Beta test & final release
Last night, I did a beta test of my big meal. I’ll also cover the final release of the meal.
Given the complexity of the meal, I knew that I had to stage the different courses, and prep what I needed prior to the big event. Let’s start with the duck (fwiw, my wife’s phone had a different idea about the spelling of that word — I’m sure if you play with the vowel, you’ll figure it out).
I had a little trouble finding duck breast that wasn’t frozen — thankfully, I succeeded on my second try. Duck, being a cold weather water bird, has a layer of below-skin fat before you get into the meat of the duck breast. In order to properly cook the duck breast, you need to render that fat out, otherwise, it’s pretty unappealing. The skin is not fat permeable, so you need to score the duck while is cold, to create exit paths for the rendering fat.
I washed the duck breasts, dried them, scored them, and then salted and peppered each side. Following that, I took a sprig of fresh thyme and ran it along the long axis of the duck breast.
From there, I covered the back side of the duck breast and thyme with bay leaves, and then flipped them over so the bay leaves and thyme were underneath the breast.
For good measure, I decided to put bay leaves on top of the duck as well. Because it’s absorbing into the meat via osmosis, I figured I couldn’t over do this — I know from smoking briskets that the only thing that really penetrates the meat deeply is salt.
From there, I put the uncovered duck breast into the fridge. I’ll pull it out just before the I cook it the following evening.
For the meal, I had four components I needed to pre-cook: Creme Anglais, Blackberry Coulis, Potato Puree, and Asparagus & Hollandaise sauce. I started with the Creme Anglais.
Creme Anglais is actually just the custard base of ice cream. Actual chefs may quibble with this statement, but it’s essentially accurate. I make ice cream often — I know how to do this. That said, I decided to forget that I knew how to do this and use the T.K. method.
To start, I took 1 kg of half and half, mixed it with sugar and fresh vanilla bean and got it on the stove. The idea behind heating the liquid is do dissolve the sugar into the liquid and allow the essential oils of the vanilla to come out of the bean and into the fat of the half & half. Incidentally, if you’ve never worked with vanilla bean, it’s kinda amazing. It’s impossible to overstate how much more refined and rich the smell and taste is of the bean vs extract. I strongly encourage using the real thing vs extract.
The milk / cream / vanilla mixture gets heated to 185 degrees and then whisked into the 10(!) egg yolks.
Thinking that I was following the recipe, I added the sugar, whisked the egg, and then folded it into the heating mixture. From there, I started whisking. I still had it over the heat at this point, thinking that I needed to get it back to 185. To be clear, this is not right.
The way you’re supposed to tell a custard is the right consistency is by seeing if it will “coat the back of a wooden spoon.” I have no idea what that means — it’s a liquid, of course it’s going to coat the back of a wooden spoon. More on that later.
As I was heating the combined mixture back up to 185, I noticed the solution was getting more viscous — exactly what we want when making either ice cream custard or a creme anglais.
…And there’s the problem. I tried to salvage my efforts, but sadly, not even a chinois could save it. After hastily cleaning everything up, I decided to use the method I know works — combining all the elements into a pan and slowly heating the mixture till I get the consistency I want.
I started out with another vanilla bean, cut it in half and scraped it, added 750g of half and half, 250 grams of heavy whipping cream, sugar, 10 egg yolks, and put it on a low flame in a big (12") saucier. From there, I stirred constantly as the mixture. In the last photo on the right, you’ll notice you can see the bottom of the pan as I move the spatula — this is what I look for when I know it’s done. Definitely a lot more testable than if a liquid covers the back of a spoon ;). From there, it was into the fridge to cool. Part one of four complete.
The next thing to make was the blackberry coulis. Coulis are incredibly easy to make — it’s fresh berries, sugar, and a tiny bit of water. In my case, I just don’t dry the fruit after I wash it.
The observant of you will have noticed that in both this and the creme anglais, I’m using non-stick pans. What gives? I love stainless steel pans and usually cook with them. When I’m working with sugar, however, they’re immensely difficult to clean afterward. These two pans are the only ones that I use with any regularity for either this kind of thing, or if I’m making my daughter mac n’ cheese.
As the berries heat, the transform in color, while giving off an amazing aroma. I wanted to make sure that the natural citrus of the berries stayed in the coulis, so I stopped while the berries were still al dente and hit it with an immersion blender.
