Monday, December 21, 2009

Perfect Pastry Cream


I have come to realize that there are many more people in the world who are way way smarter than I. A perfect example are the people at Alicia, in this particular case Ms. Laia Badal (chemist, staging at Alicia) and Mr. Pere Castells (head of the Department of Scientific and Gastronomic Research at Alicia... you may have seen him before; he was in the Anthony Bourdain episode titled "Decoding Ferran Adria", when they are in the workshop testing items out, Pere (pronounced Pera) is the person in charge of the chemistry part of things, so in my book, a rock star), the chemists responsible for developing this method which I explain below.



They do have a bunch of cutting edge equipment and technology, but most importantly, a good head on their shoulders. One of the most remarkable experiences for me during this trip was the method that they had developed to make pastry cream. The original questions for them were, how do we make it better? How can we make it faster? How do we simplify it? And I think that this may be the core of Alicia, to find better and simpler ways to make great food. The theory was that pastry cream was complicated to make and more frequently than not, the result was lumpy and grainy. Believe me, I taught this method to my students and 95% of the time the result was not optimal.

This is the classic procedure:

Place 95% of the milk (or milk and heavy cream mixture) in a sauce pot with half the sugar, in another bowl, mix the remaining milk with the cornstarch and stir well, add the egg yolks and the remaining sugar; stir until homogeneous. Must be lump free (strain if needed).

Bring the liquid to a boil and temper the egg yolk - slurry. Return the heat and bring to first boil while stirring constantly. Take the pot off the heat and stir the butter in; cool off and cover with plastic to avoid a film.

The problem with this method is that it is very aggressive on the egg yolk and the cornstarch. Egg yolks begin to coagulate at 80 degrees Celsius / 175 degrees Fahrenheit and cornstarch coagulates between 80 and 85 degrees Celsius / 175 and 185 degrees Fahrenheit. The theory was that the mixture did not need to come to a boil after tempering the egg yolks in (which is what would over-coagulate the yolks and the starch, resulting in a lumpy, grainy cream), if you could get the liquid (milk or milk and heavy cream) hot enough to coagulate the yolks and the cornstarch in one quick shot.

This is the Alicia procedure:

Mix 95% of the milk with the sugar in a sauce pot; turn on to high heat.

Meanwhile, combine the remaining milk with the cornstarch and make a slurry. Stir in the yolks. Strain through a chinois to avoid lumps.
The milk and sugar mixture need to come to a rolling boil, so you need to use a tall pot. The milk needs to come all the way to the top of the pot, then you dump it all at once into the milk-cornstarch - yolk mixture while stirring constantly (so ideally two people are involved). And then magic unfolds before your eyes as you stir. You have the smoothest, richest, creamiest, lump free pastry cream you have ever tasted.

This method does not work with recipes smaller than 1 liter, since the milk cannot reach the desired temperature that can coagulate the yolks and cornstarch instantly. But who makes only 1 liter of pastry cream?
I know a lot of folks who would frown down on this method, but, the proof's in the pudding. Enough said. Method embraced.

41 comments:

  1. Awesome! How long does it keep in the fridge? In the freezer?

    ReplyDelete
  2. I would not freeze it; I would hold refrigerated for 5 days max.

    ReplyDelete
  3. Hey, looks great! Can't wait to try it.

    ReplyDelete
  4. is there any butter in the Alicia method??

    p.s. got the new book & love the new format... can't wait to read it

    ReplyDelete
  5. There is no butter (I asked the same question); butter acts to make the pastry cream smoother in the original method. In this method it does not need it.

    ReplyDelete
  6. Thanks for the method, sounds great, will try it soon

    ReplyDelete
  7. I was working in a well busy bakery in Melbourne 4 years ago... it was so busy stirring a pot was painful so we came up with pretty much that technique. the speed of it is awesome, except when a muppet forgot to add the sugar to the milk and it burned on the base.
    P.s. Your Cafe book is in the mail, can't wait!

    ReplyDelete
  8. Chef,

    You say that it cannot be done with formula sizes less than one liter because the milk does not reach a high enought temperature. Are you saying that when the milk foams to the top of the pot, that the milk actually reaches temperatures over 100˚C? Or is it that the milk/sugar mixture has enough residual heat when temered into the egg mixture? Please let me know.

    Thanks,
    Bernie

    ReplyDelete
  9. Smaller quantities cannot carry over sufficient heat to coagulate the proteinsin the yolks or starch

    ReplyDelete
  10. would thickening the milk sugar mixture then adding slurry let thicken then let cool down a bit then adding eggs in while reheating and stirring work? so basically like making and anglaise with cornstarch at the beginning then adding egg maybe even sous vide?

    ReplyDelete
  11. well, Craig, that would defeat the puprose of simplifying the process.

    ReplyDelete
  12. haha i know i was talking for the same texture in a small batch since you need to make this in bigger batches.

