup water, 16 g whey protein, 6 egg yolks, 1/4 cup erythritol, and a pinch salt. I added stevia to taste (and lost track of how much I added). I cooked the mix (a little too much---oops), strained it. I divided it in half. To one half added 2g of cremodan 30 and to the other half I added a tablespoon of glycerin. Then I chilled the mixes in an ice bath and churned them. The first batch I churned was the cremodan mix. It chilled (much) faster (because it was in a metal bowl) and when it reached 40 deg F I churned it until it was too thick to move in the machine, about 3 minutes. I attempted to measure the overrun and estimated it at about 5% (the weight of a cup of the mix went from 8.2 oz to 7.8 oz). I chilled the second batch (with the glycerin) and it actually ended up a bit colder, about 37 F before I churned it. I'm using the Cuisinart with the bowls you chill in the freezer and I have two bowls that were both well chilled in my chest freezer, so I started with a fresh bowl each time. The batch with the glycerin had a much longer residence time in the machine, about 6 minutes. It was clearly softer at the point where it was no longer moving in the machine. I measured the overrun at 18% (the weight of a cup went from 8 oz to 6.8 oz). I do wonder what the effect of batch size is on overrun. I'm also not sure why the mix with the glycerin was 0.2 oz lighter before churning. One possible explanation is that because the mix containing cremodan was thicker, I filled the measuring cup above its sides slightly and had more mix in it. If that's the case then 5% was an overestimate on the overrun. The remaining ice cream is chilling in the freezer and I'll assess scoopability tomorrow.

I thought you didn't like their shipping rates. As I said in the other thread, I found the pistachio (and hazelnut) pastes L'Epicerie has imported from Italy to be overly roasted, but the texture is very good. I thought the domestic product from fiddyment farms had a better flavor.

You may wish to keep the scale you ordered for regular kitchen use at higher weights, especially if you do any baking. You can get one of the inexpensive low capacity scales for measuring at 0.01g accuracy and use the scale you bought for other stuff.

Dextrose is another name for glucose.

I have to admit I'd be more interested in the dulce de leche variation on the ice cream than the chocolate sweetened condensed milk version. (I do tend to find chocolate ice cream disappointingly unchocolatey). But I'll be curious to hear about your results. Be aware that the ratio I reported was the first hit on my google search, not the result of an exhaustive search. One thing to be aware of with the stevia sweeteners is that many of them are cut with something else, so for example Stevia in the Raw is actually glucose plus stevia in the packets, and maltodextrin plus stevia if you have the "bakers" version. So if you use a lot of that product you're sneaking sugar in through the back door, which would invalidate your test. The stevia is about 100 times sweeter than sucrose, so it's hard to meter it out by itself. My jar of concentrated powder came with a little scoop that is around 1/64 tsp. One way to do a texture test would be to make it unsweetened, because stevia is so concentrated it will have no effect on the texture---so you could just freeze a blob of whipped cream and see what you get. There are a few different sweetened condensed milk recipes out there that involve long cooking of milk (with erythritol) or coconut milk (with erythritol). This is a lot of work and in the context of the ice cream problem...just skip it and add the erythritol if that's the answer. I think the attraction of condensed milk is the simplicity. The problem with the condensed milk recipe you linked to is that it contains powdered milk---a lot of powdered milk, and milk contains sugar. The recipe gives the range of 2-4 cups of powdered milk, which corresponds to 139 - 278 g of lactose. So there's actually quite a lot of sugar in this condensed milk. She doesn't say what the yield is, so it's hard to compare, but a can of regular condensed milk at about 220 g of sugars seems to be comparable! Furthermore, she states in the ice cream recipe (The one for "belgian" chocolate) that it was soft after 2 hours in the freezer, but went hard over night. Note that from what I've seen, most people making sugar-free ice cream expect it to be rock hard in the freezer.