When I tasted the coulis, it was certainly tasty, but a bit generic. This got me thinking about T.K.’s conversation on flavor enhancers: salt and acid. I certainly wasn’t going to put salt into this, but maybe acid will help strengthen the natural tartness of the berry. I was already using balsamic vinegar in the duck sauce (more on that in a bit), why not try it here? The balsamic has a natural sweetness in addition to its acidity. The combination of the two was fantastic, and not something I would have thought of had I not started taking the T.K. master class. Two down, two to go.
Onto the meringue. I’ve never made one, or even really worked with egg in this fashion before. I was entirely dependent on the recipe with no other applicable experience.
The first thing that made me nervous was that I didn’t have regular granuated sugar — I had turbinado. Turbinado is actually unbleached sugar — if you’ve run across Sugar in the Raw, that’s it. I was concerned about two aspects: 1) the color, and 2), the size of the granules needing more time / heat to dissolve, given my first attempt at following instructions in the creme anglais.
To begin with, I took some of the 20 egg whites I saved from my creme anglais attempts and mixed it with the first half of the sugar. This time in a heat safe bowl. Life is a learning process.
I put the HEAT SAFE BOWL (this is still fresh for me) on a pot of bowling water and started to whisk. I needed to get the mixture up to 150 degrees. To my relief, the sugar seemed to dissolve quickly. In addition to the egg white and sugar, I added a tiny bit of vanilla extract and a bit more almond extract. I chose almond because I already had vanilla in the creme anglais and wanted something a little different. For the main event, I’ll use the other vanilla bean I have left.
As I whisked, the mixture became more and more frothy. It also started to lighten in color, to my relief. I kept checking the temperature of the liquid using my laser thermometer, which might be the single most useful kitchen gadget I own. After I got the mixture to 150, I pulled it off the heat.
From there, I took my egg mixture, and poured it into the bowl of my stand mixer. To that, I added another 150g of powdered sugar, the same amount as the turbinado I heated and whisked with the egg white and set it mixing. There’s something kind of magical about watching the egg white whip in the mixer. I knew from watching Mary Berry on the Great British Baking Show that it’s almost impossibly to overbeat the egg white in a meringue, but it’s very possible to underbeat it. With that in mind, I set the machine on high and let it go for 15 minutes. I can’t say I did this on purpose, but I do find it useful to have two different brands of smart speakers for ease of setting multiple timers (in my case, in this room, I have Microsoft and Google speakers).
As it continued to whip, the color gradually went from something that was kind of sand-like to a pure white. The consistency also changed to something close to a marshmallow fluff. After the 15 minutes was complete, it was time to make the meringues on the silpad.
The whipped sugar / egg mixture is extremely sticky. This can make forming the meringues difficult. T.K., shows a technique of using a hot water bath to keep the mixture from sticking to the spoon as you turn it out onto the pad. I found this was pretty easy, particularly when persuaded with a small spatula.
From there, they go into the oven to bake for 45 minutes. The recipe calls for them to bake for 45 minutes (!) at 180 degrees. I read that part 3 times because it seemed so weird to me. On top of all of that, my oven won’t go down to 180 degrees on bake — I did find that the warm setting was 180 degrees to into the oven they went. On warm.
After 45 minutes, the meringues were crisp on the outside, for all of a 1/16th of an inch, and entirely wet and soft beyond that, also known as undercooked. I decided that “warm” wasn’t going to cut it, put the oven on convection bake and set it for 200. After another 15 minutes, they were much better, though still possibly a little under. Next time, I’ll do the entire thing at 200 convection and start checking at 35–40 minutes.
For the final version, I did the entire thing on 200 degrees convection — 45 minutes means that there’s about 1/3" hard cooked meringue, with a creamy, marshmellowy center. I also used a fresh vanilla bean in this one (much better), instead of either vanilla or almond extract. This was a much better choice than the first time around (which was still good).
With the meringues baked, it was time to move onto the pomme puree, or fancy mashed potatoes. I will say, this was the single most physically challenging of the dishes I made, but that’s likely because I screwed something up, not because it’s *actually* physically challenging.
