Latest panettone bake and analysis

Every panettone bake is an experiment in one or more ways. This one was a trial of a chocolate panettone recipe which intrigued me because of its ingredients. 

First, it uses cocoa powder, rather than a ganache. This is easier and less costly than the cocoa mass-based approach, and also lighter. Second, it is quite high in egg yolks compared to most recipes. I wondered how this would affect the crumb; certainly the photos were encouraging.

I found the recipe to be accurate and quite workable. The dough handled well, and the only problem I had was a gap in the instructions around the use of water in the second dough. Nothing major though, and mixing went well. My starter was in good shape, as I had taken it out of storage 3 days earlier to feed it daily. 

The first impasto rose 4X in just under 12 hours, and the pH was 4.78 at completion. 

The final rise only took 3.5 hours, but I tried to stretch it out a bit longer.  The dough was scaled at 1020g in a 1 kg mould.  

I am pretty happy with the result. The loaves rose dramatically, and have a light and shreddy crumb. The chocolate flavor is not too pronounced, but it enhances the plentiful inclusions (white, milk and dark chocolate). I glazed the loaves, and found that this recipe is delicate enough to require turning the loaves periodically to avoid splitting on the side away from the convection fan. The original recipe was done with no glaze, and used the scalpare approach. This allows expansion vertically, where we want it.

But this got me thinking about how this panettone differed from my own  recipe and others I frequently bake. While the high  yolk content favors plasticity and a high rise, it comes at the cost of some reduction in tenderness, and a slightly drier crumb overall. This was in spite of a dough texture that handled very well, was moist enough to ferment very well,  and had adequate sugar saturation, salt etc.

Ternary Analysis

While I have primarily focused on sugar control, mixing technique and hydration previously, I wanted to examine some of the key ingredient relationships that vary significantly between recipes. To do this, I’ve started to use ternary plots to help me understand the “sweet spot” of enrichment. Disclaimer: I started out as a programmer, so this is also fun for me to do!

In the case of this recent bake, it seemed to me that the amount and type of enrichment relative to the amount of flour in the recipe was too “extreme” to permit smooth doming during the bake. I have a couple of initial plots which bear out my experience. 

The ternary plot allows you to look at three elements at a time, and compare multiple datasets (recipes) to consider the differences. Initially, I looked at flour, butter and yolks in four recipes I’m familiar with.

Unsurprisingly, all four recipes are in a fairly small cluster. However, zooming in we see this:

The new recipe (Lily) is skewed toward the yolk corner, as expected.  Roy’s recipe, notoriously lower in yolks and higher in butter, is also relatively higher in flour relative to enrichment. A recipe attributed to Bernardi online is higher in both yolks and flour. My own recipe is more central with yolks and butter and a bit lower in flour.

All of these recipes produce results consistent with this analysis. Modifications to hydration during the mixing process can change the crumb substantially, but panettone does usually stay in a controlled range of hydration. 

I have a number of other ternary plots in development, that should shed more light on the process. For now, I have a few insights to apply to my own bakes and more programming to do! (For instance, gray wasn’t the best color coding selection)..

Also of interest, here’s an Excel radar chart of Ingredients as a % of dough weight (without inclusions):