So with that theory covered, now back to the question of flour. And it is a question, because there’s no clear consensus on what to use. La Duree and Aoki use cake flour, Chef Eddy uses a mix, Herme uses plain flour… They do bake at different temperatures and use different milk/water and butter amounts, so it’s not as simple as just the type of flour producing a particular result. The combination is complex, but once I’ve worked through everything I’ll put it together and see what I can discern from all the variables taken together. For now, it’s all about the flour insofar as I can isolate it.
Last post it emerged that egg played a big role in the expansion of the choux, both by adding moisture but more particularly the impact of the uncoiling proteins on the ‘puffing up’ process. So one would think that the more egg you could get into the mix, without it becoming too soft, the better. Indeed, this seems a fairly universally acceped view. How could you achieve that? By using a flour that can hold a lot of moisture. This is largely (but I think not entirely) determined by protein content. For example, Chef Eddy says:
Using flour with fairly high protein content allows for extra eggs in the batter, which in turn permits lower oven temperatures but with good expansion of the products.
Chef Eddy is keen to bake at lower temperatures to avoid cracking, so he regards including a lot of eggs as really important, and reiterates elsewhere the role of a higher protein flour in enabling him to get more eggs into the paste:
I have used bread flour, with lower gluten (protein) content with good results, but mixing part bread flour and part pastry flour is very good. Certainly many chefs use all purpose flour. The reasoning for using flour with slightly higher gluten content is to permit more eggs into the paste. More eggs allow good expansion in the oven at lower oven temperatures. Lower oven temperatures help in the reduction of cracking.
Some bakers are exceedingly keen on high protein flour. Over at Iron Whisk, he runs a test that shows bread flour producing a much better expansion than AP (plain), and of cake flour he assumes it must actually be a mistake when he sees that Aoki uses it!
Now, in my test the Cake mixture absorbed the most egg, but I think that’s an aberration and I’ll test it again. However for now I think the best explanation is that the protein level alone isn’t the sole determinant of how much water flour can absorb. The type of protein and how it has been milled are also factors. Part of this is the fact that flour contains both gluten and gliadin, which are both proteins and I suppose combine to make the protein level listed on the bag. But it’s the gluten that’s important in baking, so the protein level as the sole determinant of water absorption and general performance in choux is going to be misleading. I started to research this and realized I may have gone a bit far… but if it’s your thing, start here and then go here. I would write more, but I have to restrain myself… This is why all my research papers become 20,000 word epics that take two years to get published. Enough!
What my test did show is that egg content was correlated with expansion – the Cake ones expanded much, much more than the others. So the flour that absorbs the most egg should be the one to choose every time, shouldn’t it. Why, then, if higher protein flour absorbs more egg, would anyone use cake flour? I’d say it’s the texture. The Cake choux had a much more pleasant texture, not soft but crispy like the crust of a cake and much more melt-in-the-mouth, rather than crisp and chewy like bread crust. So I imagine it’s about walking the line between absorbing egg and texture. In fact, this is exactly the conclusion others have come to. Joey Prat writes:
Most American chefs use bread flour in their choux pastry recipes. This is because bread flour, with its higher protein content produces an end product that expands better, with superior structure and a more hollow interior. It also produces a crisper shell. However, I find choux pastry items made with bread flour a bit tough and also chewier (more bread-like rather than pastry-like), and lacks the melt-in-the-mouth quality distinctive of its French counterpart. Most European chefs, on the other hand, prefer to use all-purpose flour (called plain flour in most parts of Europe; T55 flour in France), with great success. To bridge the gap between superior structure and melt-in-the-mouth tenderness.
We are of one mind!
Given all this, and given my success (this time…) with cake flour, that would be my preference. But it seems you can succeed with all three types of flour, hence this variable is one where it depends on what you’re aiming for and also your needs. Were I building a huge croquembouche, I’d make my choux with bread flour. If I wanted a delicate éclair, that didn’t need to travel much or stand up to a bit of delay before being devoured, I’d use cake flour.
I remain a little curious about the Shipton Mill Cake Flour and the mention of T55 has peaked my interest. I can feel a T55 / ‘00’ / T45 / Cake comparison coming on…. though I’m not sure the world really needs that.
A word of warning to UK bakers
‘Sponge flour’ here is low-protein, weak, fine flour BUT it almost always has raising agents in it and is therefore a big no-no in choux. I have no idea what that baking powder would do to the puffs (although it might be fun to find out. They might explode!), but I would stay away. I get my cake flour at Shipton Mill’s online shop and they also sell T55 as well as an array of interesting and unusual flours.