This project has been germinating for several weeks. We finally picked up momentum, and armed with some more detailed information and opinion, we've begun to sort out this business of stabilizers. Not only have more and more people been talking about it lately, but I've always felt like I was taking a lot for granted by using a commercially blended product. And the basic knowledge of what works and how (and why in a particular synergy) is good to have as it can lead to other things.
The idea was pushed forward when a pastry chef colleague down the street mentioned his dismay at the fact that the most readily available (at least in NYC) commercial sorbet stabilizer contains both whey and gelatin. It has bugged me too, but not enough that it kept me awake at night, that is, until I began to think about it more. It should be a given that our sorbets are, if not dairy free, then at least vegetarian. I promised that I would find the ideal solution for our own knowledge and our conscience.
The first step was to break down the Cremodan stabilizer into its labeled constituents: gelatin, locust bean gum, cellulose gum, guar gum, whey protein concentrate, and dextrose. I have all of these ingredients on the shelf, with the exception of the whey, which I'm trying to get rid of anyway. I wanted to consult a food scientist (who has had a particular interest in ice cream) on what was and wasn't necessary, and what first steps I could take to constructing my own blend.
The advice I got was that the whey powder provided no function whatsoever, and neither did the cellulose gum- carboxyl methyl cellulose or CMC- at least in this application, and when guar and locust bean gums are being used. Hands down, in his opinion, the best single stabilizing ingredient is simple gelatin. It's water-holding abilities, viscosity, mouth feel, and that it melts at body-temperature, all make it ideal. This was awesome, but it deflated my enthusiasm somewhat, as I was hoping to remove gelatin from the equation. His second choice would be locust bean gum alone or in tandem with guar gum. Our dialog on the subject is by no means finished (at least I hope not), but this was a great place to start.
I also wanted to take a look at some other 'handmade' formulas. The one that everybody seems to reference can be found in Francisco Migoya's recent book, Frozen Desserts. Now, I have to confess that I don't own it, and have only skimmed a copy, but I think it's a great resource. The components in his formula mirror that of the Cremodan blend, but at least now we get a sense of their proportions.
The next logical step would be to conduct a simple side-by-side comparison of these four variations: the commercial Cremodan stabilizer, the bespoke blend, gelatin, and finally the LBG/Guar combo. I'll apologize in advance to any real scientists out there; despite our wish to be as accurate as possible, there is certainly some inherent margin of error in our measurements. I also didn't assemble a true control (a sorbet with no stabilizer added), nor did I record every possible variable.
I chose a basic raspberry sorbet as our test subject. Looking at some general usage guidelines, I chose the percentage of each stabilizer based on their common ranges (even on such a small scale- tenths of grams- it could make a difference) while also trying to hit a fairly rounded measurement. We began with the four identical syrups, all prepared in the same manner. The Cremodan (A) was dosed at 2.25%, the Migoya blend (B) at 2.5%, gelatin (C) at 3%, and finally the gum duo (D) at 2.5%. Once prepared, matured, and cooled, I assumed that these syrups would tell us something before fruit was added and spun. As predicted, the A and B syrups appeared identical- slightly opaque but quite fluid. The gelatin-based syrup was crystal clear, with the texture of a just barely set gel. Syrup D was by far the thickest, with a good deal of elasticity; I immediately determined that the locust bean/guar percentage could be lowered by at least a half of a percentage point, if not more.
An equal measurement of each syrup was added to an equal measurement of raspberry puree. Each mixture was spun and then stored in a sub-zero F freezer. And though the anticipation was killing me, I decided to let them sit undisturbed for 48 hours. Only after some degree of storage would we be able to see how the stabilizers do their intended job. The first visual observation would be its volume or overrun. There were no noticeable differences among the group, though I don't know that such a small batch size would give us such clues.
Onto the taste test. I didn't notice any real difference in flavor, with the exception of sorbet D; the slight gumminess did seem to mask the raspberry flavor more than the others. Texturally, D was, of course, the least pleasant as predicted. The gelatin-stabilized sorbet had a more noticeable body and mouth feel than A and B; in terms of texture alone, my friend was right- gelatin won.
The other factor I wanted to look at was it's rate of melting. Here, the results become harder to analyze. The first to begin melting were A and D. After 10 minutes, the order of fastest melting to slowest was D, A, C, and then B. Yet after 20 minutes, the order had changed: B, A, C, followed by D. Taking into account which sorbet had which stabilizer, it made sense that C was among the slowest, yet D didn't contain the gelatin that A and B both had. For kicks, I stuck a thermometer into each fleeting scoop t the 30 minute mark, and surprisingly the temperatures ranged from 50F/10C down to below 40F/4C. I wouldn't think that any of these stabilizers would affect the temperature, and as you can see from the photo below, the scoops sat out within about a foot of each other. I wonder then, why such a broad difference?
So what did we learn? That the gelatin and LBG/Guar versions provided noticeably different results compared to the commercial blend and the bespoke attempt at replicating that blend. Yet on the whole, with the exception of an over-dosing in sorbet D, the sorbets were remarkably similar, more so than I would have thought. The verdict? As yet undecided, at least until I run more detailed tests. But I certainly feel just a bit smarter for having put in the effort!
While I didn't faithfully record the resulting data and observations for you, here are the syrup recipes that I started with:
Download: Raspberry Sorbet.Stabilizer Trial-Workbook 7.3.09