Okay. Brett. Let me take those in the order you gave them. So sweet flavors, so our sweet flavors business fundamentally developed as a dairy flavors business. A few years ago, as we've talked about on this call and a lot of our one on ones, we made a very concerted effort and we continue to make a very concerted effort to diversify that business into other categories such as bakery and confectionery. In some cases, we had some presence in those segments. But in other cases, we did not. So a lot of our effort in dairy, we sort of recognize some of these declines that could grip the dairy industry, particularly in North America, and so we really made a concerted effort there. A lot of where our wins are coming today are coming from bakery and confectionery-type projects. We're able to profile a lot of our sugar reduction platform. We're able to provide integrated products that also include color. For many customers, this is a very compelling aspect to development because they either lack some of the technical resources of a larger player, or they simply -- and/or they want to move very, very quickly. And so having a supplier like us who can provide color and flavor, which tend to be among the more challenging ingredients to formulate given their high intensity, in particular, in these types of applications, it's proven to be a pretty good approach that we've taken there. So our work will continue on that front. There are definitely, just to go back to dairy in general, there are definitely signs of improvement. You'd see in some of the trailing data the last 12 months, ice cream has actually been up as a category, which is a nice sign. But then you'd also see that yogurt has been down, in fact, 7% to 8% in the U.S. market. Now some brands are up, some are down. But in general, the category is down and then there are customers and products in that category that are significantly different from that average. So the program will continue to be dairy is a good business, in general, the market can grow over time. They're just -- we're kind of in a cycle right now, we're in a bit of a trough. But we're certainly not going to abandon dairy and there are definitely applications where we can begin to introduce more sugar reduction, more natural ingredients to that particular category, which I think we will be helpful. But the bigger play there is how do we continue to diversify into these other very interesting, and in some cases, particularly the bakery case, fast-growing segments. On your second question, you had a question about how is Sensient closing this cost to performance gap in natural colors. And yes, it's a very -- it's, in some cases, it is -- we've been very, very effective in terms of expanding our shade range, colors that can be suitable for different applications. So part of the gap here is just based on technology. Can you get the shade? Can you get the shade to be as richly colored as a synthetic shade because as I mentioned in my prepared comments, customers in the North America -- in the Americas and most of Asia are not going to accept a natural solution that looks anything less than a synthetic solution. And there is this talk about somehow natural color needs to look differently or if it's more whitewash and water washed, it will somehow communicate naturalness. Well, not in the eyes of the consumer. When the natural color does not match the synthetic color, the customer doesn't buy their product because more often than not, the customer thinks the flavor has changed. So color is totally linked to flavor perceptions in the minds of the consumer. So to go back to your question, how do you close this cost to performance gap, technology is a big part of that. The supply chain is another big part of that. Our acquisition of GlobeNatural was done very much with this concept in mind. How do we get more vertical on some of the key raw materials? Not all of them, but some of them that can lend itself to better economics surrounding supply -- surrounding natural colors, but also -- certainly, there's elements of availability, too, that also factor in there. But I think the other piece here is, there's a real distinction between products that are synthetically colored and they're going to be naturally colored and then new-to-the-world products. And new-to-the-world products are converting or they're being introduced on the market at a rate of 80% with natural colors. The balance of the market that has not converted or is not using natural colors is faced with this question of how do I convert and how do I make these economics work. I think that is a challenge for some customers, but it's certainly one that, again, going back to these 3 points, technology and supply chain being a big part of that, that is very much in the last 8 or 9 years, those 2 factors have gone a considerable way towards narrowing this gap between performance and cost. To your point about margins, I think the best way to explain this on natural colors is, yes, on a -- if you've a synthetic red, you want a natural red, in general, it takes more natural red to achieve that synthetic red, which translates to more volume, maybe that's at a lower gross margin. But on the EBIT margin, you should see a relatively fixed SG&A across that sale, which is to say, it should be EBIT margin neutral. But perhaps, it could be gross margin dilutive in certain cases. And then your last question, Brett, I'm not sure I totally caught it. I think it was about soy. I think Steve will take that one.