Perfecting Choux Pastry – Another Project Begins…

Last post I put my failures on display. Embarrassing, but it’s done now. The question is what went wrong? How did I go from someone who can make choux to someone who can’t? I have a few ideas, and these are leading me to some of the aspects of choux I’ll explore. The main problems were trying too hard (over-cooking the panade), not concentrating (adding the eggs to a hot panade), and using flour I hadn’t worked with before (Shipton Mill’s lovely Cake and Pastry Flour. Not so lovely in my choux).

Choux has five main ingredients flour, butter, liquid (water and/or milk), salt, and sugar. The technique for making the paste has three main steps: bring liquid and fat to the boil together; add flour and cook over heat (ie make the ‘panade); adding eggs. Within this, there are decisions about how long to cook the panade, what temperature it should be when the eggs are added and how long it should be beaten for. Then there’s piping, whether to spritz (and with what), and the baking, which itself can comprise three parts: the initial phase, the ‘letting out the steam phase’ and the drying phase.

With my last piece of research on Victoria Sponge, I explored three key variables and needed to make eight batches to test them to my satisfaction. It’s not so simple this time. Here’s a breakdown of the variables given all the elements of making choux:

  • Flour: Bread flour versus cake flour versus a mix of cake and bread flour (3 options)
  • Liquid: Milk versus water versus milk/water mix (3 options)
  • Panade: Cook until skin forms versus cook longer than skin formation point (2 options)
  • Egg addition temperature: 60C v 50C v 40C (3 options)
  • Piping: thickness let’s say 1cm versus 2cm wide eclairs (2 options)
  • Piping: round tube v star tube (2 options)
  • Spritzing: water versus oil versus icing sugar sprinkling (3 options)
  • Baking temperature: 160C versus 180C versus 200C (3 options)

There is also the addition of pate sablee on top, but I’m going to leave that out because I think one doesn’t always want to have that on choux, so it’s not integral. Also, Chef Eddy has done a pretty definitive comparison that shows the pate sablee leading to much greater expansion. LINK. I’ll take that one as proven. He knows his choux.

Even just with that many options, if I made a batch for each combination that would be a lot. If I made a mixture for each that would 1,944 to be precise. And that’s not even starting on the different recipes with their different proportions of ingredients, of which there are a very great many. It was starting to look like as I planned it out:

Planning choux project

I needed to narrow it down a bit and focus on what seems to get a lot of attention (and what might have caused my fails). Based on this, I concluded the most important variables are:

  • Flour type
  • Liquid type
  • Egg addition temperature
  • Baking temperature

Then the fairly important variables are:

  • Panade
  • Cooking time
  • Spritzing

And finally the less important variables are:

  • Piping
  • Freezing before baking (the F Migoya step)

I think these are less important because its clear that people pipe choux in both ways with success, and most people don’t seem to worry about freezing and still produce great choux.

Not helping that much. So I wondered if there were any variables that were uncontentious, either across recipes and/or where the issue could be solved with science. There was one: egg addition temperature. There’s a lot of agreement on that, which I’ll discuss in a later post. The panade cooking time might also be solvable with science — I’m going to explore that one. So that narrows it down. Three flour options, three liquid options, three baking temperature options. That’s 9 batches cooked three ways. I can mess about with spritzing and piping within batches. That’s still a whole lot of eggs and butter to potentially waste, so I’m going to knock out an option — I’m not even going to try baking at 160C. I’ve always failed at it, and only a few recipes recommend it, notably Sadaharu Aoki and Pierre Herme recommend low baking temperatures and I have a feeling it can be done but only with skill and a precise oven. I doubt I have enough of the former and don’t think my aging Neff oven matches their equipment. So that takes it down to 9 batches cooked 2 ways. Still a bit much, and the difference between the liquid variables might not be sufficient, so let’s simplify to milk/water versus just water (because most recommend a mix, but very few suggest all milk). That makes 2 x 3 = 6 batches. Spritz some, freeze some. Doable.

But which recipe? There’s a myriad of them, and that’s a whole other headache for my next post. I’ll return… with a spreadsheet.