<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
  <title><![CDATA[Moonlit Neutron]]></title>
  <link>https://moonlitneutron.com/</link>
  <description><![CDATA[Moonlit Neutron is a gelato shop in Vila Madalena, São Paulo that logs every batch and publishes what it finds. 214 flavors since 2019.]]></description>
  <language>en</language>
  <atom:link href="https://moonlitneutron.com/feed.xml" rel="self" type="application/rss+xml" />
  <item>
    <title><![CDATA[Why gelato melts faster than ice cream: a temperature and overrun breakdown]]></title>
    <link>https://moonlitneutron.com/notes/gelato-melt-rate-overrun-temperature.html</link>
    <guid>https://moonlitneutron.com/notes/gelato-melt-rate-overrun-temperature.html</guid>
    <description><![CDATA[Ask anyone who has watched a scoop of gelato collapse into a glossy puddle on a warm afternoon, and they will tell you something feels different about how gelato melts compared to ice cream. They are right — but the reason is more nuanced than "it has less fat" or "it is served warmer." Melt rate is the product of at least three interacting variables: overrun percentage, serving temperature, and the ratio of fat to water in the mix. At Moonlit Neutron, we have been logging batch data across every flavor since we opened, and last winter we finally sat down with 14 months of churn records to see whether the numbers matched our intuitions. What we found surprised us in a few places, and confirmed some long-held suspicions in others. This piece walks through the physics and the data together, so you can see exactly what is happening every time a cone gets handed across the counter.]]></description>
    <pubDate>2025-10-10</pubDate>
  </item>
  <item>
    <title><![CDATA[What Brix actually measures in gelato, and why it matters for texture]]></title>
    <link>https://moonlitneutron.com/notes/brix-gelato-texture-explained.html</link>
    <guid>https://moonlitneutron.com/notes/brix-gelato-texture-explained.html</guid>
    <description><![CDATA[Most people who eat gelato think about flavor first, temperature second, and almost never think about the ratio of dissolved solids sitting quietly inside every scoop. But that ratio — measured on the Brix scale — is one of the most consequential numbers in a frozen dessert kitchen, and it explains why two sorbets made from the same fruit can taste completely different by the time they reach the customer. At Moonlit Neutron we started keeping systematic Brix logs about two years ago, partly out of curiosity and partly because our wild-berry sorbet was melting into a puddle before it cleared the pickup counter. What we found surprised us enough to write it down. This piece is that write-up: a plain-language explainer of what Brix actually measures, followed by a walkthrough of real numbers from 34 batches in our flavor archive showing how Brix correlates with melt rate, scoop hardness, and the sweetness signal your palate receives.]]></description>
    <pubDate>2025-04-19</pubDate>
  </item>
  <item>
    <title><![CDATA[How we source milk from a single farm 65 km away and what it changes]]></title>
    <link>https://moonlitneutron.com/notes/single-farm-milk-gelato-sourcing.html</link>
    <guid>https://moonlitneutron.com/notes/single-farm-milk-gelato-sourcing.html</guid>
    <description><![CDATA[Most gelato shops list "fresh milk" as an ingredient and stop there. We wanted to know what that phrase actually means when you trace it back to a specific field, a specific herd, and a specific season. Fazenda Boa Vista sits about 65 km north of our lab in Atibaia, São Paulo state, and we have been receiving milk deliveries from them twice a week for just over two years. What surprised us — and kept surprising us — is that the milk is not the same milk every week. The fat percentage drifts with the season, the pasture, and where the cows are in their lactation cycle, and every one of those shifts shows up in the finished gelato in ways you can measure and, if you know what to look for, taste.]]></description>
    <pubDate>2026-06-13</pubDate>
  </item>
  <item>
    <title><![CDATA[The flavor archive: 214 gelatos ranked by customer rating and repeat batches]]></title>
    <link>https://moonlitneutron.com/notes/flavor-archive-214-gelatos-ranked.html</link>
    <guid>https://moonlitneutron.com/notes/flavor-archive-214-gelatos-ranked.html</guid>
    <description><![CDATA[Every gelato shop has a graveyard — flavors that arrived with ambition and left without fanfare, made exactly once and quietly retired. At Moonlit Neutron, that graveyard turns out to contain 89 flavors, or 41.6% of everything we have ever produced. The other 125 earned a second run, a third, sometimes a twentieth. This piece walks through the full flavor archive: 214 entries, 1,100-plus individual production batches logged between January 2019 and March 2026, and the customer rating data we collect after each batch cycle. The goal is not to celebrate our greatest hits or bury our misses. It is to actually look at the shape of the data and ask what it tells us about how flavors live and die in a small production gelato kitchen.]]></description>
    <pubDate>2025-07-14</pubDate>
  </item>
  <item>
    <title><![CDATA[Locust bean gum vs. Stabilizer blends: what the texture data shows]]></title>
    <link>https://moonlitneutron.com/notes/locust-bean-gum-stabilizer-gelato-texture.html</link>
    <guid>https://moonlitneutron.com/notes/locust-bean-gum-stabilizer-gelato-texture.html</guid>
    <description><![CDATA[In early 2021, Moonlit Neutron's production team pulled a five-kilogram tub of commercial stabilizer blend from the shelf and replaced it with a single-ingredient bag of locust bean gum. The decision was motivated partly by cost, partly by a suspicion that the blend's carrageenan fraction was over-firming the texture at serving temperature. What nobody expected was that the switch would produce measurable, consistent, documentable differences across nearly every texture metric collected over the following year. This piece presents those numbers as clearly as possible, explains the methodology behind them, and names the places where the data is ambiguous or incomplete. No claims are made beyond what the batches showed.]]></description>
    <pubDate>2026-03-06</pubDate>
  </item>
</channel>
</rss>