    ReplyDelete
  13. Hi,

    Wouldn't the amount of yolks and starch be scaled down in relation to the amount of milk?

    So theoretically speaking, the heat generated from 1 litre of rolling boiled milk ought to be hot enough to coagulate the relative amount yolks and starch.

    Hhmm..got to try this out...

    Thanks for the mind blowing post!

    ReplyDelete
  14. Hi again,

    Just saw that it doesn't include butter. In my understanding, butter contributes to a smoother texture but it also provides the rest of the mouthfeel experience due to its melting temperature. So I am really curious about butterless pastry creams. Interesting...

    ReplyDelete
  15. Aha... Alex Talbot of Ideas in Food was just telling me about this method (abbreviated) and I did not believe it would work. But now with the explanation of volume and the tall pot, I can see how it results in the pictures you posted. But does full coagulation takes place? What was internal temp at finish?

    Francisco - as always, thank you for sharing! T.

    ReplyDelete
  16. boil then dump into the vita. thickens nicely.

    ReplyDelete
  17. Intre: that's bad form. Plus it takes longer.
    Tina: full coagulation occurs at 85 C, so to answer your question, it is fully coagulated. what may throw you off is that it is not stiff and hard. well, it's not supposed to be stiff. We have just embraced that (much like boiling it) as it's accepted texture and it is not, which is why many people have to paddle it before using it to soften it again. why do that when it can be smooth always?

    ReplyDelete
  18. Sr. Francisco,

    Please excuse my naivety, I am curious about the absence of flour in this preparation of 'pastry cream'. Also do you have any ratios you can suggest. I look forward to attempting this recipe first thing tomorrow.

    ReplyDelete
  19. Pierre Herme's method boils the cornflour first, bringing it to coagulation, then uses that to temper the yolks and cook out. This allows the yolks not to be subjected to temperatures beyond 82C. I have found this to be the best method. I also think that butter lends more than just texture, and should not be given such short shrift!

    ReplyDelete
  20. That's great, Alastair. So I suppose you came to this conclusion after making this recipe and then made the decision?
    I don't understand why the cornsrtarch (or "cornflour" as you wrote) needs to be boiled wth the milk. If cornstarch proteins gelatinize between 80 and 85 celsius, the milk does not need to boil.
    What is "shrift"?

    ReplyDelete
  21. Now I feel like I'm being given short shrift. Excuse my british english. Yes, I've made both. There's no need to dismiss someone who's got a different view so quickly. I got excited when the method was posted at I.i.F. So I tested.

    I need not have written boil, just a turn of phrase, as of course the temperature of the starch needs only 85C, however excess temperature in this mixture is not detrimental, as opposed to the egg mixture and it also shortens the cooking time needed to bring the eggs to temperature when the two are mixed.

    Hence, in separation, the cooking of both coagulants is better controlled. I can cook the eggs to a precise temperature of my choosing. Heston would approve. :-)

    BTW "Short shrift" is a british phrase meaning being given the brush off or disregarded. I like butter. :-) I use both dairy and cocoa butter to enrich and carry flavours into pastry cream and improve mouthfeel.

    In addition, the common misconception in the orthodox method is that the mixture needs to boil. If you've truly boiled the mix then the whole cream has reached 100C. No wonder it comes out grainy... The temperature profile of a thick mix in a pan will be a function of the mixing (both in quality and speed) and the amount of heat input below. As both proteins require in the 80's, there's no need to bring it anywhere near the boil. The boil is in fact local superheating on the base of a poorly mixed pan, ending up with something akin to a hot mud spring. The mix is too thick to distribute heat by convection, so stirring is essential. The instruction to "boil" is what your 95% is made up of...If you want a more robust method for the orthodoxy, then I say give your students a thermometer, and some lessons in egg thermodynamics.

    ReplyDelete
  22. Alastair, you are absolutely right. There is more than one way to skin a cat. Can't we all just get along?
    I promise not to shrift again.

    ReplyDelete
  23. Never an intention to rant :-)

    Cat skinning. Now there's a thread!

    ReplyDelete
  24. Is it possible to simply combine everything in a bag and bring it to 82 or 83 celcius in a waterbath? It could then be whisked briefly to combine, in the same style as a sous vide creme anglaise. I'll know soon!