It could be that the proteins in the condensed milk are also playing a role, but I think the sugar is the key ingredient. Why use condensed milk instead of sugar? Because condensed milk has the sugar already dissolved in a very high concentration. To achieve the same effect with sugar you'd have to dissolve it in a liquid and to do that you'd need to apply heat, and then you'd have to wait longer for it to cool down. (Condensed milk tastes better than pure sugar because of its milk flavor. This would probably come out pretty good if you cooked the can first to make dulce de leche.) I'll bet if you used corn syrup instead of condensed milk the recipe would work. But that would make people uncomfortable. The recipe I found for this is 2 cups cream, 1 can condensed milk. So that's 1.1 cups of sugar, which is more sugar than I use in my conventional recipe. If whipped cream doubles in volume that means overrun is 60%, so I think you can mainly understand this recipe on the basis of sugar and overrun. When I tried a recipe that called for whipped cream (and whipped egg whites) without the large amount of sugar (1.25 cup heavy cream, 4 eggs, 1/2 cup erythritol) the resulting product froze hard and had an icy texture. It wasn't scoopable.

I normally make a traditional ice cream with 2 cups heavy cream, 1 cup milk, and 6 egg yolks. The fat content of the dairy portion is hence 28.5%. I have made some sugar-free formulas where the fat content is 50%. Such a high fat formula is OK with me. Formulating a mixture that has the desired fat content without using dairy milk is not difficult. Cream can be diluted with water, for example. How important are proteins in the milk for building body compared to fats? It would not have occurred to me to whip the cream before spinning. But I had a recipe that said to do this, so I tried it. This produced the only no-sugar ice cream I've made so far that didn't freeze solid in the freezer. I think one issue I have is insufficient air incorporation. As I noted before, the ice cream freezes completely in 5 minutes when I make a small batch, so there's not much time to incorporate air. IndyRob, I tried a no-sugar recipe that called for whipped cream to be placed in the freezer without churning and the result was far inferior to the result I got when I churned the whipped cream and whipped egg white mixture. Whipped cream alone tends to freeze up fairly hard into something that isn't scoopable and has a kind of dry texture. I suppose when you add the condensed milk you've got so much sugar around that it keeps it soft no matter what.

Actually they don't work the same way. Most of these substances pass undigested into the intestines, where bacteria go to work on them. But erythritol is absorbed and excreted in the urine. http://www.ncbi.nlm.nih.gov/pubmed/8457525 Note the abstract says almost all of it followed this pathway. Of course, with anything like this, assessing your tolerance before use is certainly a good strategy. So would there be any advantage in a very high fat mixture of using whey? Or casein, the other protein in milk powder? (Or even milk powder?) In this situation water is comparatively scarce and there are lots of solids (the fat globules). The point of using proteins instead of nonfat dry milk is that nonfat dry milk contains sugars. I don't have a problem with icing specifically, but I may want more overrun. I would guess that, except in the case with the whipped cream, my overrun is very close to zero. (With the whipped cream and egg whites it could have been over 100%.) I saw somebody mentioned putting lecithin into their ice cream. If I'm already using a bunch of egg yolks is there any point in that? (As an aside, I tried the Modernist Cuisine approach of using lecithin in salad dressing with strange results, like dressings that refused to emulsify, or seemed pathologically unstable. And the lecithin was very hard to clean up.) What is meant by "water activity" and how does that help keep ice cream soft?

For $45, that is suspicious. At Sartorius prices, not suspicious. I felt like I got a smooth texture frozen dessert with the various smooth pistachio pastes I used. Not "pistachio butter," mind you, but the "pistachios paste." The pistachio butters I've had have all been coarser than the pistachio pastes, and none would have produced a smooth textured gelato. I also got smooth textured results with the hazelnut pastes. It's been a year, but I believe these pistachio and hazelnut pastes were as smooth as or smoother than Jif, and I don't recall any sensation of grittiness at all in the MC gelato I made. I'm not sure what you're asking here. Are you asking what to do with the love'n bake pistachio paste that has the sugar mixed in? I mean, you may as well try using it to make the MC recipe as a trial run. See how it goes, whether you think it's gritty, etc. Add the pistachio extract if you think it will help. Sometimes those extracts have a funny flavor and make things worse. Regarding the composition of your paste, you neglected to note the protein, which is the key ingredient in determining the pistachio content, since none of the other ingredients contain protein. The protein is 4g, then that means you have 20g of pistachios. At 45% fat the pistachios contribute 9 g fat. I believe this leaves 4g for canola oil. And then we have 6g sugar to get things to add up. (Note that pistachios are 7% sugar, so some of the sugar is from the nuts). This means added sugar is only 6/30=20%, which is less than I recall being told by the manufacturer. Maybe they have lower protein pistachios? To use this product in the MC recipe you have to figure out what fraction is nuts, what fraction is oil, and what fraction is sugar so you can get the quantities right for the recipe.