I used yukon gold potatoes — they’re the closest to french fingerling potatoes that are easily available. I should have used fingerling potatoes as they’re pretty available near me, but I didn’t. The potatoes I bought were huge — like nerf football huge. Comedic effect aside, the size of the potato becomes a challenge because in this method, the potato is boiled whole with the skin on. You leave the skin on to prevent as much moisture from penetrating the potato as possible — to make more room for fat, once cooked.
After boiling for about 30 minutes, I realized that the potatoes were still raw. Back into the pot for another 25 minutes. I would have left the potatoes in for longer, but the skin started to come off the potato as it boiled. Again, I think this has more to do with the physical size of the potato vs anything else.
I started with the smallest of the three potatoes. For it, the center was nearly cooked. For the others, not so much. As it turns out, it’s hard to push a raw potato center through a fine mesh tamis.
I managed to get enough potato out of the three undercooked yukons. Between the sieve and a lot of butter and cream, it still turned out smooth and creamy. It’s easy to pass this off as “just” mashed potatos — the tamis completely changes the texture, and the butter and cream enhance the flavor as expected.
For the final meal, I used fingerling potatoes, not yukons, mostly for the size. As my high school bio teacher taught us, the ratio of surface area to volume increases as the size of the object shrinks. All that is to say that if I don’t want to push raw potato through the tamis, use small potatoes. Fingerlings, russian fingerlings in this case, have a very smooth texture to begin with, and also carry a slight sweetness to them — an nice contrast to the acid of the duck’s gravy.
Onto the actual duck. Duck cooks for an improbably long time in order to render out the sub-skin fat. For the beta test, I cooked it on a medium low heat for 15 minutes.
After 15 minutes, I flipped it over for 2 minutes, then sliced it thinly and placed it on top of a smear of the potato puree. Special note: make sure that the crispy skin is up, not down in the potatoes — This bug was noted by our testers and corrected for the final release.
We also noticed that the duck was a bit closer to medium vs medium rare — we corrected this in the final release by cutting the cooking time to 12 minutes at and finishing on a higher heat to ensure that the skin was crispy.
The gravy was something that I thought a lot about — I used no recipe here, but instead started to think about the flavors that would both complement the duck (which has salt and pepper predominantly, bay leave and thyme as a back note).
The measuring cup has:
- 1/4 cup Amaro
- 1/2 cup pomegranate juice (no sugar added)
- 1/2 cup balsamic vinegar
- 1 cup turkey bone broth
- 1/2 tsp mustard powder
- ~1/2 cup of dried cranberries
The flower is for the roux (2+tbsp), and a half cup of fresh chopped oregano. In the final release, I omitted the oregano, accidentally, but it was a happy occurrence.
To make the sauce, I needed to start with the roux — that is, adding the flour to the hot oil leavings in the pan. Once you add the flour, you need to whisk it continuously on the heat while the flower itself cooks. This is a process of making the combined flower and oil mixture ready to accept liquid. Once the flower cooked down, becoming a light brown, I added the liquid mixture into the pan and began to stir. As it turns out, I don’t have three hands, so this process was, sadly, not photographed.
Once the sauce got the right consistency, I ran it through the chinois to remove both the cranberries as well as any other lumps that might have remained. With that, I spooned the sauce onto the duck as it lounged on its bed of potato puree.
With this done, the final step was to make the hollandaise.
250g of clarified butter, 3 eggs, a little water, and a bowl that will hold the juice of 1 and 1/2 lemons. While the duck was getting cold in the beta test, I was focused on this with the help of my wonderful wife. The process of making hollandaise is whisking the egg, while incorporating the clarified butter and lemon. After that, it’s adding a little salt to taste. The net result was:
As we finished up making the hollandaise, first, it was refreshing not to slash my thumb due to the glass bowl shattering in my hand. Second, because the blanched asparagus had been in the fridge, they were very cold — making the hollandaise solidify on contact — needs to be corrected in the next release. Finally, my astute beta tester noticed that I could have made the hollandaise while the duck was rendering — a suggestion I incorporated into the final release.
Following dinner, I put the meringues I baked earlier over the blackberry balsamic coulis and creme anglaise and then sprinkled them with mint. The finished product:
This was an incredibly rewarding thing to cook — it’s also opened up lots of other ideas for me to do later. Now I need to figure out another excuse to cook an elaborate meal :)
