I've been making homemade vanilla extract for a few years now. I like being able to choose the type of vanilla beans and control the production profile. There is some hope that homemade is better than commercial, and in general one expects it to be cheaper. But when adding up the cost of mail ordering bulk vanilla beans (still a lot cheaper than buying one at a time at the store) and buying the alcohol, it doesn't come out cheaper than a quality commercial extract.
This week I impulse-bought a couple extracts from a brand that I think is fairly high quality, and made using vanilla beans I don't have access to from importers in North Amerca. I got them primarily to add variety to my available flavour profiles, but wondered how much that would show through. So I decided to do a blind tasting of the commercial extracts plus some of my homemade efforts. I put the same amount of each extract into the same amount of sweetened milk, with labels on the bottom that I couldn't see and then shuffled them around in a way that I couldn't tell which was which.
The results surprised me indeed. I hoped my own efforts would do well, but I did have high hopes for the commercial ones, because the producers have more experience with extraction and presumably access to better beans. In particular the Tahitian extract, which is made from a different variety of bean than most, smells in the bottle distinctly different from the rest, and I expected to be able to easily identify it and to prefer it in the tasting. Actually, the one that I thought was this one because of a perfume and citrus character, and that I prefered, turned out to be one of my own. The Tahitian extract came in last because I found it wasn't complex enough, so the complexity in the bottle didn't translate when put into food; my homemade extract with Tahitian beans also came in behind my other homemade extract, for the same reason. I also included an experimental extract made with bourbon and expected to identify it by the barrel char note, but never differentiated that at all, although it did come in second-last.
So I conclude there is some value in making homemade extract after all. I did notice an alcohol presence in the homemade ones that I didn't in the commercial ones, and I knew that I was using 40% alcohol to infuse the extract while commercial uses 35%, so perhaps there is value in reducing the alcohol by that small percentage. I noticed a musty character in some of the extracts, but no clear association with homemade vs commercial, except it was extra-strong in the experimental batch, which is not finished infusing, and in fact it was my longest-infused batch that noticeably did not have this character. So 6 months of infusion time seems valuable, though I should note this characteristic may be less noticeable in real-world usage even though I downgraded extracts in this tasing. And planifolia beans seem to bring both stronger vanilla and greater overall complexity than tahitensis beans to an extract even though the opposite is true of the beans themselves.
Here are the notes I took on the 5 extracts I tested, in descending order of preference:
- Homemade 2014 with planifolia beans from Uganda infused in dark rum. Different from the rest. Almost no smell, more perfume taste, almost citrus. Decent hit of vanilla, slightly musty aftertaste.
- Homemade 2015 with tahitensis beans from Papua New Guinea infused in dark rum. Slightly higher alcohol on the nose. Not as complex as some others. Taste a little more caramel, but also vanilla.
- Nielsen-Massey Mexican. Pleasant but not distinctive. Weak vanilla, less other tastes. Longer aftertaste.
- Homemade 2015 with planifolia beans from Madagascar infused in bourbon. Fragrant round flavour. Slight alcohol smell, medium vanilla. Musty aftertaste, would have ranked higher if not for that.
- Nielsen-Massey Tahitian. More floral, more vanilla, musty. Would have ranked higher but not complex enough..