I made a recipe that had (too much) xanthan gum in it. The resulting ice cream still froze very hard, so xanthan gum alone didn't seem to be helping with the scoopability problem. I also noticed that mgaretz thought a particular stabilizer combo (Cremodan) wouldn't help scoopability. I am perfectly happy using lots of eggs. But I'm also willing to use hydrocolloids--even hard-to-get ones---or proteins (gelatin). Any idea what pectin would do? It's in some recipes I've seen. Tell me more about erythritol's endothermic reaction. I've used it a bit in cakes and cookies and the cooling effect is pronounced and really annoying. Is there a possibility of eliminating or reducing that effect in this context by keeping it dissolved, somehow? (I mean, in baking water is being driven off, if any was even there to start with, so there's not much to be dissolved in.) In ice cream I don't notice a cooling reaction. Seems like that would be hard to notice given that the ice cream is cold. But I do notice that it makes the back of my throat feel funny and the funny feeling persists for a while, several minutes at least after eating the product. I don't recall having noticed this sensation from baked products containing erythritol. Also I don't notice any grittiness (e.g. from erythritol crystals that are large enough to feel). I'm not interested in making chocolate ice cream. Even with sugar I always find it disappointing. It doesn't taste enough like chocolate for me. (I want my chocolate to taste like chocolate---75% cocoa.) I've gotten satisfactory results making my own chocolate chips by mixing unsweetened chocolate with powdered stevia concentrate. I also find that when the food is cold it seems to minimize off tastes from stevia, so getting adequate sweetness is really not a concern. I'm mainly interested in eggy-vanilla, or maybe almond extract flavored, or flavoring with orange or lemon. I think smaller batches are producing harder ice creams for me. My batch sizes tend to be pretty small, since I'm experimenting and I don't want to commit a ton of ingredients to an uncertain recipe. I'm probably making about a pint at a time. I'm observing that the batch freezes in my machine (the cuisinart with the freezer-chilled bowl) in under 5 minutes. Rapid freezing should control ice crystallization, but it means there's little time for air incorporation. Also when the mix has higher viscosity I feel like air incorporation may be limited because the mix just doesn't agitate very much. When I made the ice cream that contained whipped cream (fully whipped) and whipped egg whites I didn't notice any butter formation or an unpleasant mouth feel when it was fully frozen. The texture was very weird straight out of the machine, though. I wonder if this technique might be useful where only half the cream is whipped, so I force in some air, but not so much. I actually also made an ice cream that was made with butter and mechanically emulsified with eggs and coconut milk in the blender and the mouth feel seemed acceptable to me. These mixes are higher fat than normal ice cream (the latter one was 50% fat). However, when I made a more traditional recipe with a more typical fat content (26%) based on simmering orange peels in clarified butter to extract only the non-bitter flavors I got an ice cream that did not have a good mouth feel and texture. If the ice cream is very high fat then it will "freeze" faster since fat has a lower specific heat than water. With the mix starting at 40 deg F the butterfat must be in the solid state already when it goes into the machine, so no phase transition required. Again, good for keeping ice crystals small, but bad for incorporating air. But as the amount of water declines I wonder if that changes things.

