So, I’ve recently gotten my hands on an ice cream churner with an in-built compressor. It is… well, I mean, if you’re only interested in making a batch of ice cream every month or so, or even every week, I’m not gonna say that it’s worth it. However, if you want to make a few small batches up in one day to compare them at the same age, then it’s kind of amazing. The upshot for my recipes is that the sorts of fine tuning that were difficult to do earlier have become a far more a realistic possibility.

Combining what I’ve learned from more experiments and some reading of Max Falkowitz, I’ve decided on some fairly radical revisions to my recipe procedures, and a few minor changes to the ingredients. Here are my ‘patch notes’, if you will, starting with the ingredient changes.

I’ve lowered the amount of dextrose across a lot of recipes. At one point, I added extra to specifically compensate for the lack of lactose in oat milk, even though I was already adding enough to begin with because it just felt like the right thing to do when I was originally getting my vanilla base down. The extra dextrose was making the ice-cream a little too soft, so now my recipes are firming-up better.

I did some blind taste-tests with friends on an array of fat levels and balances. The overall fat level I’m currently aiming for, about 12%, was preferred over the 7% or 16% values I tested it along side so I’m going to keep that the same, however there was a -slight- preference for a 4 to 1 ratio of coconut oil to neutral oil – but it was slight. I’ve adjusted the recipes to compensate, but it’s not a huge deal.

I’ve also been experimenting with adding some lecithin to see if that helps the resulting texture. What I’ve found out is that for certain oat milks (namely ones without added fat) it does help emulsify some of the added fats better, but otherwise doesn’t achieve anything. I’ll note this down in the glossary section for oat milk.

I also tried adding in extra maltodextrin, and did comparisons between 0, 25 and 50 grams of maltodextrin (In one-litre batches). I didn’t find a huge difference between 50 and 25 grams, but the no-maltodextrin ice cream was noticeably… worse? It’s hard to explain, but it just wasn’t as creamy. Glad my instinct to add it in the first place was well founded.

As for the procedure, I’ve reverted to a much quicker cooking time at a high temperature. Those bitter notes I noted in an earlier update were obviously just the pea protein scorching, which I don’t need to worry about as I no longer use pea protein in any of my recipes.

I’ve also stopped suggesting any set resting time, and now advise just waiting until your ice cream is cold enough to churn. This runs counter to a lot of advice given by other ice cream makers, but testing has shown that aging doesn’t really help, and damnit I like my ice cream sooner rather then later!: https://www.seriouseats.com/do-you-really-need-to-age-ice-cream-base-overnight

For test batches in my churner I don’t even wait for it to get that cold – since my churner can keep itself cool, I just let the base get to 30°C-ish and go for it! Obviously that won’t work for churners that require a pre-frozen bowl, though.

All these changes result in recipes that are marginally cheaper and easier to follow, for a gain in quality (or at least no drop in quality). They mostly affect oat milk-based recipes, and not my few homemade nut milk-based ones, but those need a separate pass later. Frankly, they just aren’t as good.