    ReplyDelete
  25. hey Chef what are the actually measurements of milk, sugar and so on i would love to try this recipe to use in my creme puffs! Thanks, Flushing NY

    ReplyDelete
  26. Stephanie, see my posting here:
    http://www.thequenelle.com/2010/01/perfect-patry-cream-recipe.html
    There's the recipe

    ReplyDelete
  27. Hi Francesco. Very interesting idea on the Pastry Cream from Alicia. This is what I understand - yes both egg yolk coagulation and flour starch gelatinization occur in the 80s degrew centigrade. Higher temperatures are not necessary. However apparently egg yolks contain a starch digesting enzyme that requires boiling to be inactivated. This is the science behind the reason why pourable custards (i.e creme anglaise are brought to 82-85 centigrade and why starch added custards (i.e pastry cream) are brought to the boil. This is a rule of cooking that has beeb passed down. Harold Magee explains that unless these egg yolk enzymes are deactivated atby boiling they will cause a stored pastry cream to liquify during storage. I have never tested the theory but have always accepted the advice and boiled starch thickened pastry cream. My method for a more tender cream is to increase the yolks (up to 16 per litre of milk) and to reduce the amount of flour (down to 40g - less for a pure starch). this gives a creamy tender cream however it has limitations in it's use due to the fact that it is not that firm. Can anyone talk of stored non-boiled pastry creams and how they change over time - do they liquify progressively? Rossa

    ReplyDelete
  28. Hello Rossa
    This is a very good point. I have observed that the next day the pastry cream looks rather lumpy and not as smooth as when it is just made, but after a vigorous mix with a whisk it goes back to a smooth consistency. I have not been able to see how it looks after more than a day, since what we make does not last us so long, but it would be interesting to see what these starch eating enzymes do. I will say though, that as far as comparing the two vis a vis, when they are just made, this new method is far superior in every single way.
    Thanks for your thought provoking comment.
    And it's Francisco, not Francesco.

    ReplyDelete
  29. Hi Francisco. I will certainly be trying your method next time I make pastry cream and will attempt to store some boiled and unboiled pastry cream over the next few weeks and try to report back on the result. Rossa

    ReplyDelete
  30. As far as I understand the magic of pastry cream, the starch is added to allow the mix to boil at a higher temperature and prevent the eggs from overcoagulating.
    Boiling the pastry cream for two minutes is super important to kill the enzyme alpha amylase in the egg yolks. If this enzyme is not killed it will cause the pastry cream to loosen when kept for storage. Then the texture will change from firm to soupy overnight.

    ReplyDelete
  31. Jenny:
    That is true to an extent. The texture however does not change overnight, in my experience you have about 48 hours. In my kitchen we make only enough for two days. It is worth it. Why would I want to make boiled out overcooked pastry cream just to have it hold for days on end? I would rather do the work to get a better product, and this is waaaaay better.
    furthermore, boiling pastry cream for two minutes will render any pastry cream inedible. If you are going for that method you just need to wait for the first boil and take the pc off the heat.
    finally, you can't "kill" an enzyme, but you can denature it.
    Thank you.

    ReplyDelete
  32. has anyone tried this method at a high altitude?

    ReplyDelete
  33. Made some really lumpy pastry cream in the kitchen yesterday. Hope this method would help out come Monday. Thanks for the tip!

    ReplyDelete
  34. I am thrilled by the idea of this method. I currently make pastry cream daily at work and love the idea of cutting out the time spent stirring on the stove. Over the last four months I have been using this new method with mediocre results. Mine never gets as thick. When the video was posted it confirmed that. It does however thicken some, but I still have to return it to the stove. In troubleshooting the only difference I can find is altitude. I work at above 5000 and things boil on average 10-15 degrees lower than 212. I'm hoping that this does not mean this method is out for those of us who work way above sea level. Any thoughts?

    ReplyDelete
  35. this turned out HORRIBLE. and i waited for a really good rolling boil. extremely disappointed and wasted a litre of milk plus eggs.

    the attached recipe is also not clear. what is 80.0 corn starch? 80 ml? 80g?

    i am VERY VERY disappointed

    ReplyDelete
  36. If I read you correctly you are saying you are disappointed. Sorry to hear that. It is 80 grams of cornstarch.
    Also? Clearly you have anger issues, you may want to look Ito that.
    I've made this recipe for 2 years now without a problem. Have a wonderful day.

    ReplyDelete
  37. I just want to thank you SO much for posting this recipe. I tried the traditional method and my custard always came out grainy (not lumpy but more like sandy). I followed this procedure step by step with the minimum requirements and it turned out PERFECT. I added vanilla extract at the very end of mixing. I was a little confused with the quantity of the ingredients, as they are by weight (grams or lbs.) I used 1 gram = 1 ml and ended up using 1 liter milk, 10 egg yolks, little under 1 cup of sugar, and 5 Tablespoons of corn starch. Does that sound about right? (I don't own a digital kitchen scale). THANKS SO MUCH!!!
    Aziz
    Houston, Texas

    ReplyDelete
  38. Thank you! Tried this method this afternoon in the kitchen, and it worked! Silky pastry cream in minutes! Will additional starch will firm it up a bit? Sandra, Melbourne - OZ

    ReplyDelete
  39. That depends on why you want to do it. I think as it is it is perfect to eat. Firmer will resemble the traditional method which is like wall paper paste, which is exactly what the recipe is trying to avoid. Glad it worked!

    ReplyDelete