I have a scale that measures 3kg in 1g increments. It cost $100, I believe, but is dishwasher safe. I killed my previous scale trying to clean fish off the buttons, so I thought the extra expense of dishwasher safe made sense, and after about 10 years with the scale I feel it was worth it. The thing to be aware of if you're buying cheap scales off ebay that ship from the far east is that precision (measuring in 0.1 g quantities) does not necessarily translate into accuracy. It doesn't do much good to read 0.3g if the error is plus or minus a gram. I personally would be reluctant to buy from such a source for this reason, and also dubious about a scale that has a huge range and a high precision. That's why the pocket scale for 0.01g readings makes sense---it's actually plausible that they might be accurate. A scale that measures to 4500g and is accurate to 0.01g has to be accurate to 2 parts per million, and you're just not going to get that accuracy for $45. Or even for $450, I'd guess. Note that 3kg isn't actually enough in the kitchen on rare occasions, usually because my tare weight eats up too much capacity, e.g. if I put a pan on the scale and tare. Yes, you can work around it by using more dishes, but that's annoying. Nevertheless, I decided that this limitation would be worth tolerating to get a dishwasher safe scale.

It's not important to the discussion whether what I'm doing is right, or makes sense to anybody. (If it's wrong, that's my problem. Note also that a few decades of eating tons of sugar and starches may have left my body in a state not encountered by evolution.) This is not an argument we need to pursue---this is a food forum, not a health forum. I don't believe I've made any health claims, nor do I believe I'm spreading any memes here. Just think of it as an interesting culinary challenge to make ice cream without the use of sugars (or starches). Maybe I want to make a savory ice cream. 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. Does milk help primary because of its sugars or because of the proteins? 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 would imagine proteins wouldn't be very effective compared to sugars, though, if molecular size is the key, since proteins tend to be big. Do you have any idea what goes wrong if you add too much whey (or egg white) protein to an ice cream? What determines the maximum reasonable amount? (Note: egg white protein molecular mass 45000 u, whey protein about 22000 u, and sucrose about 320 u, so the proteins are a hundred times larger. What's the smallest readily available (soluble) protein?)

The scale I have (which looks identical to the one I referenced on amazon) is much smaller than 8" square. It is about 5" x 3", 1/2" thick and folds up into a little box you could stash anywhere. But I only use it for measuring spices and hydrocolloids, not for general cooking. It only goes to 100g max. My general cooking scale is about the size of Lisa Shock's, though I keep it permanently on the counter because I use it every day. But it measures in grams, not 0.01g. A scale that measures 4500g in 0.01g sounds a little suspicious. I'm checking my copy of MC at Home, and the Pistachio Gelato recipes has the obvious differences from PB&J gelato, but also the sugar is increased for the pistachio recipe to 155g (3/4 cup), presumably because of the juice/water substitution. Everything else is the same, with pistachio replacing peanut and water replacing the juice.

I'll definitely give that a try next time.

I did a tiny bit of reading and some people say glycerin doesn't affect blood sugar and others say it does, perhaps with a delay. A clear consensus was not apparent. Regarding choice of sweeteners, I find that erythritol produces an annoying cooling sensation in general. I usually can't use very much of it for this reason. In ice cream, presumably this wouldn't matter, but I find that it gives me a weird feeling in the back of my throat. Stevia is tricky, because every brand tastes different. I tried a bunch until I found one I liked, and then they changed the formula. So then I tried more and found another one I liked (KAL). I've also noticed a lot of variation in how people taste stevia. Some people seem to be prone to tasting some unpleasant aftertaste or overall taste with it and others don't. Since polydextrose isn't very sweet presumably it would need to be combined with stevia or some other high intensity sweetener. I mean, the same would be true with glycerin or alcohol.

No, a postage scale is unlikely to work. I have a regular kitchen scale that measures in single grams, which is probably similar to what your postal scale does. You need about ten times as much precision to measure 0.3 g. But JNorvelleWalker is right that suitable scales are not expensive. For example: http://www.amazon.com/American-Weigh-0-01g-Digital-Scale/dp/B0012LOQUQ/ref=sr_1_1?ie=UTF8&qid=1462991258&sr=8-1&keywords=precision+scale

I think I want to avoid alcohol, though if the amount were small maybe it would be OK. When I have added small amounts of alcohol to ice cream I haven't noticed a change in texture, so that seems to suggest that bigger amounts are required. Glycerin is a weird best. I can't figure out how it's metabolized. Like a sugar? Or not?