This is a coconut and white chocolate panna cotta made with agar.
The cherry at the bottom is not a cherry at all, but a pearl created by reverse spherification, filled with hibiscus syrup.
At the very bottom is a single synthetic caviar sphere produced by normal spherification. See my post from last year on this subject.
It was all delicious - if I do say so myself.
This dessert was based on a recipe on the Flavor Bender site.
With the panna cotta I followed the recipe fairly closely. I halved the quantities, substituted thickened cream for the half-and-half and agar for the gelatin. I used 0.75 g of agar powder, which is the bare minimum needed to hold the panna cotta together. Apparently this is how a panna cotta is meant to be, but it makes it really difficult to work with. You need to turn it out really gently on the plate and you can't move it around afterwards. Here's a photo of the same recipe using twice as much agar (1.5 g). The panna cotta still tasted good and was firmer and easier to work with, but the softer textured one was definitely better.
For the pearls, I used the syrup from a jar of Wild Hibiscus Flowers in Syrup instead of brewing an hibiscus tea, but otherwise followed the recipe closely. These Wild Hibiscus Flowers, by the way, are intended for dropping into a glass of sparkling (aka champagne) and if you haven't had these before I'd highly recommended trying them.
When I made the synthetic caviar (I made a whole lot more than the single ball shown), I used calcium lactate instead of calcium chloride in the setting bath just to see what it was like. You need to use twice the quantity (by weight) and it's a little more difficult to dissolve, but it makes a noticeable improvement to the flavour and I'll be using it from now on. With the calcium chloride, the caviar has to be washed thoroughly to remove the salty flavour imparted by the chloride, and often some saltiness remains. The lactate, by contrast, is virtually flavourless.
